How xzhe0014/fit4701-4702-fyp works
Cardio Vision AI is a web app for cardiovascular disease prediction. Users enter patient data manually, upload a file, or select a dataset, then request predictions and can view results. The backend exposes prediction, dataset, patient-result, and authentication APIs; inference preprocesses inputs and invokes a loaded model. Users can also save and manage patient/result records. The repository includes a separate model-training workflow. Edges reflect excerpts, SOURCE INDEX references, or README statements; unsampled page behavior and unshown connections are omitted.
Main entry points: server.js HTTP server · PredictionPage.js Prediction page · ensemble.js Ensemble training · App.js Page routing
Components 34
Web application
- Page routing
cardio-vision-ai/frontend/src/App.js - Auth context
cardio-vision-ai/frontend/src/context/AuthContext.js - Prediction page
cardio-vision-ai/frontend/src/pages/PredictionPage.js - Dataset selector
cardio-vision-ai/frontend/src/components/DatasetSelector.js - File uploader
cardio-vision-ai/frontend/src/components/FileUploader.js - Manual entry
cardio-vision-ai/frontend/src/components/ManualEntry.js - Results page
cardio-vision-ai/frontend/src/pages/ResultsPage.js - Result display
cardio-vision-ai/frontend/src/components/PredictionResults.js - Prediction charts
cardio-vision-ai/frontend/src/components/PredictionCharts.js
Backend APIs
- HTTP server
cardio-vision-ai/backend/server.js - Auth routes
cardio-vision-ai/backend/routes/authRoutes.js - Auth controller
cardio-vision-ai/backend/controllers/authController.js - JWT guard
cardio-vision-ai/backend/middleware/authMiddleware.js - Prediction routes
cardio-vision-ai/backend/routes/predictionRoutes.js - Prediction controller
cardio-vision-ai/backend/controllers/predictionController.js - Dataset routes
cardio-vision-ai/backend/routes/datasetRoutes.js - Dataset processing
cardio-vision-ai/backend/controllers/datasetController.js - Patient routes
cardio-vision-ai/backend/routes/patientRoutes.js - Patient results
cardio-vision-ai/backend/controllers/patientController.js
Persistence
- DB connection
cardio-vision-ai/backend/config/db.js - User records
cardio-vision-ai/backend/models/User.js - Dataset records
cardio-vision-ai/backend/models/Dataset.js - Patient records
cardio-vision-ai/backend/models/Patient.js - Saved predictions
cardio-vision-ai/backend/models/PredictionResult.js
Model workflow
- Model loader
cardio-vision-ai/backend/models/loadModel.js - Input preprocessing
cardio-vision-ai/backend/utils/preprocessInput.js - Training runner
cardio-vision-ai/backend/trainedModels/trainingScript/run.js - Data processing
cardio-vision-ai/backend/trainedModels/trainingScript/processData.js - Ensemble training
cardio-vision-ai/backend/trainedModels/trainingScript/ensemble.js - Label encoding
cardio-vision-ai/backend/trainedModels/trainingScript/labelEncoder.js - Train test split
cardio-vision-ai/backend/trainedModels/trainingScript/trainTestSplit.js - Model metrics
cardio-vision-ai/backend/trainedModels/trainingScript/metrics.js
Other components
- User
- MongoDB
Connections 35
- User → Page routing: uses
- Page routing → Prediction page: routes to
cardio-vision-ai/frontend/src/App.js - Page routing → Results page: routes to
cardio-vision-ai/frontend/src/App.js - User → Auth context: authenticates
cardio-vision-ai/frontend/src/context/AuthContext.js - Prediction page → Dataset selector: renders
cardio-vision-ai/frontend/src/pages/PredictionPage.js - Prediction page → File uploader: renders
cardio-vision-ai/frontend/src/pages/PredictionPage.js - Prediction page → Manual entry: renders
cardio-vision-ai/frontend/src/pages/PredictionPage.js - Results page → Result display: renders
cardio-vision-ai/frontend/src/pages/ResultsPage.js - Results page → Prediction charts: renders
cardio-vision-ai/frontend/src/pages/ResultsPage.js - Prediction page → HTTP server: posts prediction
cardio-vision-ai/frontend/src/pages/PredictionPage.js - Auth context → HTTP server: calls auth API
cardio-vision-ai/frontend/src/context/AuthContext.js - HTTP server → DB connection: connects
cardio-vision-ai/backend/server.js - HTTP server → Auth routes: mounts
cardio-vision-ai/backend/server.js - HTTP server → Prediction routes: mounts
cardio-vision-ai/backend/server.js - HTTP server → Dataset routes: mounts
cardio-vision-ai/backend/server.js - HTTP server → Patient routes: mounts
cardio-vision-ai/backend/server.js - Auth routes → Auth controller: dispatches
cardio-vision-ai/backend/routes/authRoutes.js - Auth routes → JWT guard: protects user route
cardio-vision-ai/backend/routes/authRoutes.js - Auth controller → User records: reads and writes
cardio-vision-ai/backend/controllers/authController.js - JWT guard → User records: loads user
cardio-vision-ai/backend/middleware/authMiddleware.js - Prediction routes → Prediction controller: dispatches
cardio-vision-ai/backend/routes/predictionRoutes.js - Prediction controller → Input preprocessing: preprocesses input
cardio-vision-ai/backend/controllers/predictionController.js - Prediction controller → Model loader: loads model
cardio-vision-ai/backend/controllers/predictionController.js - Dataset routes → Dataset processing: dispatches
cardio-vision-ai/backend/routes/datasetRoutes.js - Dataset processing → Dataset records: reads and writes
cardio-vision-ai/backend/controllers/datasetController.js - Patient routes → Patient results: dispatches
cardio-vision-ai/backend/routes/patientRoutes.js - Patient routes → JWT guard: protects routes
cardio-vision-ai/backend/routes/patientRoutes.js - Patient results → Patient records: reads and writes
cardio-vision-ai/backend/controllers/patientController.js - Patient results → Saved predictions: reads and writes
cardio-vision-ai/backend/controllers/patientController.js - DB connection → MongoDB: connects
README.md - Training runner → Data processing: uses
cardio-vision-ai/backend/trainedModels/trainingScript/run.js - Training runner → Ensemble training: runs training
cardio-vision-ai/backend/trainedModels/trainingScript/run.js - Ensemble training → Label encoding: uses
cardio-vision-ai/backend/trainedModels/trainingScript/ensemble.js - Ensemble training → Train test split: uses
cardio-vision-ai/backend/trainedModels/trainingScript/ensemble.js - Ensemble training → Model metrics: uses
cardio-vision-ai/backend/trainedModels/trainingScript/ensemble.js