Artificial Intelligence–Driven Monitoring of Long-Term Rehabilitation Outcomes in Vascular Patients Undergoing Endovascular Repair

Abstract

Endovascular repair has become the chosen treatment for complicated vascular conditions because it is less invasive and takes less time to heal. But success after surgery rests on more than just how well the surgery went. It also depends on how well the patient recovers over time. Regular checks and subjective reports are common ways of tracking that don't always pick up on small signs of recovery or early signs of stagnation. To get around this problem, this study suggests using artificial intelligence to keep an eye on and guess how vascular patients will do in their recovery after endovascular repair. The framework combines clinical data, physiological signals from sensors, and patient-reported results to create a full recovery profile. It was possible to get adaptive, patient-specific insights by creating a CNN–LSTM model that can both extract spatial features and learn temporal recovery trends. This study shows that AI has the ability to make vascular rehabilitation a continuous, data-driven, and patient-centered process.