Abstract:Background: Chronic subdural hematoma (CSDH) represents a prevalent medical condition, posing substantial challenges in postoperative management due to risks of recurrence. Such recurrences not only cause physical suffering to the patient but also add to the financial burden on the family and the health care system. Currently, prognosis determination largely depends on clinician expertise, revealing a dearth of precise prediction models in clinical settings. Objective: This study aims to use machine learning (ML) techniques for the construction of predictive models to assess the likelihood of CSDH recurrence after surgery, which leads to greater benefits for patients and the health care system. Methods: Data from 133 patients were amassed and partitioned into a training set (n=93) and a test set (n=40). Radiomics features were extracted from preoperative cranial computed tomography scans using 3D Slicer software. These features, in conjunction with clinical data and composite clinical-radiomics features, served as input variables for model development. Four distinct ML algorithms were used to build predictive models, and their performance was rigorously evaluated via accuracy, area under the curve (AUC), and recall metrics. The optimal model was identified, followed by recursive feature elimination for feature selection, leading to enhanced predictive efficacy. External validation was conducted using data sets from additional health care facilities. Results: Following rigorous experimental analysis, the support vector machine model, predicated on clinical-radiomics features, emerged as the most efficacious for predicting postoperative recurrence in patients with CSDH. Subsequent to feature selection, key variables exerting significant impact on the model were incorporated as the input set, thereby augmenting its predictive accuracy. The model demonstrated robust performance, with metrics including accuracy of 92.72%, AUC of 91.34%, and recall of 93.16%. External validation further substantiated its effectiveness, yielding an accuracy of 90.32%, AUC of 91.32%, and recall of 88.37%, affirming its clinical applicability. Conclusions: This study substantiates the feasibility and clinical relevance of an ML-based predictive model, using clinical-radiomics features, for relatively accurate prognostication of postoperative recurrence in patients with CSDH. If the model is integrated into clinical practice, it will be of great significance in enhancing the quality and efficiency of clinical decision-making processes, which can improve the accuracy of diagnosis and treatment, reduce unnecessary tests and surgeries, and reduce the waste of medical resources.

Deep learning model utilizing clinical data alone outperforms image-based model for hernia recurrence following abdominal wall reconstruction with long-term follow up

Image-based Deep Learning Model to Predict Stoma-Site Incisional Hernia in Patients with Temporary Ileostomy: A Retrospective Study

Deep Learning for Automated Detection and Localization of Traumatic Abdominal Solid Organ Injuries on CT Scans

Deep learning models of ultrasonography significantly improved the differential diagnosis performance for superficial soft-tissue masses: a retrospective multicenter study

Deep Learning on Enhanced CT Images Can Predict the Muscular Invasiveness of Bladder Cancer

Development of a Dual‐Plane MRI‐Based Deep Learning Model to Assess the 1‐Year Postoperative Outcomes in Lumbar Disc Herniation After Tubular Microdiscectomy

Deep learning model combined with computed tomography features to preoperatively predicting the risk stratification of gastrointestinal stromal tumors

Deep Learning Reconstruction for Coronary CT Angiography in Patients with Origin Anomaly, Stent or Bypass Graft.

Current Status of Deep Learning in Abdominal Image Reconstruction

Preoperative CT-Based Deep Learning Model for Predicting Risk Stratification in Patients With Gastrointestinal Stromal Tumors

Combining Clinical-Radiomics Features With Machine Learning Methods for Building Models to Predict Postoperative Recurrence in Patients With Chronic Subdural Hematoma: Retrospective Cohort Study

Development of a predictive model for 1-year postoperative recovery in patients with lumbar disk herniation based on deep learning and machine learning

Deep Learning Radiomics Model of Dynamic Contrast-Enhanced MRI for Evaluating Vessels Encapsulating Tumor Clusters and Prognosis in Hepatocellular Carcinoma

Development and Comparison of Multimodal Models for Preoperative Prediction of Outcomes After Endovascular Aneurysm Repair

A new model using deep learning to predict recurrence after surgical resection of lung adenocarcinoma

A CT-based deep learning model predicts overall survival in patients with muscle invasive bladder cancer after radical cystectomy: a multicenter retrospective cohort study

Magnetic resonance imaging-based deep learning model to predict multiple firings in double-stapled colorectal anastomosis

Using deep learning to predict cardiovascular magnetic resonance findings from echocardiography videos

Deep learning‐based detection and classification of lumbar disc herniation on magnetic resonance images

Combining a deep learning model with clinical data better predicts hepatocellular carcinoma behavior following surgery

A Deep Learning Image Reconstruction Algorithm for Improving Image Quality and Hepatic Lesion Detectability in Abdominal Dual-Energy Computed Tomography: Preliminary Results