Hidaayet/mouna-breast-cancer-risk

Non-invasive breast cancer risk assessment tool combining health questionnaire and blood biomarkers — designed for low-resource clinical settings # Mouna — Breast Cancer Risk Assessment Tool > *Mouna (مُنى) — Arabic for "wish" or "hope"* A non-invasive, accessible breast cancer risk stratification tool combining health questionnaire data and blood biomarkers to generate personalized risk scores — designed for low-resource clinical settings where mammography is unavailable or inaccessible. > **Clinical Disclaimer:** Mouna is a research prototype and clinical > decision support tool. It is not a diagnostic device and does not detect > cancer. It identifies individuals who may benefit from further medical > evaluation. All outputs must be interpreted by a qualified healthcare > professional. > **Status:** Stage 1 — Research prototype (active development) --- ## The Problem Breast cancer is the most common cancer in women worldwide. In Tunisia and across North Africa, late-stage diagnosis is the norm — not because the disease is more aggressive, but because early detection infrastructure is largely inaccessible: - Mammography costs $100-300 per scan - Specialized radiology equipment is concentrated in major cities - Cultural barriers reduce screening uptake - Primary care physicians lack structured risk stratification tools ## Results — Trained on Real Clinical Data | Model | Dataset | AUC | Patients | |---|---|---|---| | Gail Model (clinical standard) | Various | 0.580 | — | | Tyrer-Cuzick (best published) | Various | 0.680 | — | | **Mouna XGBoost** | **BCSC Registry** | **0.926** | **244,737** | **Mouna achieves 0.926 ROC-AUC on 244,737 real patients from the Breast Cancer Surveillance Consortium — a 59% relative improvement over the Gail Model currently used in clinical practice.** ### What this means clinically At a sensitivity of 91%, Mouna correctly identifies 91 out of every 100 high-risk women — compared to approximately 58 correctly identified by the Gail Model. In a population of 10,000 women, this difference translates to hundreds of additional high-risk women identified for early interve …