Ultrasound Localization Microscopy (ULM) is a non-invasive technique that allows for the imaging of micro-vessels in vivo, at depth and with a resolution on the order of ten microns. ULM is based on the sub-resolution localization of individual microbubbles injected in the bloodstream. Mapping the whole angioarchitecture requires the accumulation of microbubbles trajectories from thousands of frames, typically acquired over a few minutes. ULM acquisition times can be reduced by increasing the microbubble concentration, but requires more advanced algorithms to detect them individually. Several deep learning approaches have been proposed for this task, but they remain limited to 2D imaging, in part due to the associated large memory requirements. Herein, we propose to use sparse tensor neural networks to reduce memory usage in 2D and to improve the scaling of the memory requirement for the extension of deep learning architecture to 3D. We study several approaches to efficiently convert ultrasound data into a sparse format and study the impact of the associated loss of information. When applied in 2D, the sparse formulation reduces the memory requirements by a factor 2 at the cost of a small reduction of performance when compared against dense networks. In 3D, the proposed approach reduces memory requirements by two order of magnitude while largely outperforming conventional ULM in high concentration settings. We show that Sparse Tensor Neural Networks in 3D ULM allow for the same benefits as dense deep learning based method in 2D ULM i.e. the use of higher concentration in silico and reduced acquisition time.
超声定位显微镜（ULM）是一种新兴的技术，它采用回声微泡（MB）的定位，以精细的采样和成像微循环超过超声成像的衍射极限。传统的MB定位方法主要基于考虑MB的特定点扩散函数（PSF），这导致由重叠MB、非平稳PSF和谐波MB回波引起的信息丢失。因此，必须设计能够准确定位MB同时对MB非线性和使MB PSF失真的MB浓度变化有弹性的方法。本文提出了一种基于transformer的MB本地化方法来解决这个问题。我们采用了检测Transformer（DETR）arXiv：2005.12872，这是一种端到端的对象识别方法，使用基于集合的匈牙利损失和二分匹配为每个检测到的对象检测一个唯一的边界框。据作者所知，这是第一次将Transformers用于MB本地化。为了评估所提出的策略，预先训练的DETR网络的性能进行了测试，使用迁移学习原理检测MB。我们在IEEE IUS Ultra-SR挑战组织者提供的数据集随机选择的帧的子集上对网络进行了微调，然后使用交叉验证对其余帧进行了测试。对于模拟数据集，本文支持部署基于变压器的解决方案，以实现高精度的MB定位。
CycleULM: A unified label-free deep learning framework for ultrasound localisation microscopy
Department of Bioengineering, Faculty of Engineering, Imperial College London（生物工程系，工程学院，帝国理工学院伦敦分校）；Department of Electrical and Electronic Engineering, Faculty of Engineering, Imperial College London（电气与电子工程系，工程学院，帝国理工学院伦敦分校）；Faculty of Electrical and Computer Engineering, Escuela Superior Politécnica del Litoral (ESPOL)（电气与计算机工程学院，ESPOL高级理工大学）
Efficient Image Registration for Ultrasound Localization Microscopy by Obtaining Gradients via Integration Across Iterations
Imperial College London（帝国理工学院）；State Key Laboratory of Robotics and Systems, Harbin Institute of Technology（哈尔滨工业大学机器人技术与系统国家重点实验室）；The University of Melbourne（墨尔本大学）
Automatic Aberration Correction for Transcranial Functional and Super-Resolution Ultrasound Imaging in Rodents and Nonhuman Primates
Polytechnique Montréal（蒙特利尔高等工程学院）；Université de Montréal（蒙特利尔大学）；Centre interdisciplinaire de recherche sur le cerveau et l’apprentissage (CIRCA)（大脑与学习跨学科研究中心）；Montreal Heart Institute（蒙特利尔心脏研究所）
Inverse Problem Approach to Aberration Correction for in vivo Transcranial Imaging Based on a Sparse Representation of Contrast-enhanced Ultrasound Data
Vascular Segmentation of Functional Ultrasound Images using Deep Learning
AIstroSight；Inria（法国国家信息与自动化技术研究院）；Hospices Civils de Lyon（里昂慈善医院）；University Claude Bernard Lyon 1（里昂第一大学）；Lyon Neuroscience Research Center（里昂神经科学研究中心）
De-Noising of Photoacoustic Microscopy Images by Deep Learning
University of Michigan-Shanghai Jiao Tong University Joint Institute, Shanghai Jiao Tong University（上海交通大学密西根学院）；Biomedical Engineering Department, Peking University（北京大学生物医学工程系）
