Elder Care MultiMorbidity Analysis and Prediction

• 

  ElderCare Multimorbidity Analysis and Prediction

Inspiration

My mother is approaching her senior years. The advanced years come with chronic medical complications. As a scientist with the desire to see her situation improve, I opted to use the knowledge and skills at my disposal to find a way to mitigate these complications.

What it does

A prototype tool for chronic disease management. It uses AI and multi-omics to analyze multimorbidity, build interactive chronic disease networks and predict risks from EHR, SDOH, and wearable data.

How we built it

We used synthetic datasets and integrated these different data sources into predictive and analytical methods to yield insights.

Challenges we ran into

Data Integration: Combining diverse datasets from various sources. Interpretability: Ensuring model explainability for clinical applications. Longitudinal Analysis: Addressing temporal disease progression within complex networks.

Accomplishments that we're proud of

The unique thing about the project is the inclusion of genetic information on diseases. Incorporating multi-omics across a myriad of chronic conditions while drawing insights from a cross section of the population.

We found and applied two algorithms for dimensionality reduction. The two algorithms accepted a range of diseases and allowed multiple insights.

What we learned

Integration of multi-omics data enables applications in biomarker discovery and drug target identification.

What's next for Elder Care MultiMorbidity Analysis and Prediction

Given that this is just the prototype, future extensions include integrating real-world patient feedback, testing the system on longitudinal cohorts, explaining how the disease network came about, and enhancing computational efficiency for broader adoption in clinical settings

Built With

• machine-learning
• python
• streamlit