At Rune Labs, we are constantly striving to improve the lives of those affected by Parkinson’s Disease. Today, we are excited to introduce Generative AI-enabled StrivePD Reports, our first feature leveraging StrivePD-AI. Users will receive these enhanced monthly reports of their symptoms, medications, exercise routines, and overall disease progression. With these enhanced StrivePD Reports, we aim to empower people with Parkinson’s, caregivers, and clinicians with data-driven insights to optimize care and improve outcomes.
Discover StrivePD-AI Reports: Your Personalized Health Insights
Each month, the GenAI model generates a detailed, HIPAA-compliant clinical report that is sent to each user by email. The report uses a motivational framework to meet you where you are in your Parkinson’s journey, offering encouragement to continue logging experiences and informing your doctor about falls or balance issues as these are high-risk events. The report includes a comprehensive summary of notes, medication compliance, exercise logs, and symptom fluctuations, highlighting key changes from the previous period and providing recommendations that can all be used to guide discussions during clinician appointments.
Comprehensive Insights: Symptom Trends, Activity, and Mobility Overview
The report begins with a section that highlights some key metrics that are captured from StrivePD, showing how these have changed from the previous period. For Benjamin Stecher, a dedicated user of StrivePD, these metrics provide invaluable insights into his evolving routine. Recently, Ben started an intensive exercise regime and has noticed significant shifts in his daily health data, which are reflected in the reports. Notice the increase in activity, improvement in mobility, and reduction in tremor per day.
The Symptom Log section uses AI to summarize the symptoms you’ve recorded, highlighting any concerning symptoms that may warrant follow-up with your clinician, and adding recommendations or encouragement where appropriate. It then lists the full set of symptoms that you’ve logged in StrivePD, showing the change from the previous period to help you understand in detail what symptoms have increased or decreased. For Ben, while there was a significant reduction in concerning issues, the summary still highlighted the risk of falls and encouraged follow-up with his healthcare team.
The report then shows your average Tremor and Dyskinesia over the course of 24 hours, comparing the previous period to the current period. This gives insight into how your symptoms fluctuate over the course of a day and allows you to compare how this may have changed from the previous period. The chart below shows both a total decrease in tremor per day and a reduction in the fluctuation of tremor over the course of the day. For Ben, the chart below illustrates the impact of his new exercise routine on key health markers. You can notice a marked improvement in his tremor levels and overall physical activity (listed below), demonstrating the influence of his new regimen.
The Activity section then shows a summary of the amount of weekly activity and a list of the individual activities with their total time, giving you a visual representation of how your activities have varied over the previous period. This section also includes an AI-generated summary that captures the key activities you’ve done with some recommendations on how to achieve your goals as well as encouraging a well-balanced exercise routine. In Ben’s chart below, you can see when he started his exercise journey, exceeding over 150 minutes of physical activity per week through strengthening, rowing, walking and more.
Finally, some key mobility metrics (step length, walking speed, double support %) computed from your watch and phone are listed in the Mobility section, with an indication of how these have changed compared to the previous period.
By incorporating this personalized data, StrivePD offers Ben – and others like him – a clearer understanding of how his efforts directly impact his condition. The reports reflect how his increased physical activity has positively influenced his metrics, allowing him to make informed decisions about his ongoing treatment plan.
Behind the Technology: How StrivePD-AI Reports Work
Rune Labs uses a Large Language Model (LLM), to derive insights from your data. By fine-tuning a generic LLM, giving it clinically relevant context about Parkinson’s Disease (curated by our dedicated clinical team), and directing the LLM to respond in specific ways, we’ve developed the ability to summarize the raw data you enter in StrivePD and provide personalized, meaningful summaries.
Under the covers, we use Claude 3.5 (Sonnet) as our base LLM. Claude is developed by the AI company Anthropic, which focuses on building AI responsibly and safely. We have built an architecture around Claude to manage the inputs (the ‘prompts’), tune model parameters, and evaluate the responses, allowing us to implement numerous safeguards and quality checks to the AI-generated summaries.
Your Data, Our Priority: Ensuring Safety and Privacy
As AI opens up new ways to improve patient care, Rune Labs is dedicated to making sure that our AI tools are safe and protect your personal information. We understand that trust is essential, so we have a strong system in place to ensure your data is always secure. This commitment helps build confidence within the StrivePD community by focusing on being open, respecting your privacy, following all necessary regulations, and enhancing your overall experience with us.
Getting Started: When and How You Can Access StrivePD-AI Reports:
Current members of StrivePD have access to StrivePD-AI reports by opting in to StrivePD-AI directly in the app. Advanced StrivePD-AI features will be available to StrivePD Users on a rolling basis.
Download the app on App Store today
