Skip to Main Content

An official website of the United States government

Principal Investigator
Name
Stacy Amadi
Degrees
M.S.
Institution
Northwestern University School of Professional Studies
Position Title
Student
Email
About this CDAS Project
Study
PLCO (Learn more about this study)
Project ID
PLCO-1696
Initial CDAS Request Approval
Oct 11, 2024
Title
Fairness of AI use in Healthcare
Summary
I am researching using artificial intelligence (AI) in healthcare for my data science capstone course at Northwestern University School of Professional Studies because it presents vast opportunities and challenges. The opportunities range from better care and treatment to drug research and development. However, challenges also exist in data collection and how it’s used in research and treatment of diseases. These represent the risks and ethical considerations of using AI in healthcare, such as data and patient protection, transparency and explainability, and fairness. For this project, I am researching colorectal and ovarian cancer.

Colorectal and Ovarian cancer has been dubbed “silent killers” because the early-stage symptoms are either unnoticeable or mistaken for typical ailments. Although the total ovarian cancer diagnosis rate continues to fall, the 2024 American Cancer Society projected numbers show that an estimated 65% of the diagnosed would succumb to the disease (American Cancer Society 2024). While the project number for colorectal new cases is nearly eight times that of ovarian cancer, an estimated 35% will likely die from the disease.

This study will concentrate on the fairness of data used for research and treatment and how this information can be improved. The estimated death rate from ovarian cancer is high relative to colorectal cancer, while the estimated new cases of colorectal cancer are significantly higher than ovarian cancer. More robust AI methods and applications in this sector can substantially improve these numbers in the future.



References
American Cancer Society. 2024. “Key Statistics for Ovarian Cancer.” Last updated January 19, 2024.
https://www.cancer.org/cancer/types/ovarian-cancer/about/key-statistics.html.
Aims

• Find relationships and commonalities in risk factors based on geographic locations, demographics/ethnicity, and age. Since two/three demographics represent those significantly affected by these diseases, they should be well-defined.
• Study each attribute’s (geographic location, demographics/ethnicity, and age) role in the research and treatment data.
• Investigate the impact of the research on treatment results by geography and demographics
• Research more robust AI methods and applications in this sector can substantially improve these numbers in the future.

Collaborators

None