EC.710 | Spring 2010 | Undergraduate

D-Lab: Medical Technologies for the Developing World

Projects

Project Documentation

During spring break, students traveled to Nicaragua to identify public health or medical needs that might be effectively addressed by appropriate medical technologies. Students interviewed key stakeholders, photographed target medical devices, and completed the Design Challenge Template to document their projects.

Photos and videos from the 2010 Nicaragua spring break trip

Upon returning from Nicaragua, students identified five final design projects. They formed teams, and were advised by expert mentors and course instructors. The second half of the semester focused on device design and prototyping that leveraged lessons learned from the first half of the course. The following blog entries catalogue student progress and final presentation of the group work. Teams created posters and displayed prototypes during a capstone poster session at the MIT Museum at the end of the semester.

Project Blogs

Ambuzap

BabyTrackr

EyeHeme

NebShair

Swiss Army Electrocautery

Portable low-cost defibrillator rechargeable through the ambulance’s power.

Team: Divya Srinivasan, Michael Melgar, Krithika Shanmugasundaram, and Swetha Kambhampati

This content is presented courtesy of the students and used with permission.

1 2 Next »

Hello to our Favorite Superheroes!
by Krithika Shanmugasundaram

Just wanted to give a quick shout out to all of D-Lab Health Spring 2010 and announce the presence of the Ambulance Team’s blog. I encourage you all to post your unfiltered thoughts and suggestions on this. We would appreciate any input from people that are as committed to saving the world as you all are.

One request: please post your favorite super-power at the end of every comment.

A quick piece of background and foreground on our project:

The motivation for our project was the simple (yet shocking, for me anyways) reality that hospitals in Nicaragua strip the ambulances of equipment because they claim they have a greater need for it. However, this makes it difficult to give care to patients wile transporting them, which is often necessary, as in the case of severe bleeding, cardiac arrest, and mothers in labor.

Our challenge is to incentivize the retention or, at least the return, of ambulance equipment. Ideas have included making modular instruments/equipment that would increase the ease of returning equipment to the ambulance and also make it look like it belongs more so in the ambulance than the hospital. This also combats the hazard of loose wires that are hanging from the inside of the ambulance since the equipment is literally ripped from walls it is attached to.

Photo of an ambulance interior with no visible medical supplies.

Photo by Anonymous MIT student. Hospitals had stripped the ambulance clean of its medical supplies.

Back to top

Trains of Thought
by Krithika Shanmugasundaram

These were our potential projects:

1) Integrated vital signs monitor- being able to monitor pulse, O2 saturation, respiratory rate, blood pressure, and temperature are necessary when transporting an unstable patient to the hospital. If all of the instruments needed to do this were connected to and wired into one box, it would be cumbersome and wasteful for the hospital to take the box and keep it, because they may only need to use one or two of the functions at a time for particular patients. It might also seem out of place in a hospital where large, specialized, fancy machines reign and bring pride to the physicians and nurses that provide care there.

Problems: We figured out that the ambulance usually took a doctor or nurse with them on calls, and these medical professionals had the ability and brought the equipment to monitor BP, pulse, and temperature. This makes 3/5 functions of the “vital signs box” unnecessary. It would not be effective to just have two vital signs monitored in this device because it would not provide enough incentive to keep it in the truck.

2) Low-cost AED/Defibrillator – we were inspired by the untouched EKG at the hospital in Ocotal. Although functional, people were frightened by its complexity and potential dangers with unintentional misuse and we also found that no one knew how to use it. If used properly, a defibrillator could save the lives of cardiac arrest patients, who often cannot afford to wait the 15-25 minutes it takes to drive to the hospital to get care. This is an area where en route care is a priority and can be the life-saving step.

Problems: There are dangers with working with a high-voltage circuit. The circuitry is difficult in itself, and would be more than a 4-week project.

3) Making the Defibrillator specific to the ambulance – the idea was to design a system that would be able to convert the 12 V from the cigarette lighter in the car to the 200 joules needed to deliver the adequate number of shocks (4-5) to a cardiac arrest patient. The cigarette lighter would be the only way to charge it, so it would have to make its way back to the ambulance eventually, but patient care would not be compromised, since the charged defib could be transported with the patient if needed.

-Note: other ways to provide incentive to keep the defib in the ambulance are 1) to create a flashing light or annoying beeping sound that goes off when the device is kept away from the ambulance for too long a time 2) or to actually disable the defib when it spends too long away from the ambulance.

Problems: The hospital may still just keep the defib, since it won’t run out of battery all that quickly, since it will not get used that often. The issue of making a vehicle-specific charger has also been done before (airplane flash lights charged in holder bound to the plane), so this aspect of the project would not be that interesting.

4) O2 purification system – one of the simplest items of an ambulance that can do a great amount of good in a variety of situations is the O2 tank. The shortage of O2 tanks could be compensated with an O2 generator/extractor. The idea was proposed as an “O2 catcher” that would sit on top of the ambulance and as air sped by it would only bind the O2 and keep it in a container. Our challenge is to use electrolysis to produce the pure O2, collect it efficiently, and pressurize it so that a large volume can be held (15 L/min is standard flow rate, and with an estimate of 15 min transport time, it comes out to 225 L, at least per transport)

Problems: We are in the process of hammering this one out and doing the calculations (to come soon!) to figure out how much water will be needed to produce this O2.

Those are I think projects we thought about, mostly in the order we discussed them.

{Spidey- sense}

Back to top

Conversation with Mario Luque Aguilar (83685544)
by Michael Melgar

Sivakami and I spoke on the phone with the ambulance driver we met in Sabana Grande. We spoke about the situation with cardiac care patients he deals with and about the need for oxygen on the ambulance. Interestingly, he said that he does run into shortage because there is an 8 day turnaround time for an empty tank to be shipped, filled, and returned. He usually has no oxygen available, and he estimated it is needed 2 or 3 times per week. Because of this, having a system that automatically and sustainably generates oxygen on the ambulance would be helpful. Perhaps something built-in that uses the car battery to power a device that forces the gas into a pressurized container.

I actually still prefer working on the defibrillator, though, and there are also excellent reasons to do so. Mario said that he transports about 5 cardiac arrest patients per month. These patients have no access to a defib until they reach the hospital in Ocotal or in Somoto, which takes 30-40 minutes (to the health post at Sabana Grande) plus 20 minutes (to Somoto). The nurse on board treats these patients with ‘chest massages,’ which I took to mean CPR (is this correct?). As an anecdote, he transported the same patient 3 times since our visit to Nicaragua. This man apparently suffers heart attacks almost once a week. The first two times, he was transported to the hospital, but the third time stabilization at the health post was enough. This makes me wonder if we’re talking about the same thing when we say ‘heart attack’. He also said that patients always make it to the hospital in time to be defibrillated and saved. Can 40 minutes of CPR save a heart attack patient? The symptoms he described seemed to be those of a heart attack, though: shortness of breath, loss of consciousness, convulsions, and even coughing up blood.

Some infrastructure related information we collected: Each municipality has two ambulances. Totogalpa uses one for patient transfers between health posts and hospitals and the other for emergency cases. If Mario had his way, he would simply weld a defib into the ambulance, making it impossible to remove. Lastly, we got the contact information of three other people: Ismael, the ambulance driver in San Juan de Rio Coco (88536960); Dr. Blanco, the director of the Totogalpa health post (84005324); and Dra. Janet Veillez, another doctor who periodically makes visits to the health post and has a contact in the U.S. (86282017). The next step will be to follow up with one of these people. He/she may be able to clarify what is meant by ‘heart attack’ and provide feedback on our ideas.

{I can spell ANYTHING.}

Back to top

People
by Divya Srinivasan

We haven’t been having much luck lately in contacting people. On Thursday and Friday members of our group contacted various professionals linked to AED technology. Below are the people contacted and what sort of expertise is associated with them.

People Contacted Prior to April 25, 2010:

  1. Varsha Keelara: A medical student at HMS, Varsha has a lot of experience dealing with AEDs. She also has a lot of contacts who she could refer us to, but unfortunately we haven’t heard back from her yet.
  2. Dr. Robert Malkin at Duke University: Dr. Malkin is a professional at Duke who has made a circuit that can test the voltage of the defibrillator. Our assumption is that because he has dealt with defibrillators before and high voltage circuit components, he will be able to tell us how one works or how to program the different components.
  3. Ed Boyden: He is a professor at MIT’s Media Lab and has expertise in various areas of biomedical engineering. He is an electrical engineer which will allow us to gain a better understanding of how to make circuits.
  4. Professors of 6.022: All 3 of them. We contacted them because we talked to Aubrey, a student of the class and an EMT. She talked about how she had experience with EKGs from 6.022. We’re hoping that they can help us figure out how to detect and analyze fibrillation.
  5. Jon Wu: A medical student at Tufts University, he has a fair amount of knowledge about EKGs and defibs.
  6. Aubrey Samost: An EMT and a student in 6.022 (refer to above), Aubrey gave us some helpful regarding how EKGs detect fibrillation and how it calculates the different conditions of a patient. She is a good bouncing board for ideas and could direct us to professors she has had in the past.

Here are the Questions we Asked them in our Emails:

  1. What does an AED look for?
  2. How does the AED detect the signal?
  3. How do the leads interact with each other to detect the signal?
  4. How do you analyze the signal? What does fibrillation look like versus a normally beating heart? Healthy v. shocked v. you’re a goner.
  5. Do you have any ideas for keeping a defib in an ambulance?
  6. Do you know anybody else who would be able to answer our questions?

We haven’t received responses from any of these people and are hoping to either get email responses back from them or call them directly if we can find their phone numbers. If you have any ideas of contacts we should get in touch with or any questions we should ask these professionals - that would be great! Until tomorrow, this is the “happiness-G” signing off!

Back to top

Conversation with Dr. Sergio Shkurovich, PhD
by Krithika Shanmugasundaram

We called Dr. Sergio Skurovich, the director of International Regulatory Affairs at St. Jude Medical. He works in the cardiac rhythm management division and he had some good advice for us in our pursuit of the AED.

In our e-mail correspondence, he clarified that a defibrillator senses electrical signals measured from the skin. These vectors are similar to those in an ECG, but different to the standard derivations used to collect a surface ECG:

“During a normal rhythm (called sinus rhythm), there is a clear sequence of events marked by atrial activation (P wave on the ECG) followed by ventricular activation (QRS complex on the ECG) and finally ventricular reset (T wave on the ECG). When the ventricles fibrillate, the normal electrical pattern P-QRS-T disappears and is replaced by a less organized, higher frequency and lower amplitude signal.

Sinus Rhythm: http://www.ecglibrary.com/norm.html

Ventricular Fibrillation: http://www.ecglibrary.com/vf.html”

On the phone we discussed what materials we would need to build the sensors on the defib pads and he indicated that the electrodes detect polarization and repolarization, and are usually made of silver and silver chloride. He further instructed us to take guidance a chapter from the book Medical Instrumentation: Application and Design by John Webster called “Biodetection Electrodes.” Conveniently available on Amazon. Inconveniently over our $100 budget…

Some issues he brought to our attention were that noise could be picked up from electrodes and cause false positive signaling, and that low-cost AEDs were available (from Metronic, WA and Philips for <$1000). However, $1000 is still fairly expensive for each ambulance. The components that contribute most to the cost are: a reliable, rechargeable battery, good capacitor, speedy charge delivery, and an interface for the user (i.e. voice box). These will most likely be the aspects we will focus on.

We then asked if he could direct us on how to perform signal processing and if there was a quantifiable method of detecting V-fib. He suggested we could try using frequency along with amplitude and the QRS-complex characteristics to determine an algorithm for indicating a necessary shock. He also mentioned that there is lots of literature available on this topic and we should go through that for more specific details and instructions. Programs we could use to automate the signal processing include MATLAB®.

The question of whether to make an AED vs. ED was then brought up. Sergio proposed perhaps having a laminated “instruction sheet” on the machine that clearly shows the graph of a V-fib reading and normal heart reading and instructs them to only shock when it is clearly in V-fib.

After discussing with Jose, we decided to work on an ED, perhaps using a flash from a disposable camera as our “defibrillator.”

{spidey-sense}

Back to top

"Day" Flashes
by Divya Srinivasan

Photo courtesy of David Gossard, from the MIT OpenCourseWare site 2.003 Modeling Dynamics and Control I, Spring 2002. The electric circuit from a disposable camera.

In D-Lab yesterday, we channeled our inner 3-year olds and demolished (gently) a disposable camera. What we found was a beautifully exquisite piece of circuitry with a pretty big capacitor to charge and discharge the voltage from a double-A battery. Though Michael got shocked a couple of times in the process, we found out that once the battery was in place, the capacitor was charging continually. When we connected a wire from the capacitor to a metal piece on the circuitry, the capacitor discharged and there was a resulting flash from the light attached to the circuit board.

Our goal for tonight is to attach a resistor to the capacitor so that it will cause the capacitor to charge slowly and reduce the hazard of shocking anybody. Michael and Krithika were able to meet with one of the 6.022 professors today afternoon (more to be posted later!), so we’re going to see if we can start working on the coding/sensory aspect of our defib.

What we’ve completed so far:

  1. Picking apart a disposable camera, isolating the circuit, and making the capacitor discharge to create a flash!
  2. Consulting professors on how exactly we’re going to gather data from a person and how to analyze that data to create graphs of a patient’s signals.

{Happiness-G}

Back to top

Making Progress is Wonderful! :)

We’ve made significant progress in the past couple of days– it’s pretty exciting! Yesterday, we started to make our own differential amplifier in a course 6 lab in building 38. Using the circuit diagram that we obtained from Maysun (see below), we started to piece the diff-amp together. We will be using this differential amplifier to take the voltage difference between the two leads and amplify it 100 fold. This will serve to increase the signal taken from the leads so that the computer can then analyze and produce the graphs on the PeggyBoard (as the output). Our goal for today is to finish wiring the diff-amp, acquire the leads for the defib, and the PeggyBoard, and start coding the program that will analyze the signal to produce the output graphs.

In a couple more blogposts to follow, we will detail our conversation with 6.022 instructor, Professor Mark, and other email communications we have had in the past couple of days.

Back to top

First Attempt at the Differential Amplifier and Noise Filters Done
by Michael Melgar

After long, arduous hours at D-Lab, Divya and I have built three of the circuits needed for a functional ECG. Whether or not they have been assembled correctly remains to be seen with use of an oscilloscope with Paul’s help. The differential amplifier is designed (by choice of the resistances used) so that it takes in a signal from two points and subtracts their voltages before multiplying by 100. The inputs will be connected to the electrodes attached to the patient, so our signal of ~1 mV should be amplified to ~100 mV. At times the circuit was subtracting input voltages, but failing to multiply by 100. Testing with only a multimeter, I was unable to troubleshoot the problem, but using an oscilloscope should make it much easier. (Paul’s expertise can’t hurt either).

Back to top

1 2 Next »

Labor contraction monitor for Nicaraguan hospitals

Team: Maysun M. Hasan, Grace Yao, Karina Isaak, and anonymous MIT students [LT] and [AB]

This content is presented courtesy of the students and used with permission.

1 2 3 Next »

Welcome to the Contraction Monitor team blog!

by Grace Yao

By the way, should we try to come up with a catchier name, maybe like Laborsaver or something like that? Just a random thing I thought of!

In any case, we are [AB], Maysun Hasan, Karina, [LT], and Grace Yao, and here is a little bit about the problem we are trying to solve and our goal for the solution.

In developing countries, such as Nicaragua, maternal health is very rudimentary in rural areas. Usually community health care workers make house visits to check up on the mothers, but they have limited knowledge and resources for proper maternal care. Also, mothers have to decide when to go to the health posts to address their concerns if any problems arise. When labor gets to a certain point, the ambulance has to be called to transport the mother to a hospital, which is often very far away. Often, it is hard for the mother to determine when she needs to call an ambulance. In worst-case scenarios, the mother doesn’t call in time for them to reach the hospital, and they must deliver the baby on the side of the road. To avoid such cases, we have proposed to develop a monitor to automatically alert different stages of labor based on contraction rate. The monitor should signal for when a contraction occurs, its length and rate, when the ambulance needs to be called, when critical attention is needed, when non-critical attention is needed, and when the baby needs to be delivered. The signals (in addition to the one for calling the ambulance) may aid health care workers in quickly identifying which mothers need attention at a birthing facility, helping them triage the patients.

Back to top

April 15 Post

by Karina Isaak

Hi team,

I just found this article which was published in May 2009.

Haws, Rachel A., Mohammad Yawar Yakoob, Tanya Soomro, et al. “Reducing Stillbirths: Screening and Monitoring During Pregnancy and Labour.” BMC Pregnancy and Childbirth 9, 2009.

It gives a review of the interventions used worldwide to prevent stillbirths.

Please, take a look at the section called

Monitoring in Labour  
Use of the partograph

starting on page 33.

They are talking about monitoring the contraction with a partograph, That’s kind of what we want to measure with our device. In the results section they say that even though they couldn’t find significant differences in maternal or perinatal outcomes with the use of partograph versus no partographs, it can still help in low-resource settings:

“Partographs may be comparatively more effective in low-resource settings, as the studies from Africa and Mexico in the Lavender review [140], as well as data from Southeast Asia [146] that showed reduced Caesarean section rates with use of the partograph and early intervention for slow progress of labour. The data from Southeast Asia and Indonesia also showed trends toward improved birth outcomes [146].” (Haws, et al., 2009, p.34)

Their statistical analysis was probably insignificant because of “unclear guidelines on the partograph use” (Haws, et al., 2009, p. 35). Therefore, building an intuitive device that is working with the partograph could possibly help to prevent stillbirths in developing countries.

What do you guys think?

Back to top

April 16 Post

Hey team,

Grace, I agree, we definitely need a more creative name… let’s brainstorm about it in class today. Here is a study comparing the EMG contraction monitor to a tocotransducer belt. EMG seems to be the better way to measure contractions.

Maul, H., W. L. Maner, G. Olson, et al. “Non-Invasive Transabdominal Uterine Electromyography Correlates with the Strength of Intrauterine Pressure and is Predictive of Labor and Delivery.” Journal of Maternal and Fetal-Neonatal Medicine 15, no. 5 (2004): 297-301.

Let’s first test the resistive flex sensor [LT] ordered and then start working on the EMG/EKG if the sensor turns out to be not reliable for measuring contractions.

Back to top

April 19 Post

by [AB]

Hey all,

Hope you’re enjoying the long weekend! I found another interesting article that shows how an external taco meter is just as good as an internal one, although the internal one uses pressure gauges. Just to confirm that we should make both an EMG and an external tocodynamometer. Anyways, I also think we could think of a better name. So I looked up LaborSaver, and apparently it’s already a product that allows workers to save money. How about laborhelper? Regardless, here’s the article.

Bakker, Jannet J. H., Corine J. M. Verhoeven, Petra F. Janssen, et al. “Outcomes after Internal Versus External Tocodynamometry for Monitoring Labor.” New England Journal of Medicine 362, no. 4 (2010): 306-13.

Here’s another article I found on how to make an embedded microcontroller and EMG:

Wu, Han-Chang, Chao-Hung Lin, Shuenn-Tsong Young, et al. “Monitoring Long-Term Uterine Contractions.” IEEE Instrumentation and Measurement Magazine 5, no. 2 (2002): 36-40.

I think it’s very similar to what Maysun sent out a couple days ago.

Let’s meet up some time (I think we’re agreeing for tomorrow 3 pm) to discuss specs and how to make both of these.

Back to top

Patents for Various Contraction Monitors

by Grace Yao

Apparatus for non-invasive monitoring of uterine contractions  
United States Patent 4989615

From Abstract: “An apparatus which includes a bladder element (10) which is at least partially filled with fluid, a belt-like element (12) which holds the bladder against the patient’s abdomen with some pressure, and a pressure monitoring device (18) which is connected to the bladder (10) to detect changes in the pressure of the fluid in the bladder (10) as the abdomen hardens due to uterine contractions.”

External uterine contraction monitoring device  
United States Patent 5070888

From Abstract: “An improved monitoring device for externally monitoring labor contractions preceding childbirth which does not require the use of a belt is disclosed consisting of a transducer removable assembly fixed to a base adhesively attached to the abdomen of the woman.”

Disposable tocodynamometer with self-adjusting bellows  
United States Patent 5224490

From Abstract: “A non-invasive, disposable, self-adjusting tocodynamometer (10) for monitoring uterine contractions of a patient during pregnancy, labor, and delivery The tocodynamometer includes a pressure-sensitive, fluid. filled bellows (20) responsive to changes in the hardness of the uterus during contractions The bellows has one face which projects into the patient’s soft tissue in the abdomen and adjacent the uterus, rending the tocodynamometer sensitive even for obese patients. A plate (22) supports the bellows and provides structure for attaching the tocodynamometer to the patient. A wall (30) formed on the plate receives the bellows as it is compressed during use. A conduit (14) connects the bellows to a pressure transducer (12) which, in turn, is connected to a monitor (16). The bellows, conduit, and pressure transducer form a closed system containing the working fluid.”

Back to top

Retroactive Post

by Grace Yao

This is where we were 5 days ago, before the meeting with Jose:

Below is a list of Ob/Gyn’s in Boston that I grabbed from thecityofboston.com. I think we’ve pretty much decided on using the simpler on/off indicator as opposed to measuring force with varying intensity. So what we really need to find out now on the medical side of things is:

What are the times we have to measure to know the things we want to indicate, which are

  1. when the mother should go to the hospital
  2. when the mother needs attention but not critical
  3. when the mother needs critical attention
  4. when the doctor needs to be there to deliver the baby.

Then the 5th thing we’re having a light for is an indicator of when she’s actually having a contraction, which is pretty straightforward.

Other things we have to figure out:

Materials: what kind of material for the squeezy part? The wire connecting the ball/wristband should have a covering too right? What is the wristband going to be made out of? We want to look into refractive materials so that we can just use one light to light up a lot of the wristband.

Smarts: Microcontroller? Something else? Maysun, can you maybe summarize what you talked about with Paul?

User Interface: How much information should the mother have? How are we going to display it?  
(Number/graph/both?) Do we need to design a radio component too? Would sound be better?

Acker, David B MD - Brigham Women’s Hospital

75 Francis St Ste ASB1

Boston, MA

(617) 732-5445

Etkin, Masha J MD - Vincent Obstetrics/Gynecology Associates

32 Fruit St Ste 4E

Boston, MA

(617) 726-1753

Gomez-Carrion, Yvonne MD - Beth Israel Deaconess Women’s

330 Brookline Ave Ste KS205

Boston, MA

(617) 667-2952

Johnson, Kim M MD - Kim M Johnson MD

500 Brookline Ave Ste E

Boston, MA

(617) 732-6399

Moody, David B MD - David B Moody MD

45 Francis St

Boston, MA

(617) 732-6389

Perkins, Rebecca B MD - Rebecca B Perkins MD

85 E Concord St

Boston, MA

(617) 638-8000

Rodriquez, Elisa MD - New England Medical Center

860 Washington St Ste 2

Boston, MA

(617) 636-6114

Taylor, Faye - Boston Medical Center

91 E Concord St Ste 6

Boston, MA

(617) 414-5461

Wakamatsu, May M MD - May M Wakamatsu MD

55 Fruit St Ste 148

Boston, MA

(617) 726-2000

Yum, Mimi MD - Mimi Yum MD

330 Brookline Ave

Boston, MA

(617) 667-0478

Back to top

Team Info, Revised Problem Statement, Some Early Brainstorming Thoughts (Pugh Chart, Design Specs)

by [LT]

Hi! This is also sort of a retroactive post. We are the Contraction Monitor team!

When a woman goes into labor, clinics are not equipped to monitor her contractions- an important indicator of her stage of labor and any complications. Health workers currently rely on their hands on the woman’s stomach to time the contractions, fetal heart rate, and feel for the relative intensity. This can be a time and labor intensive method, especially when 8-10 women are in labor at the same time in the same room. We have proposed to develop a monitor to automatically alert different stages of labor based on contraction rate. Additionally, it will give the relative intensity of the contraction and output a graph with the contraction frequency. The monitor should signal for when labor has started, when a contraction occurs, its length and rate, when critical attention is needed, when non-critical attention is needed, and when the baby needs to be delivered.

Some thoughts from early brainstorming:

Should we be wary of giving to much info to mother? consider external monitor

Distributed beforehand? when starts feeling contractions, puts it on, tells her when she needs to go to hospital based on timing (how much time she has left, etc.), then switches to more accurate mode with wireless broadcast. Have extra mode or signal for when to go to hospital?- Since we’ve revised our problem statement, this is no longer a priority need!

Talk to potential users about what signals to use, and whether she rather know or now know, and what would stress her out

The iPhone® app contraction monitor has start time, end time, duration, and frequency. Has a timer, and history. Another app gives the graph.

Thought for Ball contraction monitor: info sent to doctor’s phone.

See below for our Pugh chart and Prelim Design Specs!

Pugh Chart

Design specifications

Back to top

Pitch

by [LT]

A first attempt at a pitch!

In developing countries, such as Nicaragua, clinics are not equipped to monitor a woman’s contractions while she is labor- an important indicator of her stage of labor. Health workers currently rely on their hands- which can be time and labor intensive, especially when 8-10 women are in labor at the same time in the same room. Our innovation will give health workers the ability to monitor contraction rate and alert them to the different phased of labor.

When the woman is picked up from her home, the health worker places a band with a monitor around her wrist. Attached to it is a ball that the woman squeezes when she has a contraction. One of five signals will light on the wristband to indicate what the woman is experiencing, such as a contraction or complication that needs immediate attention and a contraction rate will display on the monitor. Especially when only one health worker is present in the ambulance truck and must drive, the visual cues will help health professionals know how to respond immediately.

Back to top

Re-direction of our Project

After meeting with Jose for feedback on our project, we began to change directions (and revise our problem statement). Instead of working on a squeeze ball monitor alone, we’re also prototyping a belt (testing various sensors) and comparing that with the squeeze ball. Ultimately, it will become a behavioral study.

Summary of Our Meeting:

We went over what we were planning to do and Jose began to push us to try the belt idea. He’s concerned that since the woman control the input with the ball, she can either be too urgent and squeeze it too many times to go to the hospital or not squeeze it enough. He encouraged us to do a behavioral study comparing the ball and belt to see the correlation or accuracy of signals (if we find the ball is just as good and even cheaper, then we go with the ball, for example). Basically he told us to explore both options. We also discussed technical details and the sensors. He thought what we were doing was…settling for too little and we could do more. He encouraged us to explore more bio sensors, homemade EMGs (suggested by Maysun), and other flexi sensors and to really talk to Professor Frey. Maybe even look at making our own sensors if we think we can make them cheaply.

We went over what the problem really was and it’s actually that they don’t use any monitors and need something in the hospital (not necessarily at the home to tell them when to go). The current are too expensive, so we need to make it affordable. He told us to contact Ken Endo at the biomechatronics lab, and sent us an email introducing us to some Nicaraguan contacts I think.

So:

  1. Explore homemade sensors and other sensors, continue contacting people, figure out the smarts with a computer first
  2. Design: make prototypes (ball and belt) and turn comp into smarts with micro controller
  3. Clinical: test it out, behavioral studies

Back to top

1 2 3 Next »

Non-invasive anemia diagnostic based on reflectance spectroscopy from the eye

Team: Seema Kacker, and anonymous MIT students [AJ] and [KK]

This content is presented courtesy of the students and used with permission.

1 2 Next »

Design Challenge/Problem Statement/Design Specs/Selection Matrix

Problem or Need

Pregnant women in developing countries are often anemic and do not have an inexpensive, widespread, reliable way to diagnose the severity of their anemia. Without these diagnostics, the mother and her child are severely at risk. The most effective device in the developing world, the HemoCue, is expensive and invasive. Other methods involve some expertise in diluting solutions or matching exact colors on a colorimetry scale.One approach is to apply pulse-oximetry principles and technology to measure altered blood flow due to anemia, which is similar to an existing technology, the OrSense, in the developed world. This provides a non-invasive procedure that reduces risk of infection and increases the patient’s willingness to complete the exam. We intend to create an Arduino optical blood analyzer platform that is able to, at the least, record pulse-oximetry data and anemia data. A possible modular add-on to the platform is malaria detection via magnetic principles, a non-invasive technology that is currently being researched. The platform is designed for local innovation–opening doors for new ideas to analyzing blood samples via optics or magnetism.

Background Information

(Why do you use this equipment, what is the treatment that is sought, what does the entire process entail):

This equipment can be used any time a pulse ox or anemia diagnostic is needed. It is especially useful in the field or in ambulances, since it is portable. The device is a finger clip optical sensor, attached to a battery-powered Arduino. The sensor is clipped around the patient’s finger. The SpO2 and the Hb level are printed on an LCD screen. There is also series of 3 LED lights: red (severe), yellow (mild), green (OK). One lights up based on the readings: if the patient is severely anemic, then the red light will turn on. As long as the device is on a patient’s finger and turned on, it will take measurements once every thirty seconds. A one-time check can also be performed. The pulse-ox and hemoglobin functionalities can also be turned on separately.

Technical Description/Specifications

Readout declares:

  • 9.5 < Hb < 11 = mildly anemic.
  • 7 < Hb < 9.5 = moderate anemia
  • Hb < 7 = severe anemia. AND/OR similar using hematocrit. (Hct ~ 3 * Hb).

Correlation value w/ established lab test, r > 0.75

Sensitivity > 80%

Specificity > 60%

< $100 for device (could be subsidized w/ US sales)

<$0.50 per test (average maintenance, power consumption, sterilization, etc)

How is the Local User Community Approaching thePproblem? What Type of Improvised, or Local Solutions are Being Used?

Health posts and clinics without anemia diagnostic equipment examine patients’ eyes, lips, palms, and nail beds for signs of anemia. However, this method requires training that not all medical personnel have, and only catches severe anemia. Some clinics may have WHO Hemoglobin Color Strips, but the one we visited did not. Hospitals tend to use invasive techniques that are accurate, but also time- and laboratory-intensive. Some hospitals have HemoCues, but when these break, there is no way to repair them.

Who is the Primary Contact for this Challenge? Who are the Key Stakeholders (Do you Have Their Contact Info? If Not, Get It…)

Patricia Coffey from PATH

What Relevant Resources are Available?

Consider materials (not just local materials, if they are imported, then just note it. Don’t get hung up in price. Cost is best defined as how difficult it is to obtain. Difficulty MAY include price, but may also be affected by other issues. Also consider resources to obtain relevant training, and frequency of supply. Use the parameter table.

Small electronics (regular LEDs, LCD screen), plastic. There are folks who know how to mess with electronics; most likely could figure out the Arduino, if they needed to fix it/change it. Nurses/doctors know how to read Hb and SpO2 measurements.

What resources may be needed?

IR and UV LEDs, Arduino

What are the Potential Benefits of Solving this Challenge?

More people will be diagnosed with anemia, since there is no need to wait for sterilization or an additional shipment of disposable cuvettes. This will especially help pregnant women, as anemia has been linked to both infant and maternal mortality during pregnancy. If the platform is extended, more people will also be diagnosed with malaria, which will help stop the spread of resistant malaria parasites, as well as ensuring that more victims get the treatment they need.

What are the Potential Obstacles?

Since the device is so transparent, it may be difficult to ensure standard performance across all devices. It may also be difficult to ensure accuracy, since we are not using highly-calibrated medical equipment. Making the device battery-powered may also be unexpectedly difficult. If the device uses too much power, then batteries could be impractical. Although the device is non-invasive, it is still “looking inside of you”; patients could still be fearful or wary.

What are the Risks of Undertaking this Project?

If the device is not specific enough, we could waste treatment on patients who are not really anemic. If it is not sensitive enough, we could deny patients the treatment they need.

How can you Get the Local User Community Involved in the Process?

Ask for their input – is the readout confusing? Why exactly aren’t there more anemia diagnoses? Is the lack of training, the lack of equipment, or invasiveness the biggest problem? We can then focus on the aspect of our device that is “most important.” Also, what blood diseases are most difficult for you to diagnose (in terms of time, skill needed, etc)? These diseases would be prioritized as we extended the platform.

We could also possibly involve university students – show them how to experiment with the Arduino, and see what they come up with. H-Lab could be involved in thinking of innovative ways to use optical sensor technology. Hospital technicians could also be shown how the Arduino works, so they could repair it, or even innovate upon it, as needs arise.

What Photographs and Videos Should you Take? (Take Them!!)

The photographs/videos of the maternity department of the hospital in Ocatal should be useful. Photos of the ambulance and current pulse-ox technology should be also.

What Additional Information Should you Collect? (Collect It!!)

Would qualitative LED output be useful to the users? What light wavelengths do we need to measure hemoglobin? How does the magnetic malaria detector work – what components will it need to work in our platform? What other diseases can be diagnosed optically? Can all of these light sources and magnets fit onto one clip?

Problem Statement

Anemia is a condition defined by an insufficient total volume or quantity of red blood cells, or an insufficient amount of hemoglobin in these cells. It affects nearly 2 billion people worldwide, and is associated with morbidity and mortality (WHO, 2004). Pregnant females are at an especially high risk of developing the disease – the WHO estimates that, globally, 41.8% of all pregnant women are anemic. Prevalence estimates are even higher within Africa, South-East Asia, and the Eastern Mediterranean. Maternal anemia is often caused by malaria or iron deficiency, and may result in the mortality of the mother or child, or diminished capacity and low birth weight.

This gold standard diagnostic test is an automated complete blood count (CBC), which requires a sample of blood to be drawn and reports 1. concentration of RBCs, 2. hemoglobin level, 3. “mean corpusclar volume”: size of RBCs (measured by flow cytometry), and 4. RBC distribution width. Further testing may be required to determine the exact cause of the anemia (nutritional deficiency, infection, blood loss, etc.) Simple initial diagnostics may focus only on measuring the concentration of hemoglobin in a blood sample. The least invasive procedure, the HemoCue, uses a Hematocrit measurement, and requires a pinprick collection of blood. A Hemotocrit measurement employs an existing color scale that measures hemoglobin in the blood but it is expensive, requires expert lab resources, and comparison to a color chart which is often inaccurate. This is not used widely and surveys indicate that health workers and patients would be more likely to measure hemoglobin if there was a non-invasive way to do it.

One approach is to apply pulse-oximetry principles and technology to measure altered blood flow due to anemia, which is similar to an existing technology, the OrSense, in the developed world. This provides a non-invasive procedure that reduces risk of infection and increases the patient’s willingness to complete the exam. We intend to create an Arduino optical blood analyzer platform that is able to, at the least, record pulse-oximetry data and anemia data. A possible modular add-on to the platform is malaria detection via magnetic principles, a non-invasive technology that is the midst of being researched. The platform is designed for local innovation–opening doors for new ideas to analyzing blood samples via optics or magnetism.

Design Specifications

Readout declares:

  • 9.5 < Hb < 11 = mildly anemic.
  • 7 < Hb < 9.5 = moderate anemia
  • Hb < 7 = severe anemia. AND/OR similar using hematocrit. (Hct ~ 3 * Hb).
  • Correlation value w/ established lab test, r > 0.75
  • Sensitivity > 80%
  • Specificity > 60%
  • < $100 for device (could be subsidized w/ US sales)
  • <$0.50 per test (average maintenance, power consumption, sterilization, etc)

Selection Matrix

EVALUATION CRITERIA NOTES
ON SPECIFICS 		WEIGHTS BASELINE: NIR
TRANSMISSION
(e.g., ORSENSE,
MASIMO) 		CONDUCTANCE
(CHANGE in
ELECTRICAL
COUNDUCTIVITY) 		REFLECTANCE
SPECTROSCOPY
Safety Risk of infection 2 0 0 0
Accurate Sensitivity 2 0 (94%) 0 0
Accurate Specificity (1-false positive) 1 0 (78%) -1 0
Robust Can withstand extreme conditions (weather, temperature fluctuations, transportation) 1 0 0 0
Cheap Total lifetime cost (Upfront cost + per use cost) 2 0 0 0
Maintenance / Sanitation How often, how difficult it is to fix, how much effort to maintain 1 0 -1 0
Reusable   0.5 0 0 0
Portable Can travel, either with health worker or in ambulance 1 0 0 0
Local Mfg Are the parts locally attainable? 0.1 0 -1 0
Speed of Use Turnaround time for patient use, how long it takes to get results 2 0 0 0
Usability How easy it is to use the device and to analyze results 2 0 0 0
Totals (incl. weights)     0 -2.1 0

EVALUATION CRITERIA NOTES ON SPECIFICS WEIGHTS ULTRASOUND/
OPTOACOUSTIC
SPECTROSCOPY 		SPECTRO-
PHOTOMETRIC
IMAGING
Safety Risk of infection 2 0 -1
Accurate Sensitivity 2 0 0
Accurate Specificity (1-false positive) 1 1 1
Robust Can withstand extreme conditions (weather, temperature fluctuations, transportation) 1 -1 -1
Cheap Total lifetime cost (Upfront cost + per use cost) 2 -1 -1
Maintenance / Sanitation How often, how difficult it is to fix, how much effort to maintain 1 0 -1
Reusable   0.5 0 0
Portable Can travel, either with health worker or in ambulance 1 -1 -1
Local Mfg Are the parts locally attainable? 0.1 -1 -1
Speed of Use Turnaround time for patient use, how long it takes to get results 2 0 -1
Usability How easy it is to use the device and to analyze results 2 -1 -1
Totals (incl. weights)     -5.1 -10.1

Back to top

Noninvasive Anemia Diagnostic: Mission Accepted!

So what are we doing here?

Helping the lives of 41.8% of pregnant women. Saving premature and low-birth-weight babies. Reducing the dangers of birth complications. We are… diagnosing anemia – noninvasively!

In many areas of the world, invasive diagnostics are just not feasible. They cost too much, require too much training, cause too much infection, and/or are scary! A noninvasive way to diagnose anemia could save the lives of many mothers and babies – an early diagnosis would encourage a mother to go to a clinic, rather than delivering at home. A woman whose anemia cannot be detected via clinical signs would receive the iron supplements she needs.

Now… how do we do it?

One way that’s been developed (by Orsense) uses NIR transmission – very similar to a pulse ox. This is an attractive method, because you could very easily build combination pulse-ox-hemoglobin devices. This is, infact, what Orsense did. But accuracy is a problem with NIR transmission. Are there better ways to noninvasively measure hemoglobin?

This helpful review article lists a few ways that noninvasive hemoglobin measurement has been explored. Electrical conductance, spectrophotometric imaging… we have no idea how these things work!! Hopefully we’ll figure it out soon, so we can pick a strategy and run with it.

In the meantime….full speed ahead!

Back to top

"These are not Your Children, These are Just Your Ideas"

Big Shift: Diagnosing anemia via the palpebral conjunctiva (inner eyelid) instead of the finger. When we first started the project, Amit introduced us to this amazing thing called the OrSense. It’s non-invasive and is “useful for hematocrit / hemoglobin determination, hidden blood loss monitoring and for anemia screening.”

After watching this product demonstration, I was sold.

What could possibly go wrong with a finger? We already have a platform for a PulseOx, it would be easy. We could make a platform and make it so one could add different components in the future that involve reading blood components by measuring transmittance through the finger.

The first meeting of our team to work on Pugh charts went like this:

Me: “Let’s go with the OrSense!”

Seema: “Wait, wait, wait… Don’t we need to do research first? Find out more about anemia, hemoglobin, CBC (complete blood counts), hematocrit (Hct)….”

It turns out I don’t like research so much. Jose mentioned two types of people in design processes:

  1. Whiteboard fillers. The people who can spend endless amounts of time filling up whiteboards with ideas upon ideas. But might not actually be good implementers.
  2. Task-oriented implementers. The people who need a recipe and can bring a design from Step 1 to 10, who will also tend to be quiet in these early stages of design.

(Mental waving of arms) That’s me! The second type!

Needless to say I wasn’t so warm to the idea with the eye, I wanted to start building something, anything.

It sounded like there needed to be image processing in the review article (mentioned previously). Plus, people are definitely more reluctant to have doctors looking into their eyes and would much rather give their finger for testing. (My most dreaded medical treatment ever is at an optometrist’s office when they test for glaucoma using air-puff tonometry).

Seema was really excited for the eye idea and kept pushing it. It has huge advantages, a thinner, mucosal layer that doesn’t absorb as much as the thick subcutaneous layer of finger, therefore making it much more accurate.

Jose, Amit, Seema, and [KK] met up to discuss (I was to Skype® in from Wellesley), and it seemed there was a lot of excitement surrounding the diagnosing anemia through the eye. We were given an ophthalmoscope to hack, which is fantastic. So we’ll see where we go from here!

So I need to let go of this finger PulseOx idea, and embrace the palpebral conjunctiva.

Back to top

Wait Until You See the Whites of Their Eyes…

Here’s an idea: let’s diagnose anemia… via the eye!

In this review article, we read about using light reflectance off the inner eyelid to measure hemoglobin. You don’t have to worry about melanin variation, and it’s pretty accurate. Plus, it’s new and different! Finger cuffs that measure Hb? That’s been done before. But an anemia-detecting ophthalmoscope?

Now that’s interesting.

One lesson that’s been tough to learn has been to go step by step. Yes, the ultimate goal is to design a device that will diagnose anemia accurately and efficiently. But that’s not something we can do in three weeks. We may not even be able to measure hemoglobin in the space of three weeks! But we can make the first steps. Even if our prototype can do no more than detect a pulse, that is a huge step towards the ultimate end of detecting hemoglobin and diagnosing anemia.

Anyways, so now all we have to do is learn how optics work, how reflectance works, how an ophthalmoscope works, how blood vessels in the eye work…. and then fabricate a prototype device that puts it all together. In three weeks.

Woo hoo!

Back to top

Optics, Reflectance, and Confusion

So we don’t actually know anything about optics. We knew this before, but now we really know it. What are we actually measuring? Do we want diffuse reflectance, or the other kind? Are we looking for the wavelength, or the intensity? What kind of incident light do we want? How do we filter out the incident light, so it doesn’t drown out our signal?

These are but a few of our questions. We’ve been emailing people left and right, but with very few replies. We finally lined up a meeting with a grad student in the Spectroscopy Lab here at MIT, so hopefully we will find some answers there.

In the meantime, we’ve been toying around with our difficult-to-spell ophthalmoscope. We met with the resident Expert on Everything, Dennis, who explained to us how it actually worked – light comes in the bottom, bounces off a few mirrors, and then light reflected by the patient’s eye comes back into the eye of the doctor. Maybe, we thought, this could be the incident-light-filter we’ve been looking for! After all, the doctor isn’t blinded by the light coming up from the ophthalmoscope.

We tried it with a red LED, and… nothing. We put the eyepiece of the ophthalmoscope against our light sensor, and it literally sensed zero light coming in. Can we improve the sensitivity of the sensor? Would a brighter LED help? Do we need a different sensor? The questions keep piling up. Hopefully we’ll have answered enough of them to have a decent prototype next Friday…!

Back to top

Grad Students are Awesome!

Paul Yongkeun Park from the MIT Spectroscopy Lab kindly met with us last Wednesday to answer all our critical questions that were proving to be huge blocks to our prototyping.

Big Question 1: What wavelength should our light source be and what wavelength should our detector select for? Is the reflected ray going to be at a different wavelength than our light source?

Paul’s answer: These the majority of these wavelengths are going to be the same wavelength of our light source. So really we only need to screen for the wavelength we are emitting at.

Big Question 2: Much research in non-invasive detection of Hb used “diffuse reflectance spectroscopy”, but what does this mean about what we’re detecting? Many groups have used expensive fiber-optic bundles, how can we get around this for a developing world context?

Paul’s Answer: There are two types of scattering: elastic and inelastic. Inelastic scattering is a very small portion of the scattering, therefore we only need to worry about the elastic scattering. So, if we shoot a beam directly at our sample we only need to detect what is coming straight back at the detector (ie. a very small angle to the normal).

Big Question 3: What about calibration? How can we account for the signal coming back from non-blood vessel tissue?

Paul’s Answer: Use the ophthalmoscope in the settings you are to take the measurement (ie. position away from eye, ambient light), and use a mirror instead to get a baseline reading, a certain “I0” (intensity). Then, take the real reading with the eye, intensity “I” and use the ratio “I/I0” to relate to absorbance.

Our comment: The gold standard for diagnosing anemia is drawing blood and getting the exact Hb concentration. So if we know that baseline, we can factor out the signal from the unwanted tissue.

A Problem: How will we make sure that the signal we are getting out is from the blood vessels, the Hb? This tissue will also be reflecting back at the incident wavelength…

Tissue reflectance

Our response: We don’t know really. Let’s focus on getting a signal first, some sort of good reading.

Potential directions for future projects:

  1. Increasing the path length of the beam in the blood vessel (ie. making the angle between the incident ray and the sample surface smaller), that would decrease the signal from the surface tissue. This would involve some sort of fixed geometry that current setup of the ophthalmoscope does not allow.
  2. Making a bigger physical detection window, so that more beams can enter, to get more signal, and therefore have a more accurate reading. We’re currently limited to the small detection window of the ophthalmoscope.
  3. Using two different wavelengths to get a ratio reading, in order to make a non-invasive Pulse Ox! (This would involve two photodiodes with two filters in front.)

Back to top

Roger, we have reflectance signal

Yes! We love prototyping breakthroughs! Here’s a nice timeline:

Big breakthrough 1: We decided not open up and destroy the ophthalmoscope, (even though it is fun to break things and not be able to put them back together), and decided to place the red LED, hooked up to the Arduino, into the place where the normal halogen lamp is in the ophthalmoscope. This makes a lot of sense because since the LED is such a diffuse light source, we were having problems collimating the light and making sure that incident light was not hitting the detector (we want the reflected light off of the sample to hit the detector). For a while we were trying to figure out ways to manipulate the light and put a physical barrier in front so that we could avoid this problem, but we forgot that we had the ophthalmoscope head, that does exactly what we want it to do! Let me explain a bit:

Image by MIT OpenCourseWare.

So basically, the ophthalmoscope head does exactly what we want! Using a beam splitter, the mirrors inside the ophthalmoscope direct the light to the source, but the doctor does not see the light source directly, he sees the reflected rays. Instead of a doctor looking at the sample, we replace him with a photodiode.

Big breakthrough 2: Use a filter to select for a specific wavelength. We already have a photodiode from the Arduino PulseOx project, so we just need it to select for red light. We’ll place a filter flat on the surface of the photodiode so that it can select the wavelength we want. How do we get filters for developing world contexts? Easy. We go to a florist in Wellesley and ask them for sheets of red cellophane. Wrong. Wellesley apparently doesn’t have a flower arranging major. [KK] goes to an art supply store in Boston and they give her a perfect sheet of plastic red PBC. It’s about 0.5 mm thick. It’s perfect. We cut out small rectangles to fit over the photodiode. (We haven’t secured it on yet).

A caveat: Our light source. Two questions:

  1. How did we pick red? We did it randomly and based off of convenience. Since we don’t need a ratio like in the Pulse Ox, any wavelength should suffice to find the concentration of Hb.
  2. What about the type of light source? The original halogen lamp that the ophthalmoscope came with vs. the red LED. It does not matter since both are emitting red light. The white light has components of all light and if we filter out the other wavelengths, the photodiode will essentially be only detecting red light.

Big Breakthrough 3: We got a reflectance reading from the ophthalmoscope! We used the original halogen lamp, put our finger as the sample (“patient’s eye”), and placed the ophthalmoscope in front of the photodiode and got some signal from the graphing program. Whoohoo!

Next new steps: Go to a butcher, get some very bloody meat, and get some sort of membrane or tube to hold the blood, and make measurements. (Of course lots of little details to figure out before then… but step by step…)

Back to top

T Minus 108 Hours…

We have 108 hours left!

To do:

  • make a poster
  • make an elevator-pitch slide
  • construct a testing apparatus to ensure consistency as we collect data
  • go buy meat and use it to test our device
  • build a demonstration for our prototype

At this point, I’m actually pretty confident that we can do it. Er, hold on, let me go knock on some wood. As we get down to it, so much comes down to presentation. You can have a great idea, and you can have a very high-functioning prototype, but if it doesn’t look good, then you’re not going to get attention. I was looking at a few of the posters from the IDEAS competition earlier today, and I was surprised at how terrible some of them were! One of them didn’t even state explicitly what their device did. And even though their idea was amazing, it made their project seem inferior to those with better presentation. It really brought it home for me that It’s All About Selling It. As long as you have to have the idea to back it up, of course!

P.S. Inkscape is an amazing tool for poster-making. And it’s open source!

Back to top

1 2 Next »

A breath-actuated, dosage-monitoring attachment for jet nebulizers to treat multiple patients for respiratory illnesses

Team: Caroline Hane-Weijman, Shan Tie, Mary Jue Xu, and anonymous MIT student [GK]

This content is presented courtesy of the students and used with permission.

1 2 Next »

Ideas Right Now

by Mary Jue Xu

Our blog begins! Many things have happened over the past few weeks, including speaking to many experts at Children’s Hospital in Boston. (We will post noted from that shortly. In summary, there are three large areas of focus for improving the nebulizer: dosage, interface, and sterilization. We decided to focus on dosage.

Initially, we started off with the idea of capnography, in which we would measure CO2 exhaled to correlate to dosage intake. The design would have to incorporate the fact that CO2 does not directly link to the dosage. For example, maybe the nebulizer outlet is far from the mouth or the patient breaths out too fast. Also, a physician at the Children’s Hospital informed us that a baby’s exhale was not enough to be measured with capnography.

So, it was back to the drawing boards. Right now, we are focused upon an attachment to the current compressed air nebulizers that will allow 1) dosage to be monitored and 2) multiple patients to use it. So off the end of the nebulizer motor, image a the medicine being pumped into a reservoir that stores the vapors. Then several outlets would lead to the individual patient’s interface. The interface would only open or release vapors under the direction of the patient. Maybe this is by a mechanical sensor that needs to sit on the face or maybe this is a one way valve that is breath mediated so that only with the baby’s breath will the vapors be released from the reservoir. Moving on past this part of the interface is a sensor (likely an LED sensor) that can detect and sum up the amount (volume/time possibly) that passes by and will sum up the volume or time and signal when the full dosage has been given. Jose also had the idea to use some kind of sticker on the face that is color changing after receiving a particular dose.

Back to top

Problem and Background Information Gathering

by Caroline Hane-Weijman

Just to give some background to our problem and some of the important information we are working with!

Problem Scope

The inspiration for this project came after having visited multiple hospitals, clinics, and health posts in Nicaragua. Currently, the nebulizer is commonly used in developing countries due to the high incidents of respiratory related diseases and infections (the mist from the nebulizer helps unclog liquids found in a sick patient rather than an inhaler). The current nebulizers that are used are air compressors powered by electricity, and the drug most commonly used is albuterol as it is the cheapest. These nebulizers are originally intended for home use but are used for over 30 patients/day in clinics and hospitals, Mothers stand in line with their child until it is their turn to use the nebulizer. These nebulizers do not provide a method for measuring adequate dosage and thus delivery efficacy of the drug to young patients are low, specifically infants. Infants are non-compliant to the face mask that helps administer the drug and the nozzle is just held underneath the infants’ nose without ensuring delivery and wasting a lot of medicine. Additionally, the devices do not have a system to visibly indicate that the device is sterile and ready for use for the next patient, and many are not properly sterilized.

Background

We therefore proceeded to finding more information from experts at Children’s Hospital; including a pulmonologist, respiratory therapist, and a nurse. Below are some important and interesting findings we were able to gather from all three:

Nebulizer Dosage

  • Determined dosage for children
  • Ongoing discussion amongst doctors
  • “Dosage should be same” for a child and adult because more is assumed to be wasted for the child
  • Current dosage is 2.5 mL of premixed albuterol dosage; 0.5 mL parts albuterol and 3 mL of saline
  • Takes around 15 minutes to administer
  • Crying isn’t necessarily bad- children inhale more deeply while crying
  • The way to detect inhaled dosage is through
  • a deposition tracer; using a gamma camera- something that would emit radiation- sensitive photon emitter.
  • Image of the lung to measure how much goes to the lung
  • Ventilation profusion scan
  • Size of dosage particles are very important
  • Too small particles are not helpful and too large will get stuck in upper part of the mouth and not be inhaled
  • Diseases cause under ventilation in areas of your body so medicine cannot be ensured to get into all areas of your body that it may need to- also why size matters a lot
  • Expense of the machine often correlated to how appropriate the size of the particles are
  • Depending on the disease, would ideally like to be able to alter the particle size
  • If sealed mask and if humans breath at a rate faster than the nebulizer administers dosage- can assume 100% is being received

Interface

  • Important to allow inhalation through mouth and nose
  • Especially as many children are sick or crying while they are using it so nose is probably stuffed a large percentage of the time
  • Hard to nebulize through mouth while with pacifier
  • Don’t like to use mask; sometimes put it by their face while sleeping but they are not getting full medication
  • Flavors that would taste could help
  • “Bubble masks” in the shape of fish are currently widely used - figures help compliance. The ventilation holes for this mask is located at the bottom which avoids the eyes being too exposed to the medicine
  • Mouthpiece for inhaler is used for 5-6 year-olds and older; mask is used for children under 5 years of age
  • Play therapy is the most effective way of getting a child to comply (role playing on others and yourself to the child more comfortable with the idea)

Type of Medicines

  • Steroids (more expensive)
  • Albuterol
  • Creates side effects like increased heart rate
  • Physical changes occur but difficult to use as a way to monitor since reactions are specific to patient
  • Considered a safe medicine

Ideas for monitoring airflow as a way to monitor dosage

  • Capnography- summing airflow idea
  • Intubation (exist as a color indicator)
  • Aerochamber- Meter dose inhalers/ spacers- they have whistles- at proper speed
  • Video game of trying to stay ball up in the air
  • Balls stay up- to expand lungs- incentive spirometer

Sterilization

  • Cold chemical sterilization
  • Autoclave- would melt the plastic-some plastics are resistant
  • Anything that touches the patient and anything with a backflow (no retrograde flow) should be sterilized
  • Tube that touches compressor does not need sterilization, only the cup needs sterilization
  • Here, mothers will put parts in a breast milk bag, fill with some water and place it in the microwave
  • For multiple patient use: wash with soap and water then boil
  • Ball that would expand and compress depending on heat to indicate sterilization? Color changing indicator?

Three major types of Nebulizers

  • Jet nebulizer (Air compressor as seen in Nicaragua and most commonly used worldwide)
    • Most commonly used here is the Pariproneb with 50 psi pressure (much higher than Nicaraguan which was around 15 to 20 psi)
  • Vibrating Mesh
    • New technology as of a couple of years age
    • Aeroneb Solo System (Not widely used)
      • Very cool!
      • Disposable piece that would cost $40
      • Would break up into adequate particles sizes of 3-5 microns
      • Piece connected to a controller with batteries and frequency generator (set)
      • Can be placed into series with other machines such as ventilators
      • Medicine is gravity fed through a vibrating mesh that breaks up the particles- mesh is a one-way valve
      • Solves dosage problem: device is breath-actuated so whatever medicine is used is ensured to be in the patient.
      • Problem is that reservoir gets clogged if the medicine isn’t breathed in fast enough Brainstorming ways to stop vibrating mesh if too clogged; can possibly be detected through shining an LED through it. Or if controller can only trigger for next breath?
      • Currently used with corrugated tubing can replace with see-through to look at dosage that is in it.
      • Need sealed area
      • Ideas to prolong life of disposable piece is to create a one-way valve mask that can be attached
    • Ultrasonic
      • Widely used before but it would steam copiously and would also “rip” particles rather than just breaking them up into too small particles
      • Stopped being used in the ’80s and ’90s; no considerable efforts have been put into improving technology so still room for improvement and play.
      • Idea: common reservoir with multiple outlets for multiple patient use??
        • Nurse described E-flow compressor
        • Only used for astroneum medicine
        • Works similarly to vibrating mesh
        • Patients having trouble with cleaning it
        • Fun Fact: Teenage patient used it to inhale marijuana!! (what is the world coming to.. :) )

More details coming soon!

Back to top

Broken Nebulizer in Nicaragua

by Caroline Hane-Weijman

Broken nebulizer

Back to top

Sites and AeroEclipse

by Caroline Hane-Weijman

Some useful sites that were recommended to us by Jose to explore:

Another device Jose helped us come across which almost basically solves our issue… Aeroeclipse!

  • Diaphragm idea that incorporates a breath actuated valve that will direct air from air compressor through medicine when breathing in, and will redirect away while not breathing so only air, no medicine is wasted and therefore dosage delivery can be controlled.

Back to top

Background Discussion on the Capnography Technology

by Shan Tie

The use of capnography was first considered as a way to monitor dosage. Caroline had initially discovered this technology in the context of its usage for anesthesia and intensive care. The capnogram directly measures the inhaled and exhaled concentration or partial pressure of CO2 produced by the patient. Furthermore, the device can indirectly measure the amount of arterial CO2.

There are five major physical methods for detecting CO2; however, the least expensive and most popular method for CO2 detection involves using IR technology.

The actual device interface includes a mask that contains an IR light source, some filters and focusing lenses, and an IR detector. The IR source shines IR light through the cloud of CO2 particles produced by the patient. CO2 selectively absorbs 4.3 micron IR light. The absorption amount is directly correlated to the CO2 concentration and thus the amount detected can be compared to a known standard of CO2.

The information that the capnography provides includes CO2 production, pulmonary perfusion, alveolar ventilation, respiratory patterns, and CO2 elimination. These data is presented as the inspired and/or expired CO2 plotted over time (Kodali, capnography.com).

This technology initially appealed to us because it was a technology that had concepts based very much like the pulse oximeter, is already interfaced with a mask, and is used to measure gas production. Our belief was that if we could place the sensor close between the drug outlet of the mask and the nose of the patient, the patient’s breathing can be monitored by his CO2 production and we can assume due to proximity of the drug outlet to the patient’s nose, that he would be effectively breathing the medicine. We can then use this to track the breathing of the patient over time and count up the amount of time that his CO2 production was above a minimal threshold (to account for adequate breathing) until it totally 10 minutes.

After speaking with Brian Walsh, a respiratory specialist at the Children’s Hospital Boston, we discovered some limitations in using capnography. We intended on using this technology for patients under 5 years of age. Currently, the capnograhy either directly measure the CO2 in the mask or takes measurements from side stream line that pulls out samples from the mask. Unfortunately, the amount of CO2 production by these young patients will be too low for any capnography device made cheaply to detect. Furthermore, the capnography device requires a near perfect seal between the patient and the mask. However, the current masks contain vents to allow the release of the CO2 and a perfect seal for infants is nearly impossible. Given these limitations, we decided we needed a more feasible and reliable method for monitoring dosage.

Back to top

Background on Vibrating Mesh Technology

by Shan Tie

During our visit to speak with Brian Walsh, he introduced a new nebulizer that is based on vibrating mesh instead of the standard air jet from the compressor pump.

The new aerosolizing technology was implemented in a nebulizer made by a company called Aerogen. The structure of the device is a thin mesh material with ~1000 holes. The device is battery powered to cause the mesh to vibrate at 100,000 times a second. This vibrating motion is a sieve-like motion that draws the albuterol liquid sitting above the mesh through the holes like a pump and effectively aersolizing the liquid droplets into consistently sized, aerosolized particles (Aerogen.com). The aersolized drug will then fall down due to gravity and will collect in the chamber of the nebulizer until the patient actively inhales the drug.

This technology seemed promising because it addressed two major pitfalls in the current nebulizer design. One, it created uniform aersolized particles such that it is of a breathable size that will allow for effective delivery of the medicine to the appropriate areas of the patient lung. Second, the breathe-actuated feature of the device allows for more accurate monitoring of dosage because no medicine is actively pushed out of the chamber without the patient breathing it in- thus leading to a lower medicine loss and allows for more accurate measure of how much drug is breathed in.

Back to top

Ultrasound Nebulizer Background

When we visited Brian Walsh and Chris Hug, they also mentioned a type of nebulizer that operated by sending ultrasonic waves through the drug. These were developed around 30 years ago, and consist of a piezoelectric plate which transmits high-frequency vibrations through a liquid. These devices output a far higher volume of nebulized material than traditional jet nebulizers, and since the frequency can be very easily controlled do not have the problem of varying particle size encountered by jet nebulizer. At first we had the idea to utilize the high-volume output by collecting it in a reservoir from which multiple patients could be nebulized. However, we were told that though the particle size was uniform, it could often be too small - meaning that either that the particles would only be effective on deep-lung conditions or that they would not settle in the lungs at all but be immediately exhaled. In some cases, the ultrasonic nebulizers actually ripped up the molecular structure of the drug, rendering it useless.

We decided that considering the far higher cost of ultrasonic nebulizers they may not have much advantage over the jet method - especially since clinics in Nicaragua already have air compressors and compressed air outlets.

Back to top

Design Update: Playing with Nebulizer Reservoir

  1. Drilling holes in Tupperware reservoir

Caroline bought some Tupperware that we could use for vapor reservoirs. Today in D-Lab, we drilled a hole in the bottom to serve as a connection to the medicine chamber. Grace is also going in later today to drill holes in the side of the Tupperware® to place a laser and photo detector.

  1. Condensation in the reservoir

Playing around with the nebulized vapors in the Tupperware®, we noticed a few things. There was condensation buildup on the side, and we wondered if a non-static coating would ameliorate the problem. Also, pressure could build up in the reservoir if closed, so we need to think about whether multiple users would prevent this pressure build up or whether there needs to be some release valve.

  1. Learning about how to measure vapor size

There seems to be multiple ways to measure vapor size. Paul H. talk to us about the one way he does so for the Aerovax project. A laser sits on a stand and shoots a beam of light towards a photosensor. If particles are in the way, then they will diffuse the light and less light will reach the sensor. The larger the particles–> more light scattered–>less light detected by the photosensor–>lower resistance in the detector–> higher voltage outputted and recorded. Paul built and wired this system that we will use.

We want to measure relative vapor size of particles coming out from the nebulizer directly compared to particles that have been sitting in the reservoir to see whether time in a reservoir causes clumping of particles. This is important because Dr. Brian Walsh of the Children’s Hospital mentioned that particle size will dictate efficient delivery of the medicine in to the lungs. So we want to ensure that our reservoir is not changing the particle size.

Today, we were orientated to the setup and we plan on doing some tests during the week.

Back to top

Background on Anti-Static Coating for the Reservoir Surface to Minimize Aerosol Particle Coagulation

by Shan Tie

Anti-Static Methods

Chemical solutions to static electricity: http://www.explainthatstuff.com/howantistaticcoatingswork.html

Glass pipes with a protective antistatic coating: (PDF)

The optimal composition of the applied film is found to be: 50% SnCI~.5H20 with a concentration of 1% SbCI3.

Anti-static coating and its method of preparation:

This patent talks about composition of anti-static material composed of a synthetic resin base without any metallic particles. However, it is in the context of using this coating for the fuselage of airplanes; therefore it might be a bit more than what we need for coating our nebulizers.

%%%

Found various anti-static sprays:

  1. Endust Anti-static spray

  2. Static Guard Anti-Static Spray 5.5 oz (156 g)

  3. Aeros anti static spray (END096000) Category: Surface Cleaners

Found various anti-static tape:

  1. 3M 40PR 1/4 in x 72 yd Clear Anti-Static Tape

  2. Dadas Tapes 1662-05 - Anti-Static Transparent Tape, 3" Core, 1/2" …

  3. Botron B1651 (1/2"x36yd Clear ESD Tape) - 1/2"x36yd Clear ESD Tape

Back to top

Pictures of our Reservoir Prototyping Round 1

by Caroline Hane-Weijman

Vapor laser sensor connected to the Arduino can measure particle size and concentration

We need to explore a valve system so that the pressure does not get too high inside the reservoir.

Next step: Build the patient tube with a one-way valve and another vapor sensor!

Experimental setup with patient tube, one-way valve, and second vapor sensor

Back to top

1 2 Next »

A combination scalpel and extendable electrocautery tip for rural surgical procedures

Team: Sonya Makhni, Victoria Lee, and anonymous MIT students [AK] and [SG]

This content is presented courtesy of the students and used with permission.

1 2 3 Next »

Post 1

Our project will address some of the problems the rehab center in Nicaragua faces with their electrosurgery machine. Right now, the machine runs a huge risk of burning the patients because there is a faulty circuit between the return pad and the machine’s generator. They used to have a means for checking that the circuit is functioning properly, but this no longer works. As a result, several patients experience severe burns during surgical procedures.

Our first idea was to develop a low cost, easy to use electrocautery device that would have rechargeable batteries. The electrocautery machine would be able to function when plugged in to a power source and if it were simply using the batteries. Another idea we had was to develop a censor-based system that would be used with their existing electrosurgery machine itself. This system would contain an auto-shutoff mechanism that would detect if the circuit between the pad and generator was broken. In addition, this system would contain a series of indicators that would alert the user as to which portion of the circuit was not functioning properly (probably via series of on/off lights).

After a couple weeks of discussions, meetings, and research, our team has decided to shift our project’s focus. We first learned that chargeable electrocautery pens could be purchased for $15, a very inexpensive price. This made us rethink our first idea of creating a cheap pen, as they already exist. We also decided to not focus on the electrosurgery idea, for this would involve hacking a system we do not have access to. We are on the road to coming up with a more innovative solution…

Back to top

Pricing for Electrocautery

by [AK]

The cheap ($30-45) electrocautery pens were just the pen unit (no tips or one complementary tip included); typically this can be autoclaved for a certain number of times (40 times was what the representative said). On these, a box of tips (blade or needle type) can be bought for about $50 dollars as Sonya said. The tips must be disposed after each surgery because they cannot be autoclaved (we need to find out why).

The expensive pens (around $80-$125) are actually a box of ten Sterile pens with the pen attached to the tip. Each of these is disposed completely after thesurgery because it cannot be autoclaved. The low temp (which Sonya found out can only coagulate not cut) is a bit cheaper than the high temp (which apparently can cut as well, we need to confirm this).

To do some math:

For 10 surgeries: 1 autoclavable pen + 10 disposable tips = approx $90-$100 (after 40 surgeries a new pen must be bought because it can only be autoclaved 40 times)

For 10 surgeries: 10 nonautoclavable (Sterile) pen+tip units = approx $100

Basically, tips cannot be autoclaved, ever….?

Back to top

Company Contacts for Electrosurgery

Response from Future Health Concepts 
Rob Gorgas 
List of Contacts 
Production Engineering - Medical Equipment Division 
6035 E. 38th Avenue, Denver, Colorado 80207 USA (303) 393-7800 FAX (303) 393-1482

Valleylab - Covidien 
Tel: 1-800-722-8772

Conmed Electrosurgery 
Telephone: 800.448.6506 
5088135544 Erica McLaughlin 
Denver

Back to top

Talk with Ailis…

by Sonya Makhni

Hey Guys,

I just got off the phone with Ailis (pronounced like alyse–whoops, I totally mispronounced her name when I called her). But I got a lot of good info, so here is a recap of the conversation:

  • She really likes our idea (the autoclavable tip + way to modify cheap cautery pens to cut AND coagulate)…She had actually been thinking about something like this before, so she is really excited about this!

Electrocautery vs. Electrosurgery

Both electrosurgery (they call it a bipolar machine) and electrocautery are used. Electrocautery is used more often. They use elctrosurgery for procedures that involve really important structures that they don’t want to damage (i.e. big vessels). It’s used for a more focused area and you can dial up the intensity. Electrocautery is more general and used for a broader number of surgeries. Because its more general, you run the risk of injuring neighboring structures because you essentially have less control than with a bipolar device.

Cutting and Coagulating: Settings

The electrocautery pens have two buttons–blue for coagulate, yellow for cut (or something like that). The surgeon presses the desired button and can immediately perform either function. Sometimes the surgeon uses a foot pedal instead of a button so that he doesn’t have to touch it.

Cut vs. Coagulate

We know that they operate off of different temperatures, but we (before) were wondering why you can’t coagulate with the higher temperature setting typically used for cutting. This is because cutting causes much more damage to neighboring tissue and coagulating uses the minimum temperature needed to get that job done. In other words, the high temperature used for cutting is excessive for the purposes of coagulation and would cause unnecessary damage to neighboring structures/ tissues.

What MGH Does

From what it sounds like, the elctrocautery MGH uses is the kind that runs off of AC current that is attached to a giant machine. She was interested in knowing more about the fancy cautery pens that run off of batteries and can do cutting/coagulation, but I forget exactly where I saw this. Help if you know what I’m talking about on this one please.

They throw away the entire pen. Because it is normal for the tips to get a little bit charred during a surgical procedure (from the vaporized tissue), it seems useful to have replaceable tips. The surgeons have a scratching pad to remove the char, but if it gets to be too much they’ll just replace the tip altogether.

After a surgery they get rid of the whole pen.

The pen they use is from ConMed Corp.

What She Will Do:

She will also get in touch with Yaffe to see if we can visit his OR

She’ll try to get in touch with a medical device representative (the people who come and give demos of their products).

She knows another surgeon that is really into technology for the developing world and will try to help us meet him.

And, she said she could probably get her hands on an electrocautery pen/ tips for us.

What WE Need to Do:

  1. She said she sees no reason why we can’t make these tips out of stainless steel. But, we still need to talk to somebody to make sure there is no reason behind not doing this. This should be our first task to complete.
  2. Find out the mechanism behind how one cautery pen can immediately switch temperature settings. We need to find out how to achieve this same duality on a cheaper, single-purpose coagulation cautery pen.
  3. We also need to let her know of our availability to come in and meet her/ see a surgery. She isn’t free this week, but she will be next week.
  4. Continue to talk to the engineers at these companies to prod them for information.

K, that’s all I think. Let me know if you have any questions, and please respond with comments about your ideas.

Back to top

Focusing on Tips

by [AK]

  1. URLs of products of interest:

Electrocautery pens and pencils (corded - cutting)

Cordless low and high temp

Tips of interest:

  1. Possible Materials:

From customer service rep -

silicone and rubber (?)

stainless steel

tungsten

copper

stainless steel

  1. Company contact - Bovie Medical

Vice President of Regulatory Affairs

I just emailed him. If he does not respond in a couple of days, I will call customer service again and other companies.

Back to top

Questions that Need to be Answered…

by Sonya Makhni

Hey guys, these are the big questions that we need to get answered (we’ve discussed these all before, but I just thought I would post as complete list as I have on our blog)

  1. Are cordless cautery pens that can switch between high and low temperatures used? Can cordless pens cut and coagulate?
  2. Or, is it more common to use the cautery pens that are attached to a generator that runs off of AC current? Why?
  3. What are the tips most commonly made of?
  4. Can they be made of stainless steel? According to one company, stainless steel tips are not used because they aren’t “approved” in hospitals…
  5. How does a cautery pen switch instantaneously between high and low temp with the push of a button (mechanistically)? Is there different wiring for the two settings or does the same system accommodate both temperatures?
  6. We understand that it is common practice to use a cautery tip once…is it possible to autoclave them? Or is there something inherent in the material that does not make this ideal? We cannot figure out if they are disposed of out of convenience, cost (for hospitals), or because the company says they should be…
  7. How exactly are the tips heated? (Does a current pass directly through them, or is heat conducted from the base of the tip?)

Back to top

Questions for Anna…

by Sonya Makhni

Hey Anna, here are a couple questions on the materials side that we said we would post. Please let us know if you have any suggestions/ ideas/ answers! Also, I don’t know if the picture will go through via this email, but I’m also attaching the picture and link of a sample cordless electrocautery pen. This particular one is only high temp, another one is low temp. The company hasn’t gotten back to me about the questions I asked them, but they said they would tomorrow or so.

http://www.medexsupply.com/products/pid-32662/AaronMedicalFineCauteryHighTem.htm?zmam=34602484&zmas=1&zmac=22&zmap=32662

  1. What are some autoclavable materials?
  2. Heat conductivity of stainless steel vs. other materials
  3. Why coat tips with silicon? Does this affect how tips can be autoclaved?

Thanks!

Back to top

Material Science Contact

I will be speaking with Professor Kimerling and Anu Agarwal (Principal Research Scientist in the medical area) about some of our materials. I’m planning on asking

  1. What are some materials that can 

    a) withstand temperatures up to 2000 F 

    b) heat up/cool down quickly (from 1000 F to 2000 F) 

    c) be autoclaved

  2. Why coat tips with silicon?

From the initial conversation I had with Professor Kimerling, he said:

“My guess, at this point, is that silicon is i) stable at the temperatures you mention, ii) has a low thermal expansion coefficient, iii) and grows a surface oxide during sterilization that is hard, protective and relatively chemically inert.”

I have to clarify why the last 2 are desirable qualities, but if there are any other questions you think I should ask, let me know.

Answers from Ailis (Correspondence 2)

by Sonya Makhni

Hey, these are some Q&A’s we were wondering about from Ailis.

  1. MGH does use a grounding pad for their electrocautery machine…they place it on the patient’s thigh.
  2. She’s never seen anyone with a cordless Bovie/ an autoclavable tip…we should still continue to look into this, though.
  3. She can give us the exact model of the instrument when we meet with her

Back to top

Bovie Electrocautery Tip Material

Just got a response from Bovie Medical, the manufacturers of the Change-A-Tip electrocautery pens.

I asked about the tip material:

“For Bovie’s cautery replacement tips, the body material is ABS with brass post.”

ABS is a plastic (Acrylonitrile Butadiene Styrene) that is a common thermoplastic. It is good for impact resistance, it is tough, light, and a good electrical insulator"

I also asked about the two different temperature settings:

“When referring to the H100 (low temperature) and H101 (high temperature) there are no design differences. This was placed into Bovie’s catalog as quick references for the end user using low/high temp cauteries.”

Back to top

1 2 3 Next »

Course Info

Instructor
Departments
As Taught In
Spring 2010
Level
Undergraduate
Topics
Learning Resource Types
Design Assignments with Examples
Download Course
MIT Open Learning
Over 2,500 courses & materials
Freely sharing knowledge with learners and educators around the world. Learn more
© 2001–2024 Massachusetts Institute of Technology
Accessibility
Creative Commons License
Terms and Conditions
Proud member of:
© 2001–2024 Massachusetts Institute of Technology