1) We met with Paul Hlebowitsh and found that using the ophthalmoscope + LED did produce enough signal for the photodiode. So we need to ditch that idea and come up with something else.
So the big problem was: how do we make sure we're only getting reflectance signal?
Seema suggested: Why do we need to only get the reflectance signal? We could subtract out the baseline signal from the LED and measure the change in signal…
So that's where we proceeded… and made this makeshift prototype:
So as you can see, the photodiode is right next to the LED, picking up the LED's light, but also the reflected light off of the "blood" (red paint + water) in the cuvette. (The Helping Hands were holding the cuvette securely and the Arduino board taped in place).
Seema talked with Andrew Berger from the spectroscopy lab at MIT, and he gave us a few very valuable tips:
In our final prototype, we may want to use lenses to collimate the light from the LED. This would be the most expensive part of the final prototype, since these lenses would cost $20-30.
Prototype 1: Superbright red led + cuvettes of watered-down paint
We get our first exciting measurements: the red cuvette reflects much, much more than the blue! The difference in reflectance of red and blue is analogous to the difference between hemoglobin and tissue, so this was great news.
The only problem was with consistency. The difference between red and blue varied greatly… sometimes, blue inexplicably reflected more! The sensor often refused to settle on a single value, and would keep increasing or decreasing.
So after talking with a Professor Andrew Berger, we tried a different design!
Prototype 2: Red laser + pools of dyed milk
This prototype was… painful. It was late, we were tired… and we had tape. Lots and lots of tape. Which we used. Everywhere.
The laser (suggested by Prof. Berger) gave us a little more consistency – if we shined the laser directly at the sensor, we got the same values every time. Plus, we could tell exactly where the light was going. But, it was much, much harder to keep the laser still. At first Seema just held it, but we couldn't get consistent values, so we tried aluminum foil, cardboard, ruler and clamps (shown on the left)… none of them got us consistent data.
So… we took it to D-Shop.
Prototype 3: Constructed setup + red laser + pool of dyed milk
Thanks to D-Shop and the Wellesley chemistry labs, we have a secure set up! You can see our three samples in the back: red ("blood"), pink ("anemic blood"), and white ("tissue"). The white circles sitting in the milk represent the "eyeball". There is a hole in the middle of the setup, in which a sample tightly fits.
This seemed very promising! But, long story short, it wasn't. We got a lot of good data, but sometimes white still reflected way more than red, which was very unsettling.
We're hoping some of this inconsistency is due to the different volumes of milk that were used… if not, I have no idea what's going on!! In the meantime, for our poster, we'll have to use the data we have – there's no time to do more tests until tomorrow.
Yesterday we presented our eyeHeme, which is the "hip" name I came up with and I guess other people liked as well. Amit suggested using a hip name instead of a more standard name.. so there we are!
It was a great time to see the devices that all the other D-lab classes came up with. Really interesting stuff…
Seema gave an awesome elevator pitch, and in general people seemed really interested in our device. Our device is different from other teams in that it's solving worldwide, huge, very apparent need. The other projects have a lot of potential to solve problems of poverty. They identified a problem they observed in the field, and are solving it, but it seems that it will only affect a select group of people. So, this is probably the eyeHeme's strength, that we have potential to affect 2 billion people, 42% of pregnant women around the world, 40-50% of children in developing countries…
We also had someone run up to us as we were cleaning up named Sarah Bird, who is the Chief Technology Officer at IRD (Interactive Research and Development), a medical devices organization in Pakistan. She was really excited about our project and offered to help us set up some clinical trials when we had a patient-ready device built. Very nice!
Hey everyone – I'm new to the blogging thing, but I'll do my best to keep up with my awesome team members!
I wanted to write a bit about my relationship with the eye. As [AJ] mentioned in one of her earlier posts, I was really into this idea of obtaining a signal from the eye instead of other tissue. Why?
Evidence for diagnostic potential:
The eyes are affected by several conditions stemming from other parts of the body (from The Merck Manual's "Diseases of the Eye"). This all suggests that the eye may already be a familiar target for informal/qualitative diagnoses (so our diagnostic may be easily integrated into current practice), and that tools may already exist to better visualize parts of the eye.
Extraneous Personal background:
Please note that I do not prescribe to alternative medicine techniques like "eyelogy" or "iridology" or "scelerology." My impression is that although the eye can be a very useful tool for some diagnostic purposes, these fields are generally accepted as bogus.
Over the weekend at the D-Lab showcase event, I represented our team by giving an elevator pitch on the eyeHeme. Each team (from every D-Lab class) gave a quick presentation on their device, inviting the audience to come visit their poster to learn more. I wrote up something while thinking about what to say, although I ended up just informally mentioning some main points. Below is the formal, written version:
Anemia is a condition characterized by red blood cells that do not supply an adequate amount of oxygen to tissues in the body. This affects an estimated 2 billion people worldwide, and is associated with decreased productivity, increased risk of infection, and rapid disease progression. The WHO describes anemia as a leading contributor to the global burden of disease, and the World Bank estimates that, worldwide, $50 billion in GDP is lost each year due to anemia.
Pregnant women and young children are at especially high risk, and maternal anemia is associated with maternal and infant morbidity and mortality.
Effective treatment for most anemia cases is well-understood, but especially in developing world contexts, anemia remains underdiagnosed and undertreated. Current diagnostic methods – generally using analysis of a blood sample – are invasive and costly, and often require additional skilled labor and infrastructure for sample analysis, infection control, and waste management. Non-invasive techniques –mostly based on observation – are criticized as qualitative and inaccurate, and are only sensitive enough to detect severe cases. A significant bottleneck in addressing the worldwide challenge of anemia is the lack of an appropriate and effective diagnostic.
So we ask you now to envision the eyeHeme, our novel non-invasive system used to diagnose anemia. Our design uses reflectance from blood vasculature in the inner eyelid to measure the concentration of hemoglobin, the iron-containing compound of RBCs that carries oxygen. This region is a highly vascular area, and an excellent target for an optical detection device. Specifically, it is not affected by scattering in thick tissue or variation in melanin – two challenges that other attempted optics-based diagnostic devices have faced. We believe that the reflectance signal we detect is both sensitive and specific enough to serve as an initial screen for anemia. We welcome your comments and feedback – please come see our table upstairs. Thank you.
Anemia is a condition defined by the inability of red blood cells (RBCs) to carry enough oxygen to the tissues of the body.
This can involve:
Obvious changes in the palpebral conjunctiva → useful for diagnostic purposes?
(Poster and presentation files removed due to copyright restrictions.)
We also got invited to present at the The World Health Medical Technology Conference at BU this Monday! There are some really exciting presentations during the day. I'm bummed I can't go…
Mary Xu, in our D-Lab class sent us a link about Biosense, a non-invasive anemia screening that is an Echoing Green Fellow receiving up to $90,000 in seed funding and technical support.
[KK] had gotten in contact with them as well and they were willing to meet up with us in Seattle or New York, but I guess we didn't get the chance.
Take a look at Biosense, with their product ToucHb. ToucHb uses a modified pulse-oximeter using the finger as a measuring point for hemoglobin. It looks like they had gone to clinical trials this past January. It would be interesting to see the results of their clinical trials and how accurate their readings were considering the same problems we faced about subcutaneous tissue being different in each person, thus affecting the Hb readings.
But congrats to Biosense for drawing more attention to the worldwide anemia problem!