-
-
Patient Profile
-
Medicine taken ,missed logs(plots)
-
Report Generation and AI recommendations
-
Medicine Dose-Event History(Audit Trails)
Inspiration
Medication schedules can be difficult to follow consistently, especially when patients take multiple medicines at different times of the day. Missed or delayed doses can affect medication routines, making it important to identify adherence patterns and provide timely support.
This inspired us to develop MediSense AI, a medication adherence monitoring system designed to help track dose records, identify potential adherence risks, and provide personalized habit recommendations.
What We Built
MediSense AI is a healthcare dashboard that brings medication adherence information into one place. It allows users to explore patient records, monitor medication-taking patterns, and view adherence and risk summaries.
The application includes patient-specific dashboards, dose-event tracking, risk analysis, alerts, personalized recommendations, and downloadable PDF reports. Users can also simulate dose events to observe how changes affect the displayed adherence metrics and risk indicators.
How We Built It
We developed the application using Python and Streamlit for the interactive dashboard. The backend processes patient profiles and medication records to calculate adherence metrics and assess potential risks. A recommendation component provides habit suggestions based on the available information.
We focused on connecting these components into a straightforward workflow, allowing users to select a patient, review relevant metrics, explore alerts, and generate an individual report.
Challenges We Faced
One of our main challenges was integrating the backend logic with the dashboard while keeping patient selection, calculations, and reports consistent. We also worked through application errors during integration and tested the report-generation workflow to ensure that patient-specific information was reflected in the output.
Another challenge was presenting multiple types of information, including adherence percentages, risk levels, dose statuses, and recommendations, in a way that was easy to understand.
What We Learned
Building MediSense AI helped us understand how data processing, rule-based risk analysis, and interactive interfaces can work together in a practical healthcare-oriented application. We gained experience integrating Python modules, debugging application issues, organizing patient data, and generating PDF reports.
The project also taught us the importance of communicating the limitations of healthcare prototypes. MediSense AI is intended to support medication adherence monitoring and awareness; it does not replace medical advice or clinical decision-making.
What's Next
We would like to improve the system with more extensive testing, better visualization of adherence trends, and further evaluation of its risk assessment and recommendation methods. Our goal is to make medication tracking more understandable and useful while keeping the system practical and accessible.
Log in or sign up for Devpost to join the conversation.