Study Rationale:
The rich and longitudinal (gathered over time) dataset available through the Parkinson's Progression Markers Initiative provides a unique opportunity to gain insights into the progression of Parkinson's disease. By using this data, we hope to find patterns in cohorts of patients to provide new insights into the disease.
Hypothesis:
Using all components of the data including unstructured text and machine learning tools and algorithms, it is our hope that otherwise unnoticed patterns will be identified. Using these tools in other disease states, we have successfully identified this type of detail.
Study Design:
Initially, data will be pre-processed, that is, run through existing algorithms to determine how much information can be gained with current tools. We will create a robust assessment score after tuning the models for the specific disease findings. With feedback from subject matter experts at MJFF, we will then refine the models and further gather input from subject matter experts.
Impact on Diagnosis/Treatment of Parkinson's Disease:
Understanding how the disease progresses in different patient cohorts may eventually inform more personalized treatments and, hopefully, improve quality of life and outcomes.
Next Steps for Development:
After this initial phase where patient records and genomic information will be separately evaluated, we hope to combine patient record data and genomic data into a combined model.