Periodic Reporting for period 1 - ADAPTE (A novel and accurate emotion recognition system for real-time and continuous patient monitoring in psychiatry)
Reporting period: 2023-11-01 to 2025-04-30
Mental healthcare support faces three critical challenges: low treatment adherence (>60% dropout rate), difficulty in tracking emotional states between clinical visits, and insufficient between-session support. Current methods rely heavily on patient self-reporting and memory, leading to potential gaps in care continuity.
ADAPTE aims to enhance mental healthcare support through an AI-powered patient-centric software solution that:
- Facilitates emotion tracking through natural speech analysis
- Supports patient engagement and adherence
- Bridges communication gaps between clinical sessions
- Provides objective insights to supplement clinical care
The project initially focuses on supporting individuals with Major Depressive Disorder (MDD), with potential expansion to other mental health conditions. Expected impact includes:
- Making emotional wellbeing support accessible to >95% through smartphone-based tools
- Supporting treatment adherence improvement to 75% (from current 40%)
- Enhancing patient-clinician communication
ADAPTE aims to enhance mental healthcare support through an AI-powered patient-centric software solution that:
- Facilitates emotion tracking through natural speech analysis
- Supports patient engagement and adherence
- Bridges communication gaps between clinical sessions
- Provides objective insights to supplement clinical care
The project initially focuses on supporting individuals with Major Depressive Disorder (MDD), with potential expansion to other mental health conditions. Expected impact includes:
- Making emotional wellbeing support accessible to >95% through smartphone-based tools
- Supporting treatment adherence improvement to 75% (from current 40%)
- Enhancing patient-clinician communication
Technical Developments:
- Enhanced speech-based emotion recognition through advanced AI
- Developed robust quality assurance processes
- Created personalised engagement frameworks
Clinical Validation:
- Established partnerships with psychiatric hospitals covering 6,000 MDD patients
- Successfully tested patient conversational pathway v1.0
- Validated patient engagement modules with clinicians
- Implemented secure data management protocols
- Enhanced speech-based emotion recognition through advanced AI
- Developed robust quality assurance processes
- Created personalised engagement frameworks
Clinical Validation:
- Established partnerships with psychiatric hospitals covering 6,000 MDD patients
- Successfully tested patient conversational pathway v1.0
- Validated patient engagement modules with clinicians
- Implemented secure data management protocols
Technical Innovation:
- Novel application of time-series AI analysis to speech patterns
- Integrated emotion recognition with patient engagement tools
- Developed scalable cloud-based architecture for widespread deployment
Market Impact:
- First-to-market patient-centric solution combining continuous monitoring with engagement
- 95% accessibility through smartphone-based approach
- Demonstrated feasibility of continuous emotion monitoring in real-world settings
Next Steps:
- Enhance AI models with larger datasets
- Ensure product-market fit with stakeholders’ feedback
- Iterate patient-centric software with early adopters
- Create sustainable business model
- Novel application of time-series AI analysis to speech patterns
- Integrated emotion recognition with patient engagement tools
- Developed scalable cloud-based architecture for widespread deployment
Market Impact:
- First-to-market patient-centric solution combining continuous monitoring with engagement
- 95% accessibility through smartphone-based approach
- Demonstrated feasibility of continuous emotion monitoring in real-world settings
Next Steps:
- Enhance AI models with larger datasets
- Ensure product-market fit with stakeholders’ feedback
- Iterate patient-centric software with early adopters
- Create sustainable business model