Program Type

Undergraduate

Faculty Advisor

Dr. Robin Ghosh

Document Type

Presentation

Location

Face-to-face

Start Date

18-4-2024 10:20 AM

End Date

18-4-2024 10:50 AM

Abstract

While cancer impacts all segments of the United States population, specific groups experience a disproportionate burden of the disease due to social, environmental, and economic disadvantages. This research examines the correlation between socioeconomic factors and the accessibility of cancer clinical trials across U.S. counties, employing a comprehensive dataset, County-Level Socioeconomic and Cancer Clinical Trial Data from Noah Ripper, and advanced machine-learning techniques. Our findings, derived from regression analysis and machine learning models like gradient boosting, highlight significant disparities in trial availability linked to socioeconomic indicators, including poverty rates, population estimates, median income, incidence rates, and mortality rates. Many regression models such as gradient boosting, random forest, linear regression, and K-neighbors, were implemented to find the best fit for the data. The models, particularly gradient boosting, showed about 75% prediction accuracy. The other three showed initial accuracy of 72.79%, 67.16%, and 62.25% respectively. These insights provide a foundation for targeted interventions and resource allocation to improve healthcare equity and cancer treatment outcomes. In the future, we plan to refine the best-performing model to enhance predictive accuracy to at least 90%, aiming for more precise and actionable insights to guide interventions and improve equitable access to cancer care.

Comments

https://kray19.github.io/Map-Website/

Share

COinS
 
Apr 18th, 10:20 AM Apr 18th, 10:50 AM

Analyzing the Impact of Socioeconomic Factors on Cancer Clinical Trials Accessibility in the U.S. Using Machine Learning

Face-to-face

While cancer impacts all segments of the United States population, specific groups experience a disproportionate burden of the disease due to social, environmental, and economic disadvantages. This research examines the correlation between socioeconomic factors and the accessibility of cancer clinical trials across U.S. counties, employing a comprehensive dataset, County-Level Socioeconomic and Cancer Clinical Trial Data from Noah Ripper, and advanced machine-learning techniques. Our findings, derived from regression analysis and machine learning models like gradient boosting, highlight significant disparities in trial availability linked to socioeconomic indicators, including poverty rates, population estimates, median income, incidence rates, and mortality rates. Many regression models such as gradient boosting, random forest, linear regression, and K-neighbors, were implemented to find the best fit for the data. The models, particularly gradient boosting, showed about 75% prediction accuracy. The other three showed initial accuracy of 72.79%, 67.16%, and 62.25% respectively. These insights provide a foundation for targeted interventions and resource allocation to improve healthcare equity and cancer treatment outcomes. In the future, we plan to refine the best-performing model to enhance predictive accuracy to at least 90%, aiming for more precise and actionable insights to guide interventions and improve equitable access to cancer care.