Explainable machine learning for breast cancer prediction in resource-constrained settings: A multi-algorithmic framework integrating shap-based transparency with clinical decision support
Erklärbares maschinelles Lernen zur Brustkrebsprognose in ressourcenarmen Settings: ein multi-algorithmischer Rahmen mit SHAP-Transparenz und klinischer Entscheidungsunterstützung
PLOS Digital Health · 11.09.2026 · 5 (9) · e0001706
DOI: 10.1371/journal.pdig.0001706 ↗Auf einen Blick
Auf dem Wisconsin Breast Cancer Diagnostic Dataset wurden 8 Klassifikationsalgorithmen verglichen. Alle Ensemble- und regularisierten Modelle erreichten im Test AUCs über 0,98; XGBoost und SVM lagen bei 0,996.
Fragestellung
Wie gut und wie interpretierbar lassen sich ML-Modelle zur Brustkrebsdiagnose in ressourcenarmen Settings einsetzen?
Methode
Vergleich von 8 überwachten Klassifikationsalgorithmen mit stratifizierter 10-fach-Kreuzvalidierung und unabhängigem Hold-out-Testset (80:20); SHAP-Analyse zur Interpretierbarkeit.
Ergebnisse
Datensatz mit 569 Feinnadelaspiraten und 30 Merkmalen. XGBoost und SVM: höchste AUC 0,996; Logistische Regression: höchste Genauigkeit 98,25 % und MCC 0,962. SHAP nannte worst perimeter, worst concave points und worst area als wichtigste Prädiktoren.
Schlussfolgerung der Autor:innen
Die Autor:innen berichten ein reproduzierbares, transparentes Framework; die SHAP-Signaturen stimmten mit zytopathologischen Prinzipien überein und unterstützen die Integration interpretierbarer ML-Modelle in ressourcenarme Diagnostik.
Kernaussagen
- Alle Ensemble- und regularisierten Modelle: Test-AUC >0,98
- XGBoost und SVM: höchste AUC 0,996
- Wichtige SHAP-Prädiktoren: worst perimeter, worst concave points, worst area
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
Kurz gefasst
Fachartikel in der Zeitschrift PLOS Digital Health, online erschienen am 11.09.2026. Verfasst von Oluwaseun Adebayo Bamodu und 2 weiteren Autor:innen. Beteiligte Einrichtung: Building Research and Implementation to Drive Growth and Equity in Africa (BRIDGEAfrica), Dar Es Salaam, Tanzania. Die Arbeit nennt 44 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Breast cancer remains the most commonly diagnosed malignancy among women globally, with disproportionately higher mortality rates in low- and middle-income countries (LMICs) where diagnostic delays and limited specialist pathology capacity are widespread. While machine learning (ML) approaches achieve strong predictive performance for cancer classification, algorithmic opacity and absence of interpretability frameworks tailored to resource-constrained environments have impeded clinical adoption. This study bridges the translational gap between predictive accuracy and clinical utility by developing an explainable artificial intelligence (XAI) framework specifically designed for breast cancer diagnosis in underserved healthcare settings. Using the Wisconsin Breast Cancer Diagnostic Dataset (569 fine-needle aspirate cytological specimens with 30 nuclear morphometric features), we systematically benchmarked eight supervised classification algorithms: Logistic Regression, Random Forest, XGBoost, LightGBM, Support Vector Machine (SVM), Gradient Boosting, Decision Tree, and K-Nearest Neighbors, using stratified 10-fold cross-validation and an independent hold-out test set (80:20 split). Performance was evaluated across discriminative and probabilistic metrics, including AUC-ROC, F1-score, Matthews Correlation Coefficient (MCC), and Brier score, and interpretability was operationalized through SHapley Additive exPlanations (SHAP) analysis with global feature importance, cross-model consensus ranking, and individual-level dependence characterization. All ensemble and regularized models achieved test-set AUCs above 0.98, with XGBoost and SVM attaining the highest AUC of 0.996, and Logistic Regression the highest accuracy (98.25%) and MCC (0.962). SHAP analysis consistently identified worst perimeter, worst concave points, and worst area as the dominant predictors, with strong concordance across gradient-boosted models (pairwise Spearman rho: XGBoost–LightGBM 0.86, XGBoost–Random Forest 0.82, Random Forest–LightGBM 0.67). Logistic Regression also demonstrated superior probability calibration, a critical requirement for clinical risk stratification. Collectively, these findings deliver a reproducible, transparent framework whose SHAP-derived signatures align with established cytopathological principles, supporting responsible integration of interpretable ML into resource-limited diagnostic workflows and providing a template for equitable AI deployment in global oncology.
Autor:innen und Institutionen
- Building Research and Implementation to Drive Growth and Equity in Africa (BRIDGEAfrica), Dar Es Salaam, Tanzania.
- Bethesda Achievers Consultancy, Frederick, Maryland, United States of America.
- Department of Prevention and Community Health, Milken Institute School of Public Health, The George Washington University, Washington, District of Columbia, United States of America.
- Department of Neurology, Taipei Medical University-Shuang Ho Hospital, New Taipei City, Taiwan.
- Department of Neurology, School of Medicine, College of Medicine, Taipei Medical University, Taipei City, Taiwan.
- Taipei Neuroscience Institute, Taipei Medical University-Shuang Ho Hospital, New Taipei City, Taiwan.
ORCID-Profile: Oluwaseun Adebayo Bamodu, Sumaiya Nezam, Chen-Chih Chung
Zitieren
APA
Bamodu, O. A., Nezam, S., & Chung, C. C. (2026). Explainable machine learning for breast cancer prediction in resource-constrained settings: A multi-algorithmic framework integrating shap-based transparency with clinical decision support. PLOS Digital Health, 5(9), e0001706. https://doi.org/10.1371/journal.pdig.0001706
Vancouver
Bamodu OA, Nezam S, Chung CC. Explainable machine learning for breast cancer prediction in resource-constrained settings: A multi-algorithmic framework integrating shap-based transparency with clinical decision support. PLOS Digital Health. 2026;5(9):e0001706. doi:10.1371/journal.pdig.0001706
BibTeX
@article{Bamodu2026,
author = {Bamodu, Oluwaseun Adebayo and Nezam, Sumaiya and Chung, Chen-Chih},
title = {Explainable machine learning for breast cancer prediction in resource-constrained settings: A multi-algorithmic framework integrating shap-based transparency with clinical decision support},
journal = {PLOS Digital Health},
year = {2026},
volume = {5},
number = {9},
pages = {e0001706},
doi = {10.1371/journal.pdig.0001706}
}
EUVIMED