Ammar Khan
Pronouns: He/Him
Research Mentor(s): Jonathan Reader
Research Mentor School/College/Department: Michigan Alzheimer’s Disease Research Center / Medicine
Program: UROP
Session: Session 5 (2:40pm – 3:30pm)
Authors: Hasan Khan, Sam Chuisano, Melissa DeJonckheere
Abstract
Recruiting participants into Alzheimer’s disease and related dementia (ADRD) research studies is crucial for understanding disease progression and consequences, but enrolling participants into research is often a difficult task. Traditional methods of participant recruitment such as manual categorization are time-consuming and inaccurate or incomplete data resulting from participant self-report can lead to suboptimal referrals to various studies. This time, effort, and money spent on recruiting participants who do not qualify or are at risk of not enrolling wastes limited resources. Enrolling participants into the best studies for them based on a series of characteristics (age, sex, MRI eligibility) is made even more difficult when considering how to co-enroll (enroll in multiple studies) participants. In order to address these complications, we developed a novel categorizing program that uses algorithms to efficiently and accurately sort participants based on their specific characteristics and qualifications. Using a real-world dataset from the Michigan Alzheimer’s Disease Research Center, this program was tested for participant referral accuracy. The goal of the program was to quickly and efficiently sort participants into appropriate research programs, resulting in a more efficient and effective recruitment process. The results of the project demonstrated the effectiveness of using machine learning to generate a decision tree for sorting research study participants based on diagnosis, sex, and age. The decision tree provided an efficient and accurate method for categorizing participants into appropriate studies based on their characteristics and the needs of each study. Overall, the findings highlight the potential of machine learning and decision trees as valuable tools for improving participant recruitment and study efficiency in the field of participant-based research. A number of limitations are important to note: It is important to note that the specific techniques and methods used in this study may not be generalizable to all research studies or datasets. Additionally, the quality and completeness of the data collected may have impacted the accuracy and effectiveness of the models generated. Nonetheless, the methods used in this study provide a useful starting point for future research in the area of sorting research participants into studies. Implementation could lead to an increase in the number of participants who enroll in ADRD research studies and help to speed up the discovery of new treatments and therapies for the disease.
