EAAMO ’24 Social Hackathon
One of the most impactful initiatives at EAAMO ‘24 was the Social Hackathon: Breast Cancer in Mexico, aimed at addressing the country’s growing challenges in breast cancer detection, treatment, and care. We detail the description and outcomes of the hackathon below.
A total of 31 individuals expressed interest in participating in the social hackathon. After a month of work involving collaboration and problem-solving sessions, four teams reached the final stage. The winning team, Team Sana, composed of Harvard students, presented a groundbreaking solution that impressed both judges and policymakers. The team will continue collaborating with key decision-makers to refine and implement their innovative approach, contributing to meaningful advancements in breast cancer care in Mexico.
The winning team Sana, submitted by a team of Harvard University students Sarah Shirley, Alina Yu, Fiona Millan, Jasmine Andresol, Audrey Chang, Diana Yue, and Richael Saka, proposed a solution that is a public health tool designed to assess personalized breast cancer risk while raising awareness of environmental factors contributing to the disease. Focused on the region of San Luis Potosí, the tool combines a mobile app and a web dashboard to empower individuals and policymakers. For residents, it provides real-time local data on air and water quality, UV exposure, and personalized risk assessments based on demographic and lifestyle inputs. For policymakers, Sana visualizes environmental breast cancer risks using machine learning models that predict cancer prevalence across geographic regions. This centralized solution addresses gaps in breast cancer awareness and prevention, especially in Mexico’s resource-limited settings. By integrating data, educational content, and potential government collaboration, Sana aims to reduce disparities, improve early detection, and inform effective policies for breast cancer prevention and care.
The runners-up were also Harvard University students Khushi Kohli, Coby Garcia, Cody Chou, and Joaquin Alvarez with their proposed solution - solution ALMA. ALMA (sp. Apoyo Local para Mujeres con Análisis) is a data-driven platform aimed at improving early diagnosis, access to screening, and treatment of breast cancer in underserved areas of Mexico. By leveraging environmental, demographic, and healthcare data, ALMA identifies key risk factors, such as air carcinogens, age, and geographic disparities. Using a machine learning model with 97.7% accuracy, the platform categorizes women into low-, medium-, and high-risk groups based on environmental exposure, age, and regional factors. The solution highlights urban-rural disparities, with urban areas experiencing higher mortality rates, and identifies emission hotspots like Tabasco and Nuevo León. Designed to inform public health strategies, ALMA provides actionable insights for policymakers to target interventions, regulate environmental pollutants, and improve healthcare access. Future steps include developing a user-friendly mobile app for individuals to assess personal risk, visualize data interactively, and receive guidance on preventive measures.
Team ALMA presenting their solution
A team of Mexican students Héctor Segura Quintanilla, Ericka Ovando-Becerril, and Daniela Sánchez-Batallas won third place in the hackathon, with a solution Parallelized interventional and surveillance system for breast cancer in Mexico. This solution aims to address breast cancer challenges in Mexico through an integrated data-driven and AI-powered approach. The solution focuses on creating a centralized database of mammograms labeled with expert diagnoses to enhance early detection efforts. A deep learning model, specifically a ResNet-based Convolutional Neural Network, was proposed to assist medical professionals in classifying mammograms efficiently, reducing delays, and increasing diagnostic capacity. The solution also introduces a Breast Cancer Risk Index, which combines environmental pollutants, healthcare accessibility, and social marginalization data to identify high-risk municipalities. By leveraging common machine learning models, the system pinpoints geographic and socio-environmental disparities, providing actionable insights for targeted public health interventions. To enhance accessibility, particularly in underserved and indigenous communities, the project highlights the need for multilingual communication systems using large language models to overcome cultural and linguistic barriers. Future goals include nationwide expansion of the risk index, retraining diagnostic AI with mammogram datasets, and developing predictive models for breast cancer incidence. This multidimensional solution integrates AI diagnostics, data infrastructure, and cultural inclusivity to optimize early detection, improve access to care, and inform policy for effective breast cancer management in Mexico.
Parallelized interventional and surveillance system for breast cancer in Mexico presentation
An honorable mention went to Shobhit Jagga (Quantitative Researcher, Quadeye) for his submission Contribution based Budget Allocation for Breast Cancer Mortality Reduction. Shobhit tackles breast cancer mortality in Mexico using a predictive AI framework that analyzes environmental, healthcare, and socioeconomic factors. Focusing on excess deaths instead of Standardized Mortality Rates, his model identifies key contributors such as carcinogenic pollutants, marginalization, and healthcare access. A Random Forest model predicts mortality risks, while a budget allocation system optimizes resources based on cost-effectiveness using a combinatorial approach. The solution features a functional web prototype with an interactive risk visualization map, a “what-if” simulation tool for data updates, and a budget allocation tool to prioritize interventions. Designed with ethical safeguards and explainable insights, the system empowers policymakers to make evidence-based decisions, addressing regional disparities and improving resource allocation to reduce breast cancer mortality.
We believe that social hackathon exemplifies the power of collective action and creative thinking from three areas unique to EAAMO – student enthusiasm, EAAMO’s willingness to bridge research and practice, and policy-makers’ domain knowledge and expertise.