Short-ECG Cardiac Amyloidosis Detection: CNN-LSTM vs. Foundation Models

Alejandro Tostado1, Xabier Arana Achaga2, Cristina Goena Vives2, Iñaki Villanueva Benito2, Ane Alberdi Aramendi3, Maitane Barrenetxea Carrasco4, Unai Ayala Fernandez4
1Tecnológico de Monterrey, México. Faculty of Engineering, Mondragon Unibertsitatea, España, 2Heart Failure and Inherited Cardiac Diseases Unit, Department of Cardiology, Donostia University Hospital, Donostia, Spain Biogipuzkoa Health Research Institute, Donostia, Spain, 3Mondragon Unibertsitatea, 4Faculty of Engineering, Mondragon Unibertsitatea, Arrasate 20500, Spain


Abstract

Cardiac amyloidosis remains underdiagnosed, partly because its electrocardiographic signatures are not reliably detected during routine clinical interpretation. While deep learning methods have enabled automated detection from ECG signals, most high-performing models depend on large labeled datasets and long signal durations, limiting their use in data-constrained environments with short-duration recordings.

We evaluated pre-trained transformer models and convolutional architectures trained from scratch on a balanced dataset of 461patients with 2.5-second twelve lead ECG recordings, using a 5-fold cross-validation strategy. Pre-trained transformer representations from HuBERT-ECG and ECG-JEPA were assessed under a linear probing setup without fine-tuning, while two CNNs (small and medium) and a hybrid CNN-LSTM were trained from scratch using identical validation splits and consistent preprocessing conditions.

The CNN-LSTM achieved the highest performance, yielding a cross-validated AUC of 0.832 (± 0.019). ECG-JEPA reached a comparable AUC of 0.827 (± 0.060), whereas HuBERT-ECG underperformed with a maximum AUC of 0.735 (± 0.060).

These results showed that, in settings with limited data and short signal durations, a hybrid CNN-LSTM architecture that jointly captures morphological and temporal structure retains a measurable advantage over frozen foundation models. However, the competitive performance of ECG-JEPA suggests that fine-tuning or the use of longer recordings may reduce this gap, supporting further exploration of self-supervised transformer models for cardiac amyloidosis detection under limited data conditions.