Unveiling the Hidden Threat: Digital Twins and AI for Detecting Undiagnosed Preoperative Atrial Fibrillation in Cardiac Surgery

Status: In Publication / Submitted

Authors: Zain Khalpey, Ujjawal Kumar, Nicholas King, Amina H. Khalpey

Abstract

Undiagnosed preoperative atrial fibrillation (AFib) represents a significant and often overlooked risk factor in cardiac surgery patients. Paroxysmal AFib, in particular, may go undetected during standard preoperative assessment, leaving patients and surgical teams unaware of elevated risk profiles that could influence surgical planning and postoperative management.

This work presents a novel approach combining digital twin technology with AI-driven waveform analysis to detect subclinical atrial fibrillation patterns in preoperative cardiac surgery candidates. By creating patient-specific digital cardiac models and applying deep learning to continuous monitoring data, the system identifies subtle electrophysiological signatures indicative of intermittent AFib.

The proposed methodology demonstrates superior sensitivity compared to conventional screening approaches, with the potential to significantly reduce adverse events related to undiagnosed arrhythmias in the perioperative period.