👋 Welcome to Self-Harm Risk Assessment (Deep ANN–powered)
Self-Harm Risk Assessment is a quick, confidential Self-Harm Risk Assessment screening tool that analyzes your responses using a Deep Artificial Neural Network (ANN) trained to detect patterns associated with emotional distress and risk.
📬 Contact & Collaboration
For questions, feedback, academic collaboration, or responsible use inquiries:
Dr. Vikas Gaur — drvikasgaur@gmail.com
🧠 Why this is NOT an ordinary mental-health scale
Traditional screening tools typically use fixed scoring rules (simple sums + thresholds). Self-Harm Risk Assessment uses a Deep ANN that, during training, self-learned complex relationships from the same 33 features, including:
- Non-linear effects (a “Yes” may matter more only in certain combinations),
- Feature interactions (two mild signals together can become meaningful),
- Hidden response signatures that fixed scoring cannot represent.
This enables the model to capture patterns beyond ordinary scales.
🔒 Important: the model does not learn from your answers in real time while you use the app. It runs a fixed trained model to keep screening consistent and safe.
🔍 Explainable-AI style transparency
After your result, the app will show:
- Which themes were activated (based on items you marked “Yes”), and
- Which specific questions you marked “Yes” to under those themes, so you can see why the model is leaning toward a particular screening category.
⚡ High-throughput / mass screening (deployment-scaled)
Each assessment is a lightweight CPU inference. With proper deployment scaling (multiple instances/replicas + load balancing + adequate compute), the system can support large population screening programs (from thousands up to very large cohorts).
✅ How to use
- Click Start Screening
- Choose language (English/Hindi)
- Read the ethical notice and give consent
- Answer all 33 questions and submit
🛡️ Data is used only for this session and not stored permanently (unless the deployment adds storage). ⚠️ This is screening/education only — not a medical diagnosis.