How samridh-kumar/cifa works
Cifa is presented as an AI-assisted cancer-diagnosis website. The implemented workflow is image selection in the browser, submission to Flask’s brain or oral prediction endpoint, model inference, and display of the returned prediction. The repository also contains training datasets and model notebooks. The README describes account management, Firebase storage, lung-cancer detection, and care recommendations, but the sampled runtime source does not establish those flows; they are not drawn as implemented components.
Main entry points: app.py Flask API · script.js Upload and submit · index.html Diagnosis page · Dataset_for_Brain_tumor Brain image data
Components 13
Web interface
- Diagnosis page
index.html - Upload and submit
script.js - Prediction display
result.html
Image inference
- Flask API
app.py - Brain inference
app.py - Oral inference
app.py - Brain model
models/brain_tumor_model.pkl - Oral model
models/oral_cancer_mode.keras
Model development
- Brain image data
Dataset_for_Brain_tumor - Oral image data
Oral_cancer_dataset - Brain model notebook
Full code for models/BrainTumor_ML_Model.ipynb - Oral model notebook
Full code for models/Oral_Cancer_ML_Model.ipynb
Other components
- Patient
Connections 11
- Patient → Diagnosis page: opens
- Diagnosis page → Upload and submit: loads
- Upload and submit → Flask API: posts image
- Flask API → Brain inference: dispatches
- Flask API → Oral inference: dispatches
- Brain inference → Brain model: predicts with
- Oral inference → Oral model: predicts with
- Flask API → Upload and submit: returns prediction
- Upload and submit → Prediction display: navigates with result
- Brain image data → Brain model notebook
- Oral image data → Oral model notebook