The S.A.M.D.S.T.

Inadequate access to prenatal Down Syndrome testing is a global issue, largely affecting regions without proper laboratories to run these screens, specifically rural Peru. The lack of proximity to working laboratories combined with the high expense associated with running Down Syndrome tests makes prenatal Down Syndrome screening completely unviable for over 7 million rural Peruvians.

Down Syndrome is the most common chromosomal disorder and the leading cause of mental and physical disability, with a frequency rate of approximately 1 in 700 babies. Despite this, Down Syndrome is the least funded major genetic condition by the National Institute of Health (approximately $20 million dollars per year out of a $50 billion dollar annual budget) and receives little to no funding in the developing world.

Algorithmic photo-analysis of lateral flow assays (LFAs) will determine the chance a fetus has Down Syndrome by calculating free-beta human chorionic gonadotropin (hCG) and pregnancy-associated plasma protein A (PAPP-A) levels in maternal blood via pixel analysis of assay test lines.

PAPP-A and hCG are both biochemical markers that indicate a woman’s state of pregnancy, as well as the developing fetus’s risk of having Down Syndrome, i.e. trisomies 16, 18, or 21. Low concentrations of PAPP-A and high concentrations of hCG in pregnant mothers (during the first trimester) are both correlated with the fetal development of Down Syndrome.

Currently, hCG and PAPP-A concentrations are measured in laboratories through blood sampling; while accurate, this method is not plausible where laboratories aren’t accessible. Our solution is to determine hCG and PAPP-A concentrations via lateral flow assays, which are cheap, portable, and easy to use.

Lateral flow assays (pictured above) are self-administered devices that require mere drops of human fluids - in our case, blood - to detect the presence of the target analytes (hCG and PAPP-A) in a matter of minutes. They function by having a test line that darkens in the presence of these analytes.

The darkness of these test lines exhibits a strict mathematical relationship to the concentration of the given analyte, where the test line’s darkness (as measured through pixels of an image) varies logarithmically with the concentration of the analyte. Our solution will exploit this equation to create a machine learning algorithm that will analyze user-submitted images of their completed LFA’s test lines in order to derive the concentrations of hCG and PAPP-A present in the user’s blood. The algorithm will then use these calculated hCG and PAPP-A concentrations to determine the likelihood that the user’s fetus has Down Syndrome, as can be done to great accuracy.

Our target population is the 100 thousand “soon to be parents” in rural Peru who have yet to be informed of their unborn child’s health conditions - possibly Down Syndrome - but do not have the access to or money for laboratory blood testing. The S.A.M.D.S.T. will provide a cheap, accessible, and accurate alternative to laboratory testing that does not require the proctoring by a clinician. It will be self-administered and exhibit high ease-of-use, aiming to quell all worries a future parent might have regarding finances, travel, or user error.

Tancrede Roy of Boston, Massachusetts is the founder of the Tancrede Swim Challenge for Down Syndrome which began six years ago. To date, the challenge has raised more than $50,000 towards pioneering water therapy as a treatment for Down Syndrome; their focus has now moved onto the investigation of biological markers that may help determine if a fetus has Down Syndrome via the novel use of lateral flow assays.

John Paul Anderson, a physics major at University of Massachusetts Amherst, is currently in direct contact with neonatologists in Peru and is continually gathering testimonies that affirm the lack of healthcare infrastructure in rural Peru. His dedication to the betterment of living standards is evidenced by his own non-profit group, Amautas, which has already worked to close the technological gap for rural schools in Peru. He and his brother, Kyle, both Peruvians, have extensive connection to Peru; this has brought much insight to the struggles that face the country. Through “Spark”, a three week program by the NSF and administered by MIT, we have conducted 12 interviews with mothers and doctors in rural Peru. Our team will also partake in the follow up program “Fusion”, where 12 additional interviews will be conducted.

Recent additions to our team include Kyle Anderson, a biochemistry major at UCLA, and Hai Nguyen, a computer science major at the University of Massachusetts Amherst, who have been tasked with working on the LFA design and machine learning algorithm respectively.

Furthermore, our team puts substantial effort towards understanding the ethical dilemma behind Down Syndrome treatment. We respect any decision that the parent of a baby with Down Syndrome may make and only seek to provide the parent with as much information as they desire - nothing further. We have ensured that our solution remains considerate to the perspectives of the Down Syndrome communities around the world.

Solve would provide our team with the resources necessary to ensure that our solution remains technically sound. Technical expertise in the biomedical industry would greatly assist us in the manufacturing and production of our novel lateral flow assays. Furthermore, connections with senior programmers in machine learning would provide guidance as we navigate the creation of our unique algorithm.

Funding from MIT Solve and investors would support our travels to rural Peru, during which we will perform the field work essential to maintaining cultural connection with our target group. Beyond conducting interviews, we will move forth with sampling human data to bolster our algorithm, specifically in the communities we are serving. We have already received doctoral consent to a testing body comprising rural Peruvians, who remain largely underrepresented in medical studies. Solve would facilitate and expedite this process.

Through our literature review and numerous interviews, we can confirm that there is no current prenatal test for Down Syndrome in Peru other than a laboratory test in Lima. Our device would be the only test accessible to and affordable for the aforementioned 7 million rural residents. We have also secured patient testing from neonatologists in the regional hospital of Cusco, where most of these patients would be medically understudied rural citizens.

1. Establish a statistical baseline for comparison of hCG and PAPP-A concentrations with their testline darkness

2. Create a prototype LFA that reads for both hCG and PAPP-A without cross-interference at a minimum accuracy of 80% (as is the standard for hCG and PAPP-A biotests)

3. Develop and run approximately 200 tests on the app/software for our prototype, optimizing the algorithm intermittently

4. Secure a deal with manufacturers to produce our prototype on the scale of hundreds

5. Make, publish, and share our app in Cusco

6. Distribute devices to our neonatologist contacts in Cusco

7. Get direct feedback from the rural recipients and revise as needed

Our solution is aligned with the UN Sustainable development goal in the listed manners:

• Increase reproductive, maternal and child healthcare access

• Lessen healthcare inequality and augment equitable access (our LFA is self-administered, inexpensive, and widely accessible)

• Ease the burden on healthcare workers (our device does not require technicians or a blood lab because it is self-administered)

• Increase the medical surveillance of a demographic who would otherwise remain ignored due to monetary expenses, lack of trained technicians, and lack of blood testing facilities in the region

With the end goal of total access to prenatal Down Syndrome testing, we believe the steps towards that are:

• The device functions from a technical standpoint; it can accurately read hCG and PAPP-A to great sensitivity

• The app functions from a software standpoint; it can analyze all reasonable test photos and convert their data to an associated risk.

• Both the device and app function from a user standpoint; users can operate both with ease and minimal error. 

• An overall accuracy of at least 80% for the algorithm-LFA test; this quota is in accordance with standard blood testing for Down Syndrome.

• Widespread access to the device/app; remove all financial, proximal, and cultural barriers.

Our team has conducted interviews with rural Peruvian doctors to assess the potential impact our device would have. They stated that the S.A.M.D.S.T. would eventually relieve much burden off of technicians in their regional hospital, as they are currently understaffed. They also believe our device would allow their doctors to record previously unavailable data on their rural patients. In addition to interviewing doctors, our team has communicated directly with Peruvian mothers regarding their opinion towards prenatal Down Syndrome testing. They were all open to the use of our device, as they all desired to know the most about the health of their growing fetus. 

The larger picture is Peru’s relationship with Down Syndrome. Our team has interviewed dozens of Peruvian teachers, and their consensus is that the lack of funding is disproportionately affecting students with learning disabilities. Parents who know - with months of advance - if their child has Down Syndrome can more effectively prepare for their upbringing and education as necessary.

Our solution relies on the novel repurposement of a commonplace technology, the lateral flow assay, via introducing the measurement of both hCG & PAPP-A concentrations. A single drop of blood would be put in a two-track lateral flow test to reveal hCG and PAPP-A presence or lack thereof. Our LFA would be the first to measure the two of these analytes.

Furthermore, we are developing a new app in house that reads and analyzes the data from a user-submitted image of the LFA test lines to provide the patient with an accurate likelihood for the chromosomal disorder in their fetus. As of now, ideal photo conditions and a proper camera would facilitate the algorithm’s function greatly, but a more advanced AI is certainly required for use in the developing world, where camera quality and conditions may be less than ideal. Our lead programmer Hai is currently studying machine learning to better create this algorithm, and we have contacted experts in the specific field of LFA photo-analysis for guidance.

Beyond the US, our small team of four has members representing Peru and Vietnam, with extensive family in each. As mentioned before, we partook in the “Spark” program by I-Corps, NSF, and MIT, where we conducted 12 interviews with our target demographic. We will continue this model of diverse outreach with our participation in the following “Fusion” program, in which we may receive a micro-grant for travel to rural Peru for more interviews of these historically underrepresented demographics. This direct contact allows us to hear the problem from the people’s standpoint and to get their honest feedback on what solution would aid their underserved community best.

As part of the Spark program, we have specialized our model further towards our target demographic. Our value for the expecting parent is the removal of both travel and high expenses. For health care workers, our device will provide them with more data on their patients, as well as ease some of the burden off technicians in laboratories testing for Down Syndrome. These testimonies have been given by rural Peruvian citizens in addition to doctors from a regional hospital in Cusco.

The Tancrede Swim Challenge for Down Syndrome has raised more than $50,000 over the past six years. Our plan is to offer our device at an exceptionally low price and to purpose the revenue towards funding further research. We are applying to the NSF’s Small Business Innovation Research (SBIR)/Small Business Technology Transfer (STTR) program, which grants startups up to $2,000,000 in funding. Future grants and fundraising will continue to provide the financial stability for our research to prosper.

We have been invited to submit a full submission to the NSF’s SBIR program, after an initial smaller pre-proposal which they accepted. As mentioned previously, we have raised over $50,000 in funding through donations as part of the Tancrede Swim Challenge.