Related Experiment Video
Updated: Oct 19, 2025
Improved Registration of 3D CT Angiography with X-ray Fluoroscopy for Image Fusion During Transcatheter Aortic Valve Implantation
Published on: June 3, 2018
Deep learning-based prediction of early cerebrovascular events after transcatheter aortic valve replacement
Taishi Okuno1, Pavel Overtchouk2,3, Masahiko Asami1
1Department of Cardiology, Inselspital, Bern University Hospital, University of Bern, Bern, Switzerland.
Insights
A new deep learning tool can predict cerebrovascular events (CVE) after transcatheter aortic valve replacement (TAVR). This advancement aids in identifying high-risk patients, potentially improving outcomes for TAVR procedures.
Area of Science:
- Cardiovascular Medicine
- Medical Imaging
- Artificial Intelligence in Healthcare
Background:
- Cerebrovascular events (CVE) are a significant feared complication following transcatheter aortic valve replacement (TAVR).
- Current clinical predictors inadequately explain the multifactorial origins of TAVR-related CVE, highlighting a need for improved predictive capabilities.
Purpose of the Study:
- To develop a deep learning-based predictive tool for identifying patients at risk of cerebrovascular events (CVE) post-TAVR.
- To integrate both clinical and advanced imaging characteristics for enhanced CVE prediction accuracy.
Main Methods:
- A prospective TAVR registry was utilized, analyzing integrated clinical and imaging data from 1492 patients.
- Recursive feature reduction was employed to select predictive variables for training an autoencoder model.
- Model performance was evaluated using the area under the curve (AUC) for predicting 30-day CVE.
Main Results:
- The study included 1492 patients with a median age of 83 years and an STS score of 4.6%.
- Acute (1.3%) and subacute (2.4%) CVE occurred in 19 and 36 patients, respectively, with CVE associated with increased mortality risk (HR 2.62).
- The developed deep learning model, utilizing under 107 variables, achieved an AUC of 0.79 (0.65-0.93) for predicting 30-day CVE.
Conclusions:
- A deep learning-based predictive algorithm can effectively predict TAVR-related cerebrovascular events.
- The developed model demonstrates promising performance and is available online for broader clinical application and research.
Abstract:
Cerebrovascular events (CVE) are among the most feared complications of transcatheter aortic valve replacement (TAVR). CVE appear difficult to predict due to their multifactorial origin incompletely explained by clinical predictors. We aimed to build a deep learning-based predictive tool for TAVR-related CVE. Integrated clinical and imaging characteristics from consecutive patients enrolled into a prospective TAVR registry were analysed. CVE comprised any strokes and transient ischemic attacks. Predictive variables were selected by recursive feature reduction to train an autoencoder predictive model. Area under the curve (AUC) represented the model's performance to predict 30-day CVE. Among 2279 patients included between 2007 and 2019, both clinical and imaging data were available in 1492 patients. Median age was 83 years and STS score was 4.6%. Acute (< 24 h) and subacute (day 2-30) CVE occurred in 19 (1.3%) and 36 (2.4%) patients, respectively. The occurrence of CVE was associated with an increased risk of death (HR [95% CI] 2.62 [1.82-3.78]). The constructed predictive model uses less than 107 clinical and imaging variables and has an AUC of 0.79 (0.65-0.93). TAVR-related CVE can be predicted using a deep learning-based predictive algorithm. The model is implemented online for broad usage.
More Related Videos
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
09:57Four-Dimensional Computed Tomography-Guided Valve Sizing for Transcatheter Pulmonary Valve Replacement
Published on: January 20, 2022