Ultrasound Segmentation Task

Ultrasound segmentation aims to automatically identify and delineate specific anatomical structures within ultrasound images, improving diagnostic accuracy and efficiency. Current research emphasizes developing robust and fair models, often employing attention mechanisms and autoencoder architectures, to address challenges posed by image complexity and variability across patient subgroups. This work is crucial for advancing automated medical diagnosis, particularly in echocardiography, by reducing reliance on human annotation and improving the consistency and speed of analysis. Furthermore, research actively addresses issues of model bias and incorporates statistical validation to ensure reliable and ethical deployment of these algorithms.

Papers

September 11, 2024

EchoDFKD: Data-Free Knowledge Distillation for Cardiac Ultrasound Segmentation using Synthetic Data
Grégoire Petit, Nathan Palluau, Axel Bauer, Clemens Dlaska
Synthetic Data Human Annotation Data Free Knowledge Distillation Ejection Fraction Ultrasound Video Human Annotated Label Ultrasound Segmentation Task

December 5, 2023

Inspecting Model Fairness in Ultrasound Segmentation Tasks
Zikang Xu, Fenghe Tang, Quan Quan, Jianrui Ding, Chunping Ning, S. Kevin Zhou
Deep Learning Model Fairness Fair Model Ultrasound Dataset Ultrasound Segmentation Task

October 2, 2023

BAAF: A Benchmark Attention Adaptive Framework for Medical Ultrasound Image Segmentation Tasks
Gongping Chen, Lei Zhao, Xiaotao Yin, Liang Cui, Jianxun Zhang, Yu Dai
Ultrasound Image Parallel Attention Ultrasound Segmentation Task

August 2, 2023

Decomposing and Coupling Saliency Map for Lesion Segmentation in Ultrasound Images
Zhenyuan Ning, Yixiao Mao, Qianjin Feng, Shengzhou Zhong, Yu Zhang
Ultrasound Image Human Saliency Lesion Segmentation Ultrasound Segmentation Task

April 3, 2023

Efficient human-in-loop deep learning model training with iterative refinement and statistical result validation
Manuel Zahn, Douglas P. Perrin
Deep Learning Training Data Human in the Loop Iterative Refinement Experimental Validation Training Deep Quantitative Segmentation Ultrasound Segmentation Task

February 27, 2023

EDMAE: An Efficient Decoupled Masked Autoencoder for Standard View Identification in Pediatric Echocardiography
Yiman Liu, Xiaoxiang Han, Tongtong Liang, Bin Dong, Jiajun Yuan, Menghan Hu, Qiaohong Liu, Jiangang Chen, Qingli Li, Yuqi Zhang
High Efficiency Masked AutoEncoder Student Encoder Ultrasound Segmentation Task

Ultrasound Segmentation Task

Papers

EchoDFKD: Data-Free Knowledge Distillation for Cardiac Ultrasound Segmentation using Synthetic Data

Inspecting Model Fairness in Ultrasound Segmentation Tasks

BAAF: A Benchmark Attention Adaptive Framework for Medical Ultrasound Image Segmentation Tasks

Decomposing and Coupling Saliency Map for Lesion Segmentation in Ultrasound Images

Efficient human-in-loop deep learning model training with iterative refinement and statistical result validation

EDMAE: An Efficient Decoupled Masked Autoencoder for Standard View Identification in Pediatric Echocardiography