Temporal Representation Learning for Real-Time Ultrasound Analysis

• URL: http://arxiv.org/abs/2509.01433v1
• Date: Mon, 01 Sep 2025 12:43:38 GMT
• Title: Temporal Representation Learning for Real-Time Ultrasound Analysis
• Authors: Yves Stebler, Thomas M. Sutter, Ece Ozkan, Julia E. Vogt,
• Abstract summary: temporal dynamics is essential for assessing motion patterns in applications such as cardiac monitoring, fetal development, and vascular imaging.<n>Current deep learning models often overlook the temporal continuity of ultrasound sequences.<n>We propose a method for learning effective temporal representations from ultrasound videos.
• Score: 16.047239074927234
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Ultrasound (US) imaging is a critical tool in medical diagnostics, offering real-time visualization of physiological processes. One of its major advantages is its ability to capture temporal dynamics, which is essential for assessing motion patterns in applications such as cardiac monitoring, fetal development, and vascular imaging. Despite its importance, current deep learning models often overlook the temporal continuity of ultrasound sequences, analyzing frames independently and missing key temporal dependencies. To address this gap, we propose a method for learning effective temporal representations from ultrasound videos, with a focus on echocardiography-based ejection fraction (EF) estimation. EF prediction serves as an ideal case study to demonstrate the necessity of temporal learning, as it requires capturing the rhythmic contraction and relaxation of the heart. Our approach leverages temporally consistent masking and contrastive learning to enforce temporal coherence across video frames, enhancing the model's ability to represent motion patterns. Evaluated on the EchoNet-Dynamic dataset, our method achieves a substantial improvement in EF prediction accuracy, highlighting the importance of temporally-aware representation learning for real-time ultrasound analysis.

Related papers

• Improving segmentation of retinal arteries and veins using cardiac signal in doppler holograms [2.1194651766074424]
  We propose a simple yet effective approach for artery-vein segmentation in temporal Doppler holograms.<n>By incorporating features derived from a dedicated pulse analysis pipeline, our method allows conventional U-Nets to exploit temporal dynamics.<n>These findings demonstrate that time-resolved preprocessing can unlock the full potential of deep learning for Doppler holography.
  arXiv  Detail & Related papers  (2025-11-18T16:49:20Z)
• VA-Adapter: Adapting Ultrasound Foundation Model to Echocardiography Probe Guidance [57.43511837589102]
  We adapt medical knowledge learned by foundation models from vast datasets to the probe guidance task.<n>We meticulously design a parameter-efficient Vision-Action Adapter (VA-Adapter) to enable foundation model's image encoder to encode vision-action sequences.<n>With built-in sequential reasoning capabilities in a compact design, the VA-Adapter enables a pre-trained ultrasound foundation model to learn precise probe adjustment strategies.
  arXiv  Detail & Related papers  (2025-10-08T09:38:30Z)
• MOOSE: Pay Attention to Temporal Dynamics for Video Understanding via Optical Flows [21.969862773424314]
  MOOSE is a novel temporally-centric video encoder that integrates optical flow with spatial embeddings to model temporal information efficiently.<n>Unlike prior models, MOOSE takes advantage of rich, widely available pre-trained visual and optical flow encoders instead of training video models from scratch.
  arXiv  Detail & Related papers  (2025-06-01T18:53:27Z)
• EchoWorld: Learning Motion-Aware World Models for Echocardiography Probe Guidance [79.66329903007869]
  We present EchoWorld, a motion-aware world modeling framework for probe guidance.<n>It encodes anatomical knowledge and motion-induced visual dynamics.<n>It is trained on more than one million ultrasound images from over 200 routine scans.
  arXiv  Detail & Related papers  (2025-04-17T16:19:05Z)
• CTPD: Cross-Modal Temporal Pattern Discovery for Enhanced Multimodal Electronic Health Records Analysis [46.56667527672019]
  We introduce a Cross-Modal Temporal Pattern Discovery (CTPD) framework, designed to efficiently extract meaningful cross-modal temporal patterns from multimodal EHR data.<n>Our approach introduces shared initial temporal pattern representations which are refined using slot attention to generate temporal semantic embeddings.
  arXiv  Detail & Related papers  (2024-11-01T15:54:07Z)
• Self-supervised inter-intra period-aware ECG representation learning for detecting atrial fibrillation [41.82319894067087]
  We propose an inter-intra period-aware ECG representation learning approach. Considering ECGs of atrial fibrillation patients exhibit the irregularity in RR intervals and the absence of P-waves, we develop specific pre-training tasks for interperiod and intraperiod representations. Our approach demonstrates remarkable AUC performances on the BTCH dataset, textiti.e., 0.953/0.996 for paroxysmal/persistent atrial fibrillation detection.
  arXiv  Detail & Related papers  (2024-10-08T10:03:52Z)
• TEAM PILOT -- Learned Feasible Extendable Set of Dynamic MRI Acquisition Trajectories [2.7719338074999547]
  We introduce a novel deep-compressed sensing approach that uses 3D window attention and flexible, temporally extendable acquisition trajectories. Our method significantly reduces both training and inference times compared to existing approaches. Tests with real data show that our approach outperforms current state-of-theart techniques.
  arXiv  Detail & Related papers  (2024-09-19T13:45:13Z)
• Spatiotemporal Representation Learning for Short and Long Medical Image Time Series [8.344657414044153]
  Analyzing temporal developments is crucial for the accurate prognosis of many medical conditions. tracking longer term developments that occur over months or years is essential for accurate prognosis. Despite importance of both short and long term analysis to clinical decision making they remain understudied in medical deep learning.
  arXiv  Detail & Related papers  (2024-03-12T10:47:29Z)
• Deep Separable Spatiotemporal Learning for Fast Dynamic Cardiac MRI [22.7085949508143]
  MRI plays an indispensable role in cardiac diagnosis. To enable fast imaging, the k-space data can be undersampled. This challenge necessitates extensive training data in deep learning reconstruction methods. We propose a novel and efficient approach, leveraging a dimension separable learning scheme that can perform exceptionally well even with highly limited training data.
  arXiv  Detail & Related papers  (2024-02-24T23:56:15Z)
• Intelligent Robotic Sonographer: Mutual Information-based Disentangled Reward Learning from Few Demonstrations [42.731081399649916]
  This work proposes an intelligent robotic sonographer to autonomously "explore" target anatomies and navigate a US probe to a relevant 2D plane by learning from the expert. The underlying high-level physiological knowledge from experts is inferred by a neural reward function. The proposed advanced framework can robustly work on a variety of seen and unseen phantoms as well as in-vivo human carotid data.
  arXiv  Detail & Related papers  (2023-07-07T16:30:50Z)
• Motion Magnification in Robotic Sonography: Enabling Pulsation-Aware Artery Segmentation [44.868281669589194]
  In order to improve the artery segmentation accuracy and stability during scans, this work presents a novel pulsation-assisted segmentation neural network (PAS-NN) Motion magnification techniques are employed to amplify the subtle motion within the frequency band of interest to extract the pulsation signals from sequential US images. The extracted real-time pulsation information can help to locate the arteries on cross-section US images.
  arXiv  Detail & Related papers  (2023-07-07T16:14:17Z)
• Microvascular Dynamics from 4D Microscopy Using Temporal Segmentation [81.30750944868142]
  We are able to track changes in cerebral blood volume over time and identify spontaneous arterial dilations that propagate towards the pial surface. This new imaging capability is a promising step towards characterizing the hemodynamic response function upon which functional magnetic resonance imaging (fMRI) is based.
  arXiv  Detail & Related papers  (2020-01-14T22:55:03Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.