MilkGuard: Predictive Modeling and Mobile App Development for Affordable, Usable Breast Milk Diagnostic
Date of Award
Santa Clara : Santa Clara University, 2021.
Breast milk is considered the gold standard of infant nutrition, but some infants around the world lack access to it due to maternal health complications or other considerations. Human breast milk banks do exist to try to alleviate this problem, but most are underfunded and have high operational costs, making it difficult for some infants to obtain safe, reliable donated breast milk.
Existing methods of testing breast milk are expensive, so the MilkGuard project was conceptualized in 2017 as a fast, economical, and highly usable bacterial contamination detection system. Prior to this year, previous MilkGuard teams had developed a system that was faster and more affordable than prior methods, but its main drawbacks were that it was difficult to use and that it lacked the sensitivity to detect low Escherichia coli (E. coli) contamination levels. To ameliorate these drawbacks, our goals for this year were 1) to improve MilkGuard’s sensitivity to the Human Milk Banking Association of North America’s (HMBANA) lower limit of detection standard of 104 CFU/mL, 2) to increase the ease of the assay process, and 3) to achieve these objectives in an economical and environmentally-friendly way.
Through COMSOL Multiphysics software simulations, we proved the possibility of realistically optimizing biosensor parameters on a computer. Since the simulations were virtual, we discovered an optimal biosensor configuration without the need to purchase, manufacture, and test hundreds of physical sensors. Future teams can quickly confirm these results by building a physical sensor in the lab. We also developed the MilkGuard app, which greatly simplifies the colorimetric analysis process for the user. This mobile app uses our improved color-analysis algorithm which improves detection sensitivity around the HMBANA’s lowered limit of detection standard, given the same image data to analyze. The efficacy of our new color analysis algorithm can be confirmed by future teams in the lab, and our current regression curve can be made more robust with a larger sample size.
Taken together, our developments this year have increased the usability and sensitivity of the MilkGuard system, which can improve bacterial contamination testing by milk banks and move one step closer to equitable access to safe breast milk for infants around the world.
Hsia, Beau and McCurry, Emma, "MilkGuard: Predictive Modeling and Mobile App Development for Affordable, Usable Breast Milk Diagnostic" (2021). Bioengineering Senior Theses. 109.