What problem does this project tackle?
Jaundice occurs due to excessive bilirubin content in the body, leading to yellow or green pigmentation in the skin or sclera. While mild jaundice often resolves naturally, severe cases can lead to kernicterus, a condition that may cause long-term neurological damage, including cerebral palsy, hearing loss, seizures, or intellectual disabilities. The current gold standard for diagnosis, bilirubin plasma tests, and treatment, phototherapy, are expensive and often inaccessible in low-income and rural communities, making early detection and treatment difficult.
Who experiences this problem in the world?
Jaundice is highly prevalent among newborns worldwide. Approximately 50% of full-term and 80% of preterm newborns develop clinical jaundice within the first week after birth, with symptoms appearing within three days. Globally, more than 85% of newborns experience jaundice. In severe cases, untreated jaundice is a leading cause of neonatal death, affecting 5-10% of newborns. The burden is especially high in low-resource settings where access to diagnostic tools and treatment is limited.
Why is this problem important?
Jaundice is a major neonatal health concern due to its high prevalence and potential for severe complications. While many cases resolve naturally, undiagnosed or untreated jaundice can lead to kernicterus, causing permanent neurological damage or death. The lack of affordable, accessible diagnostic methods disproportionately affects newborns in low-income regions, where advanced medical care is scarce. Developing cost-effective, widely available screening solutions could significantly improve early detection and reduce the risk of severe outcomes.
What specific technical problems did you encounter:
One of the key technical challenges in developing this web-based jaundice detection application is ensuring accurate classification across various lighting conditions and image qualities. Additionally, ensuring data security and compliance with medical privacy regulations is also critical, as the application deals with sensitive health information.
Describe the coolest thing about your project:
The coolest thing about this project is its potential to provide a fast, accessible, and non-invasive jaundice screening tool. This approach makes early jaundice detection more accessible, especially in low-resource settings where traditional diagnostic tools may be unavailable. With just a smartphone or computer, healthcare providers and even parents could use the application to assess risk levels and seek timely medical intervention.
Describe any sustainable design considerations driving your solution:
By using machine learning, the system reduces reliance on physical medical infrastructure, minimizing the need for disposable testing materials and lowering overall healthcare costs. Additionally, the use of widely available devices like smartphones and computers ensures accessibility without requiring specialized hardware, making the solution both scalable and environmentally friendly.
Project Name: Jaundice Tracker
Team Members: Mashal Ahmed, Kailey Christensen, Alexandra Dakoyannis, Marissa Huddy
Team Mentors: Oluwadamilola Oke
Team Sponsors: N/a