AIquitas: LLM Solutions to Improve Efficiency and Inclusivity in Clinical Trial Recruitment Process
#LLM #BioBert #Healthcare #Inclusivity #UX #UI #Research #Prototyping #ProjectManagement

Project Overview
AIquitas results from an intensive one-year project created for the prestigious competition known as the 3rd Coast Augmented Intelligence for Health Bowl. This esteemed competition, organized by the Institute for Augmented Intelligence in Medicine (I.AIM) at Northwestern University, offers an annual platform for higher education institutions to tackle Health Disparities through the power of AI.
The Solution
AIquitas is a groundbreaking solution developed to address the significant disparities in clinical trial participation and its impact on diverse populations.
Patient-Centered Approach:
AIquitas places patients at the forefront by leveraging advanced Machine Learning (ML) and Natural Language Processing (NLP) techniques to parse clinical trial criteria and patient billing codes. This enables AIquitas to suggest eligible clinical trials to patients and their healthcare providers based on a calculated compatibility score, ensuring a personalized and patient-centric approach to trial participation.
Streamlined Recruitment Process:
AIquitas streamlines patient recruitment by intelligently matching patients with appropriate clinical trials. By automating and optimizing this critical aspect, AIquitas saves researchers, healthcare providers, and patients time and resources, ensuring efficient trial enrollment.
​
Addressing Disparities and Bias:
AIquitas also provides capabilities for analyzing patient demographics for any trial, allowing for explainable findings and potential improvements in trial language for clinical trial researchers and regulatory agencies. By tackling the persistent underrepresentation of marginalized and underrepresented communities in clinical trials, AIquitas aims to rectify the biases and limitations that arise from the lack of diversity in collected data.

Award
1st Place Winner of $35,000
Team:
Kate Yuwei Guo
Alexander Chih-Chieh Chang
Rabira Jemal Tusi
Katelin Lauren Rimando Avenir
Aditya Singh
Anirudh Vaidhyaa Venkatasubramanian
Shiqi Liang
Significance: Health Disparity in Clinical Trial
Clinical trial participation in the United States does not reflect the diversity of disease populations.
While about 43% of the US population is non-white, less than 10% of the population in clinical trials over the last 25 years were people of color.
This imbalance in clinical research inclusion has been shown to result in severe limitations and biases in the collected data, with significant implications for downstream analyses, from developing therapeutic indexes to drug safety solutions, risk predictions, and toxicity estimates, resulting in poorer health outcomes for these excluded and vulnerable populations.
​
These issues raise additional concerns that underrepresented minorities and marginalized populations may not have equitable opportunities to access potentially life-saving treatments from clinical trials.
More Projects
Voice User Interface User Experience, Multimodal Design
Healthcare, Digital Service, User Experience, User Interface
AR, User Experience, User Interface, Multimodal Design
Key Feature Prototype
AIquitas can be integrated with any hospital Patient management or EHR system to introduce patient-centered clinical trial selection and match service for patients in need. It would be used by both patient and physicians to collaborate in clinical trial selection process.
​
Below is the prototype that shows the process step by step.
3. Review the trial details: AIquitas technology would highlight the matched inclusion and exclusion criteria text based on Patient EHR, this would facilitate physicians' clinical decision-making process by reducing their cognitive load.

1. Patient onboarding: The patients could indicate their preference and sync their medical profile with AIquitas for the best clinical trial match.


My Role
-
Contribution to research and product management through product thinking and problem-framing expertise.
-
Led stakeholder interviews to identify target stakeholders in the complex landscape of clinical trials.
-
Utilized product impact metrics (social impact, economic viability, data availability, and technological feasibility) to guide decision-making.
-
Created a compelling prototype for the final competition presentation, showcasing the effectiveness and accessibility of our solution.
More Details are coming...
2. Filter trials according to patient medical record: The patients could select potential trials according to compatibility and distance, they can also forward the information to their physicians.

4. Compare clinical trials: The user could select two clinical trials and compare their compatibility details side by side.



