Dominion Akinrotimi
Back to Projects
Python Streamlit Scikit-Survival Machine Learning

OncoVision: METABRIC Risk Stratifier 🧬

The problem: doctors need to know which breast cancer patients face the highest risk, so the most aggressive treatment and closest follow-up go where they matter. OncoVision sorts patients into low, medium and high risk from clinical data available at diagnosis.

Open the live app View GitHub Repo
OncoVision dashboard

Model comparison

Every model was scored with 5-fold cross-validation on 1,584 training patients. The winner was picked on that score alone, and the 397 test patients were used once, at the end.

ModelC-index (5-fold CV)C-index (test)10-year AUC (test)
Random Survival Forest0.686 ± 0.0180.6910.742
Gradient Boosting Survival0.686 ± 0.0140.6920.749
CoxNet (elastic net)0.680 ± 0.0150.6680.711
Cox proportional hazards0.678 ± 0.0140.6640.707

The two tree ensembles are effectively tied. The Random Survival Forest was kept because it was slightly ahead in cross-validation and three times faster to train.

Risk groups on held-out patients

Risk groupPatientsDied during follow-up10-year survivalMedian survival
Low12632.5%81%257 months
Medium13956.8%64%172 months
High13282.6%32%86 months
Kaplan-Meier survival curves for the low, medium and high risk groups, clearly separated

The curves separate clearly (log-rank p = 3 × 10-22).

What drives the risk

Permutation importance and SHAP agree on the top factors, and they're the ones clinicians already use:

How I kept it honest

Data: METABRIC, 2,509 patients. I dropped 528 with no recorded outcome rather than guess them, which left 1,981 patients (1,144 deaths) and 21 clinical features.

No leakage: relapse status and vital status restate the outcome, so they were removed. The 1,258 missing feature values are filled inside the model pipeline, learned from training patients only.

Split: stratified 80/20 by outcome, with fixed hyperparameters, so nothing was tuned on the test set.

Research prototype. Not a medical device and not for treatment decisions. Risk groups describe groups of patients, not individuals.

Need an AI or machine learning system like this?

I build and deploy models that solve real problems, from clinical research to products in production.