Machine learning
Machine learning (ML) is a field of artificial intelligence concerned with computational methods that identify patterns in data and use them to generate predictions, classifications, or decisions.
Machine learning (ML) is a field of artificial intelligence concerned with computational methods that identify patterns in data and use them to generate predictions, classifications, or decisions. In contrast to rule-based systems, ML models estimate relationships from examples, such as clinical variables, laboratory measurements, medical images, text, or molecular profiles. Common approaches include Logistic Regression, support vector machine, random forest, XGBoost, and LightGBM; deep learning represents a related family of models that uses multilayer neural networks. Model performance may be assessed with measures such as Accuracy, sensitivity, specificity, F1 score, and the Area Under the Receiver Operating Characteristic Curve.
In biomedicine, ML is used for risk stratification, diagnosis, prognosis, image segmentation, clinical decision support, and analysis of high-dimensional biological data. Its potential value lies in integrating multiple predictors—including static patient characteristics and changing biomarkers—to estimate individualized risks. Interpretability methods such as SHapley Additive exPlanations and XGBoost-SHAP can help describe how input variables contribute to a prediction, while Decision Curve Analysis can assess the potential clinical usefulness of a model. ML is therefore a computational tool rather than a biological treatment or pharmacological mechanism; its clinical significance depends on data quality, validation, interpretability, and successful integration into care.
- Comparison of an ensemble machine learning model to a Cox regression model to predict colorectal cancer risk among people with HIV using retrospective nationwide cohort data in Sweden: a study protocol. PMID 42716689
Where the papers sit
12 papers study machine learning directly. Those 12 are one subject: Clinical Outcome Prediction. Clinical and radiomic data are being combined to predict cancer risk, treatment response, complications and mortality. Interpretable models and comparisons with conventional regression recur, but the applications remain disease- and endpoint-specific. No way of splitting those 12 scores better than chance. 1 new direction follows.
Machine learning is used to prescribe adaptive treatment regimes rather than merely predict clinical outcomes
Smoking cessation trial data and ordinal-outcome treatment regimes are used in the Bayesian machine-learning study to estimate optimal dynamic treatment regimes, with decision rules adapted to patients’ treatment histories and changing disease status rather than using machine learning only for prediction, classification, or risk stratification. This changes its role from forecasting an outcome to selecting a sequence of treatments while quantifying uncertainty in those decisions 42639730Aug.
Recent Findings on Machine learning
Clinical Outcome Prediction: Patients with cancer, critical illness, hematological malignancies, and surgery are being risk-stratified for recurrence, mortality, thrombosis, weakness, and residual disease using clinical data, biomarkers, and MRI radiomics 42679396Sep42642795Aug42618846Aug42530089Jul42527084Jul. The models increasingly combine dynamic biomarkers or radiomic features with clinical predictors, although clinical variables supplied most predictive value for residual meningioma and radiomics added only modestly 42618846Aug42679396Sep. Bayesian machine learning estimates optimal dynamic treatment regimes with ordinal outcomes, while GUSL provides interpretable prostate MRI segmentation and embedding models detect NPD-trait markers 42639730Aug42364419Jun42526023Jul. Random forest, XGBoost, and LightGBM showed useful discrimination in selected cohorts, but local-recurrence prediction after carbon-ion radiotherapy had ROC-AUC 0.622, and external mortality discrimination fell to 0.764 42642795Aug42618846Aug42530089Jul42527084Jul. SHAP, calibration, Decision Curve Analysis, and web-based calculators make individual risk estimates more interpretable and actionable 42642795Aug42618846Aug42530089Jul. Prospective validation and workflow integration remain priorities: a Swedish protocol will compare ensemble stacking with Cox regression, while surgical and chemotherapy teams are being studied for ML-CDSS adoption 42716689Sep42150783May42150828May.
Written from 12 PubMed abstracts, each one cited by PMID above. Published: 2026-08-29. Last written: 2026-09-11 by GPT. Drafted by language models from published abstracts; not medical advice.