Machine Learning Prediction of Treatment Failure After Middle Meningeal Artery Embolization for Subdural Hematoma
Abstract
BACKGROUND AND OBJECTIVES: Middle meningeal artery embolization (MMAE) is effective for chronic subdural hematoma (SDH), but failure rates vary and risk prediction tools are lacking. This study developed and validated a machine learning (ML) model to predict MMAE treatment failure for SDH. METHODS: We conducted an international, multi-institutional retrospective cohort study of patients who underwent MMAE for SDH between 2019 and 2025. The cohort was split into training and test sets, stratified by treatment failure, with all hematomas from each patient assigned to a single cohort to prevent data leakage. Multiple ML algorithms were evaluated using 5-fold cross-validation, with gradient boosting selected as the final model. The primary outcome was treatment failure, defined as rescue surgical evacuation or persistent/recurrent SDH requiring retreatment. Model performance was assessed using area under the curve-receiver operating characteristic. Patients were stratified into low-risk (<5%), medium-risk (5%-15%), and high-risk (>15%) groups using a priori clinically defined thresholds. RESULTS: Among 692 treated hematomas in the test cohort, 64 (9.2%) experienced treatment failure. The gradient boosting model achieved an area under the curve-receiver operating characteristic of 0.856 (95% CI, 0.803-0.902) before calibration and 0.832 (95% CI, 0.766-0.889) after isotonic calibration, with excellent calibration (Brier score, 0.053). The most important predictors were age-platelet interaction, platelet-hemoglobin ratio, platelet count, SDH thickness, and age. Observed treatment failure was 1.6% (95% CI, 0.4%-5.7%) in the low-risk group (n = 124; 17.9%), 6.0% (95% CI, 4.3%-8.4%) in the medium-risk group (n = 516; 74.6%), and 59.6% (95% CI, 46.1%-71.8%) in the high-risk group (n = 52; 7.5%) ( P < .001 for trend), a 37.2-fold difference between high- and low-risk groups. CONCLUSION: This ML model stratified MMAE patients into risk groups with a 37.2-fold difference in treatment failure. High-risk patients had 59.6% failure, whereas low-risk patients had 1.6% failure. Risk stratification may improve patient selection and monitoring.
// Source
Authors: Bluyé DeMessie, Muhammed Amir Essibayi, Hamza Adel Salim, Alireza Karandish, Ahmed Y. Azzam, Deepak Khatri, Neil Haranhalli, Amanda Baker, Richard L. Zampolin, Allan Brook, Seon-Kyu Lee, Nimer Adeeb, Dhairya A. Lakhani, James T. Grist, Fulvio Zaccagna, Yan-Lin Li, Davide Simonato, Diego Alejandro Ortega, Nicole Mariantonia Cancelliere, Jose Danilo Bengzon Diestro, Joseph A. Carnevale, Craig Schreiber, Atakan Orscelik, Zachary A. Abecassis, Spencer Raub, Georgios S. Sioutas, Abdulrhman Alsalama, Clairice Pearce, Giancarlo Salsano, Rahim Abo Kasem, Svetlana Kvint, Andrew Falzon, Vance Cantrell, Brian Holliday, Jefferson O. Abaricia, Pedram Daniel Maleknia, Leonardo Cruz‐Criollo, Samantha J. Schimmel, Basel Musmar, Matthew David Alexander, Jorge Rios Zermeno, Prateeka Koul, Ahmed Aljuboori, Dominik F. Vollherbst, Domagoj Gajski, Jared Cooper, Omar Alwakaa, Mohamad Ezzeldin, Christopher S. Ogilvy, Fawaz Al‐Mufti, Vladimir Kalousek, Markus Alfred Möhlenbruch, Luca Scarcia, William Walker Wroe, Hussein A. Zeineddine, Tang Si Zhao, Peter Bernhard Sporns, Anil Gopinathan, Robert William Regenhardt, Spiros Blackburn, Clemens Maria Schirmer, Thien J Huynh, Rabih G. Tawk, Fabio Settecase, Stavropoula I. Tjoumakaris, Pascal M. Jabbour, Kunal V. Vakharia, Mario Zanaty, Santiago Ortega‐Gutiérrez, Jesse George Atherton Jones, Marco Colasurdo, Hussein H. Nasser, Sri Hari Sundararajan, Pascal J. Mosimann, Erez Nossek, Eytan Raz, Sami Al Kasab, Alejandro M. Spiotta, Lucio Castellan, Bruno Del Sette, A Alaraj, Caterina Michelozzi, Davide Saraceno, Pietro Panni, Visish Mani Srinivasan, Jan‐Karl Burkhardt, Gaultier Marnat, Pietro Mario Santini, Michael R. Levitt, Giuseppe S Lanzino, Waleed Brinjikji, Jared Knopman, Tareq Kass‐Hout, Julian Spears, Thomas R. Marotta, Vítor Mendes Pereira, Maurizio Fuschi, Adam Andrew Dmytriw, David Altschul, on behalf of the MESH Collaborators
Institutions: Brigham and Women's Hospital, University of Chicago, University of Washington, Harvard University, Montefiore Medical Center, University of Alabama at Birmingham, University Hospital of Basel, Nuffield Orthopaedic Centre, University of Oxford, University of South Florida, Albert Einstein College of Medicine, Oregon Health & Science University, Cornell University, University of Iowa, Heidelberg University, University Hospital Heidelberg, National University of Singapore, National University Health System, The University of Texas MD Anderson Cancer Center, University of Illinois Chicago, NYU Langone Health, New York University, Université de Bordeaux, Mayo Clinic in Arizona, St. Michael's Hospital, Toronto Western Hospital, Beth Israel Deaconess Medical Center, West Virginia University, Mayo Clinic, Medical University of South Carolina, Louisiana State University in Shreveport, Presbyterian Hospital, Ospedale Policlinico San Martino, New York Medical College, Westchester Medical Center, California Pacific Medical Center, The University of Texas Health Science Center at Houston, Mayo Clinic in Florida, Thomas Jefferson University, San Raffaele University of Rome, Jacksonville College, Centre Hospitalier Universitaire Henri-Mondor, Hospital of the University of Pennsylvania, National University Hospital, Bordeaux Population Health, University of Iowa Health Care, St George’s University Hospitals NHS Foundation Trust, Sisters of Charity Hospital, Hôpitaux Universitaires Henri-Mondor, University of Iowa Hospitals and Clinics, Lehigh Valley Hospital-Pocono, Lehigh Valley Health Network, Lone Star College Kingwood, Stadtspital Waid, Sutter Medical Center