Neural Network QRS Duration Estimator for EP Studies Enhanced by Foundation Model Morphological Context

Yan B Werneck1, Thaís de Jesus Soares2, Lucas Arantes Berg2, Tiago Dutra Franco3, Fabrício Santos3, Raul Pereira Barra3, Bernardo Martins Rocha3, Joventino de Oliveira Campos2, Rafael Sachetto Oliveira4, Marcelle Cristina da Silva Bastos Vasconcelos5, Thiago Goncalves Schroder e Souza6, THAIZ RUBERTI SCHMAL7, Rodrigo Weber dos Santos2
1PPGMC-UFJF, 2Federal University of Juiz de Fora, 3Universidade Federal de Juiz de Fora, 4Univerisidade Federal de S�o Jo�o del-Rei, 5Department of Internal Medicine, Federal University of Juiz de Fora (UFJF), Juiz de Fora, Brazil, 6University Hospital of the UFJF, 7EBSERH University Hospital Juiz de Fora


Abstract

Electrophysiological studies with programmed ventricular stimulation require QRS duration measurement across sinus and stimulus beats acquired at varying coupling intervals, a regime whose paced and aberrantly conducted morphologies existing detectors were not designed for and where expert annotation is scarce. We propose a neural delineator trained on 715 annotated beats from four non-ischemic dilated cardiomyopathy patients. Each lead is scored independently by a shared dilated convolutional backbone into onset and offset attention maps; a frozen HuBERT-ECG encoder supplies morphology-aware weights that pool the twelve maps, and soft-argmax yields the boundaries. A per-boundary uncertainty width is predicted alongside, and the reported predictor is the per-beat median over eight independently seeded models. Under leave-one-patient-out cross-validation, the model attains a QRS duration MAE of 19.1 ± 1.6 ms for unseen patients, comparable to expert inter-observer variability. The residual error is dominated by a per-patient offset shift, and the predicted uncertainty correctly orders both beats and patients by difficulty on onset but not on offset.