How vaibhav-nhayade/medinsight-ai works
MedInsight AI accepts medical-report uploads, stores document metadata and files, parses report text into structured test results, classifies results against report-provided ranges, and returns findings and a summary. The backend API and document services connect upload and analysis to those AI stages. The README also describes RAG, historical comparison, patient questions, OCR, and safety/privacy capabilities, but the current tree does not show their implementations; the extraction pipeline shown supports PDF text extraction. Edges reflect sampled code, SOURCE INDEX references, or README statements; unsampled or planned features are not represented as implemented.
Main entry points: orchestrator.py Analysis orchestrator · reports.py Report endpoints · analysis.py Analysis endpoint · document_service.py Document lifecycle
Components 26
API and workflow
- FastAPI application
backend/main.py - Report endpoints
backend/routes/reports.py - Document endpoints
backend/routes/documents.py - Analysis endpoint
backend/routes/analysis.py - Report responses
backend/schemas/report.py - Report analysis service
backend/services/report_service.py
Document management
- File storage
backend/services/file_storage.py - Document lifecycle
backend/services/document_service.py - Document record
backend/models/document.py - Document repository
backend/repositories/document_repository.py - SQLite database
backend/database/connection.py
Report extraction
- Report parser
ai/extraction/report_parser.py - Document processing
ai/extraction/pipeline.py - PDF text extraction
ai/extraction/pdf.py - Test result extraction
ai/extraction/result_parser.py - Parameter extraction
ai/extraction/parameter_extractor.py - Reference extraction
ai/extraction/reference_extractor.py - Extracted result model
ai/extraction/schema.py
Result analysis
- Analysis orchestrator
ai/analysis/orchestrator.py - Range analysis
ai/analysis/service.py - Range classification
ai/analysis/classification.py - Analyzed result model
ai/analysis/result.py - Finding generation
ai/analysis/finding_service.py - Medical findings
ai/analysis/finding.py - Analysis summary
ai/analysis/summary.py
Other components
- Report user
Connections 29
- Report user → Report endpoints: uploads report
README.md - FastAPI application → Report endpoints: registers route
backend/main.py - FastAPI application → Document endpoints: registers route
backend/main.py - FastAPI application → Analysis endpoint: registers route
backend/main.py - Report endpoints → File storage: stores file
backend/routes/reports.py - Report endpoints → Document lifecycle: registers document
backend/routes/reports.py - Report endpoints → Report responses: returns response
backend/routes/reports.py - Document endpoints → Document lifecycle: manages document
backend/routes/documents.py - Analysis endpoint → Report analysis service: requests analysis
backend/routes/analysis.py - Analysis endpoint → Document lifecycle: checks document
backend/routes/analysis.py - Analysis endpoint → Report responses: returns response
backend/routes/analysis.py - Document lifecycle → Document repository: persists metadata
backend/services/document_service.py - Document repository → SQLite database: reads and writes
backend/repositories/document_repository.py - Document repository → Document record: maps records
backend/repositories/document_repository.py - Report analysis service → Analysis orchestrator: runs analysis
backend/services/report_service.py - Analysis orchestrator → Report parser: parses report
ai/analysis/orchestrator.py - Analysis orchestrator → Range analysis: analyzes results
ai/analysis/orchestrator.py - Analysis orchestrator → Finding generation: generates findings
ai/analysis/orchestrator.py - Analysis orchestrator → Analysis summary: builds summary
ai/analysis/orchestrator.py - Report parser → Document processing: processes document
ai/extraction/report_parser.py - Report parser → Test result extraction: parses results
ai/extraction/report_parser.py - Document processing → PDF text extraction: extracts PDF text
ai/extraction/pipeline.py - Test result extraction → Parameter extraction: extracts parameters
ai/extraction/result_parser.py - Test result extraction → Reference extraction: extracts ranges
ai/extraction/result_parser.py - Parameter extraction → Extracted result model: builds results
ai/extraction/parameter_extractor.py - Reference extraction → Extracted result model: builds ranges
ai/extraction/reference_extractor.py - Range analysis → Range classification: classifies values
ai/analysis/service.py - Range analysis → Analyzed result model: builds results
ai/analysis/service.py - Finding generation → Medical findings: creates findings
ai/analysis/finding_service.py