Childhood leukemia survival rates show a stark contrast between countries in the Global North and Global South: 80% versus less than 30%, respectively, as reported by the World Health Organization. The primary challenge in lower-income regions is the delay in diagnosis and the absence of timely treatment.
Founded by Claudia Durkin, Vincent Gu, and Sam Won, students in the Master of Information and Data Science program at the University of California, Berkeley, A World for Every Child is a capstone project focused on improving childhood cancer diagnosis rates for underserved children.
The project was supervised by Korin Reid, PhD and Puya Vahabi, PhD.
A World for Every Child capstone project aims to work collaboratively with communities in under-resourced regions to improve childhood cancer diagnosis rates using innovative machine learning techniques.
Our mission is to create a diagnostic tool that facilitates early and accurate diagnoses, reducing the diagnostic burden and increasing scale, ultimately broadening accessibility, thus offering every child, irrespective of their geographical location, a fair chance against cancer.
Using machine learning techniques, we have developed two complementary models to diagnose childhood Acute Lymphoblastic Leukemia (ALL) from peripheral blood smears (PBS), enhancing early diagnosis, offering an alternative to conventional manual methods. This tool does not replace medical advice or care, but rather, aims to enhance diagnostic capabilities using machine learning models vis-à-vis expert manual analysis.
For more information, visit our project page.