Traditional assessment ratings are

• Expensive to administer
• Vulnerable to bias
• Imprecise
• Plagued by poor resolution and sensitivity to treatments effects
• Gathered in an office or clinic, not in the real world
• These shortcomings decrease the signal-to-noise ratio of clinical studies and increase costs

Digital biomarkers increase the likelihood of successful clinical trials by

• Improving and enriching patient selection
• Monitoring interview quality
• Validating endpoints
• Characterizing treatment-responsive groups
• Hypothesis generation
• Measuring the impact of treatment on real world function

Challenges in the use of digital biomarkers

• Plagued by poor resolution and sensitivity to treatment
• Gathered in an office or clinic, not in the real world
• Complexity
• High burden for participants and sites
• Lack of clear purpose
• Black box modeling that is
  • Uninterpretable
  • Unreproducible

Our solutions

• “Sparse” approaches to technology & hardware
• No- to low-burden strategies for patients & sites
• Limited, clinically relevant feature sets
• “Intelligent” AI/ML with a priori selection of variables for inclusion
• Convert massive datasets to interpretable findings
• Evaluate progress in real-time in the participant’s natural environment