How rohithsuresh21/mental-health-digital-twin-ai-assistant works
Mental Health Digital Twin is a local-first mental-health monitoring and early-warning product. The patient workflow is daily check-in or chat import, feature extraction, personal baseline calibration, forecasting and anomaly analysis, clinical risk classification, then explanations and patient-facing results. The graph separates the daily portal and persistent records from the analysis pipeline, and includes the guide and PDF output as documented product capabilities. Some subsystem links rely on README descriptions or the SOURCE INDEX; unsampled implementation details are not inferred.
Main entry points: unified_pipeline.py Pipeline Orchestrator · routes.py Daily Portal API · App.tsx Patient App · app.py Flask API
Components 25
Patient Portal
- Patient App
User Interface/src/App.tsx - Check-in & Import
User Interface/src/components/PatientIntakePortal.tsx - Patient Data View
User Interface/src/components/PatientTelemetry.tsx - Patient Data Hook
User Interface/src/hooks/usePatientData.ts
Data Intake
- Daily Portal API
daily_portal/routes.py - Chat Import Parsers
data_parser - Daily Records
daily_portal/db.py - Record Assembly
daily_portal/dataset.py
Analysis Pipeline
- Feature Extraction
Stage_1/Extract_features.py - Personal Baseline
stage_2/baseline.py - Temporal Binning
stage_2/temporal_bin.py - Risk Forecasting
stage_3/tft_model.py - Anomaly Ensemble
stage_4/anomaly_pipeline.py - Risk Classification
Stage_5/inference.py - Pipeline Orchestrator
unified_pipeline.py - Analysis Runner
pipeline_runner.py
Results & Guidance
- Feature Explanations
XAI/XAI.py - Medical Summary PDF
pdf_generator.py - Analysis Results
User Interface/src/insights/ResultsPage.tsx - Explanation View
User Interface/src/insights/ExplainPage.tsx - Forecast View
User Interface/src/insights/ForecastPage.tsx - Offline Guide
User Interface/src/guide/Guide.tsx
Local Runtime
- Flask API
app.py - Desktop Launcher
desktop/server.py
Other components
- Patient
Connections 29
- Patient → Check-in & Import: enters check-ins
README.md - Check-in & Import → Daily Portal API: submits entries
README.md - Patient → Chat Import Parsers: imports chats
README.md - Daily Portal API → Chat Import Parsers: parses imports
daily_portal/routes.py - Daily Portal API → Daily Records: saves entries
daily_portal/routes.py - Daily Portal API → Feature Extraction: extracts features
daily_portal/routes.py - Daily Portal API → Personal Baseline: refreshes baseline
daily_portal/routes.py - Record Assembly → Daily Records: reads entries
daily_portal/dataset.py - Patient App → Check-in & Import: renders portal
User Interface/src/App.tsx - Patient App → Patient Data Hook: loads patient data
User Interface/src/App.tsx - Patient Data Hook → Daily Portal API: requests status
User Interface/src/hooks/usePatientData.ts - Patient Data View → Patient Data Hook: uses patient data
User Interface/src/components/PatientTelemetry.tsx - Flask API → Daily Portal API: registers portal API
app.py - Flask API → Analysis Runner: runs analysis
app.py - Analysis Runner → Record Assembly: uses assembled records
pipeline_runner.py - Analysis Runner → Pipeline Orchestrator: runs pipeline
pipeline_runner.py - Pipeline Orchestrator → Feature Extraction: extracts vectors
unified_pipeline.py - Pipeline Orchestrator → Personal Baseline: normalizes vectors
unified_pipeline.py - Pipeline Orchestrator → Temporal Binning: bins entries
unified_pipeline.py - Pipeline Orchestrator → Risk Forecasting: forecasts risk
unified_pipeline.py - Pipeline Orchestrator → Anomaly Ensemble: scores anomalies
unified_pipeline.py - Pipeline Orchestrator → Risk Classification: classifies risk
README.md - Pipeline Orchestrator → Feature Explanations: generates attributions
unified_pipeline.py - Flask API → Medical Summary PDF: generates summary
app.py - Patient App → Analysis Results: shows results
User Interface/src/App.tsx - Patient App → Explanation View: shows explanations
User Interface/src/App.tsx - Patient App → Forecast View: shows forecasts
User Interface/src/App.tsx - Patient App → Offline Guide: opens guide
User Interface/src/App.tsx - Desktop Launcher → Flask API: serves locally
desktop/server.py