Benchmarking Self-Supervised Models for Cardiac Ultrasound View Classification

Youssef Megahed; Salma I. Megahed; Robin Ducharme; Inok Lee; Adrian D. C. Chan; Mark C. Walker; Steven Hawken

eess.IVFeb 17, 2026

Benchmarking Self-Supervised Models for Cardiac Ultrasound View Classification

Authors: Youssef Megahed, Salma I. Megahed, Robin Ducharme, Inok Lee, Adrian D. C. Chan, Mark C. Walker, Steven Hawken

Organizations: Department of Systems and Computer Engineering, Carleton University, Ottawa, Ontario, Canada · Department of Methodological and Implementation Research, Ottawa Hospital Research Institute, Ottawa, Ontario, Canada · Department of Acute Care Research, Ottawa Hospital Research Institute, Ottawa, Ontario, Canada · Department of Obstetrics and Gynecology, University of Ottawa, Ottawa, Ontario, Canada · School of Epidemiology and Public Health, University of Ottawa, Ottawa, Ontario, Canada · Department of Obstetrics, Gynecology & Newborn Care, The Ottawa Hospital, Ottawa, Ontario, Canada · International and Global Health Office, University of Ottawa, Ottawa, Ontario, Canada

Abstract

Reliable interpretation of cardiac ultrasound images is essential for accurate clinical diagnosis and assessment. Self-supervised learning has shown promise in medical imaging by leveraging large unlabelled datasets to learn meaningful representations. In this study, we evaluate and compare two self-supervised learning frameworks, USF-MAE, developed by our team, and MoCo v3, on the recently introduced CACTUS dataset (37,736 images) for automated simulated cardiac view (A4C, PL, PSAV, PSMV, Random, and SC) classification. Both models used 5-fold cross-validation, enabling robust assessment of generalization performance across multiple random splits. The CACTUS dataset provides expert-annotated cardiac ultrasound images with diverse views. We adopt an identical training protocol for both models to ensure a fair comparison. Both models are configured with a learning rate of 0.0001 and a weight decay of 0.01. For each fold, we record performance metrics including ROC-AUC, accuracy, F1-score, and recall. Our results indicate that USF-MAE consistently outperforms MoCo v3 across metrics. The average testing AUC for USF-MAE is 99.99% (+/-0.01% 95% CI), compared to 99.97% (+/-0.01%) for MoCo v3. USF-MAE achieves a mean testing accuracy of 99.33% (+/-0.18%), higher than the 98.99% (+/-0.28%) reported for MoCo v3. Similar trends are observed for the F1-score and recall, with improvements statistically significant across folds (paired t-test, p=0.0048 < 0.01). This proof-of-concept analysis suggests that USF-MAE learns more discriminative features for cardiac view classification than MoCo v3 when applied to this dataset. The enhanced performance across multiple metrics highlights the potential of USF-MAE for improving automated cardiac ultrasound classification.

Figures & tables

Category	Number of Images
Total Images	37,736
A4C	7,422
PL	6,102
PSAV	5,832
PSMV	6,014
Random	6,021

Table 1: Dataset distribution of cardiac view classes and data split used in stratified 5-fold cross-validation.

Figure 1: A) Original ultrasound frame with overlay annotations. B) Cropped and cleaned, processed and inpainted image used for analysis.

Component	MoCo v3 [ 12 ]	USF-MAE (ours) [ 2 ]
Backbone architecture	ViT-B/16	ViT-B/16
Pretraining paradigm	Contrastive SSL	Masked autoencoding (MAE)
Pretraining dataset	ImageNet-1K	OpenUS-46
Dataset domain	Natural images	Ultrasound images
Number of images	$\sim$ 1.28M	$\sim$ 370K
Pretraining objective	Instance discrimination via momentum contrast	Patch reconstruction (MSE over masked patches)

Table 2: Comparison of pre-training configurations for MoCo v3 and USF-MAE.

Figure 2: Comparison of MoCo v3 and USF-MAE Pipelines.

Fold	Accuracy (%)		Recall (%)		F1-score (%)		AUC (%)
Fold	MoCo v3	USF-MAE	MoCo v3	USF-MAE	MoCo v3	USF-MAE	MoCo v3	USF-MAE
1	99.18	99.35	99.18	99.35	99.18	99.35	99.98	99.99
2	99.01	99.35	99.01	99.35	99.01	99.35	99.97	99.99
3	98.91	99.42	98.91	99.42	98.91	99.42	99.97	99.98
4	98.97	99.22	98.97	99.22	98.97	99.22	99.98	99.99
5	98.90	99.31	98.90	99.31	98.90	99.31	99.98	99.98

Table 3: Per-fold performance across 5-fold cross-validation.

Figure 3: Classification performance for cardiac view classification: A) normalized confusion matrix and B) per-class ROC curves showing near perfect discrimination across all classes.

Explore similar work

May 19, 2026

Synergistic Foundation Models for Semi-Supervised Fetal Cardiac Ultrasound Analysis: SAM-Med2D Boundary Refinement and DINOv3 Semantic Enhancement

We present a semi-supervised framework for joint segmentation and classification of fetal cardiac ultrasound images. Built upon the EchoCare multi-task backbone, our method integrates SAM-Med2D for boundary refinement and leverages DINOv3 to enhance pseudo-label quality. We introduce view-specific hard masking along with a two-stage optimization strategy: an EMA phase to consolidate segmentation capabilities, followed by a Classification Fine-Tuning phase that freezes segmentation parameters and resets the classification head to recover classification performance without compromising segmentation gains. Evaluated on the FETUS 2026 leaderboard, our method achieves a Dice Similarity Coefficient at 79.99%, Normalized Surface Distance at 61.62%, and F1-score at 41.20%, validating the effectiveness of our approach for prenatal congenital heart disease screening. Source code is publicly available at: https://github.com/2826056177/zcst_fetus2026.

Tonghao Zhuang, Shanglong Hu, Yongsheng Luo +2

Zhuhai College of Science and Technology, Zhuhai, China

Ultrasound Image Segmentation Medical Imaging Foundation Models Medical Image Classification Ultrasound Imaging Semi-Supervised Learning

Ask PDF

May 25, 2026

Anatomy-Anchored Self-Supervision: Distilling Vision Foundation Models for Invariant Ultrasound Representation

Self-supervised pre-training paradigm has gained increasing prominence for learning transferable representations in medical imaging, yet existing methods for ultrasound (US) images operate at the image or frame level, overlooking the anatomical context for clinical-aligned representation learning. In this work, we propose an anatomy-anchored ultrasound self-supervision framework ANAUS that shifts representation learning from generic visual regions to clinically meaningful anatomical structures. Utilizing a learnable latent prompt engine alongside a one-time domain adaptation on existing public image-mask pairs, we empower the LP-SAM module to achieve annotation-free anatomy delineation at scale. Building upon this anatomical grounding, we propose a dual-policy self-supervised learning paradigm consisting of inter-view semantics-aware anatomy-separating alignment and contextual core-region prediction to enhance representation learning. Specifically, the former enforces feature invariance within identical anatomical regions while promoting discriminability across distinct structures; the latter compels the model to reconstruct corrupted regions, thereby capturing fine-grained structural details. Extensive evaluations on six public datasets demonstrate that ANAUS consistently outstrips current state-of-the-art methods while maintaining the computational efficiency essential for clinical deployment. Code is available at https://github.com/zhcz328/ANAUS.

Chunzheng Zhu, Yijun Wang, Jianxin Lin +5

Hunan University, Changsha, China · Shenzhen Maternity and Child Healthcare Hospital, Shenzhen, China

Invariant Representation Learning Medical Image Representation Learning Self-Supervised Visual Representation Learning

Ask PDF

Sep 15, 2026

Which Pretext Task Transfers? Self-Supervised Pretraining Objectives for Lung Ultrasound

Self-supervised learning (SSL) can reduce the need for labelled medical images, but the choice of pretext objective remains unclear for lung ultrasound (LUS). Contrastive learning, masked reconstruction, and joint-embedding predictive architectures (JEPA) differ in the space in which their targets are defined, yet existing ultrasound studies compare them under different corpora, backbones, and evaluation protocols. We compare these three objective families using the same encoder backbone, pretraining corpus, optimisation schedule, and frozen-evaluation protocol. Encoders are pretrained on COVID-BLUeS LUS videos and evaluated with linear,

k

NN, and attentive probes at 5%, 10%, 50%, and 100% label budgets. Evaluation is performed on POCUS using patient-level five-fold cross-validation and on the independently acquired Mendeley-Uganda dataset, which is excluded from both pretraining and probe fitting. At the full label budget under linear probing, VideoMAE and V-JEPA achieve

66.5 \pm 13.1

and

65.4 \pm 11.7

balanced accuracy on POCUS, while MoCo achieves

42.1 \pm 1.2

. On Mendeley-Uganda, the ranking reverses: MoCo performs best at

62.7 \pm 1.0

, followed by VideoMAE at

53.8 \pm 2.8

, while V-JEPA falls near chance at

35.1 \pm 4.9

. These results show that POCUS probe accuracy alone does not identify the objective that transfers best across datasets. We also outline planned representation-level analyses to examine this reversal. Code is publicly available at https://github.com/moeinheidari7829/LUSVideoSSL.

Moein Heidari, Junbo Rao, Jai Choraria +3

University of British Columbia, Vancouver, BC, Canada · Rutgers Cancer Institute of New Jersey, New Brunswick, NJ, USA

Contrastive Learning Masked Autoencoders Self-Supervised Pre-Training Joint-Embedding Predictive Architecture Ultrasound Imaging Medical Image Representation Learning

Ask PDF