Anatomical Modeling of Brain Vasculature in Two-Photon Microscopy by Generalizable Deep Learning

作     者：Waleed Tahir Sreekanth Kura Jiabei Zhu Xiaojun Cheng Rafat Damseh Fetsum Tadesse Alex Seibel Blaire S.Lee Frédéric Lesage Sava Sakadžic David A.Boas Lei Tian 

作者机构：Department of Electrical and Computer EngineeringBoston UniversityBostonMAUSA Department of Biomedical EngineeringBoston UniversityBostonMAUSA Biomedical Engineering InstituteÉcole Polytechnique de MontréalMontréalQCCanada Institute of Neurological Sciences and PsychiatryHacettepe UniversityAnkaraTurkey Department of RadiologyMassachusetts General HospitalHarvard Medical SchoolCharlestownUSA Neurophotonics CenterBoston UniversityBostonMAUSA 

出 版 物：《Biomedical Engineering Frontiers》 (生物医学工程前沿（英文）)

年 卷 期：2021年第2卷第1期

页      面：103-114页

核心收录：

学科分类：0831[工学-生物医学工程（可授工学、理学、医学学位）] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：supported by NIH 3R01EB021018-04S2 

主　　题：Deep network overcome 

摘      要：Objective and Impact *** of blood vessels from two-photon microscopy(2PM)angiograms of brains has important applications in hemodynamic analysis and disease ***,we develop a generalizable deep learning technique for accurate 2PM vascular segmentation of sizable regions in mouse brains acquired from multiple 2PM *** technique is computationally efficient,thus ideal for large-scale neurovascular *** segmentation from 2PM angiograms is an important first step in hemodynamic modeling of brain *** segmentation methods based on deep learning either lack the ability to generalize to data from different imaging systems or are computationally infeasible for large-scale *** this work,we overcome both these limitations by a method that is generalizable to various imaging systems and is able to segment large-scale *** employ a computationally efficient deep learning framework with a loss function that incorporates a balanced binary-cross-entropy loss and total variation regularization on the network’s *** effectiveness is demonstrated on experimentally acquired in vivo angiograms from mouse brains of dimensions up to 808×808×702μ*** demonstrate the superior generalizability of our framework,we train on data from only one 2PM microscope and demonstrate high-quality segmentation on data from a different microscope without any network ***,our method demonstrates 10×faster computation in terms of voxels-segmented-per-second and 3×larger depth compared to the *** work provides a generalizable and computationally efficient anatomical modeling framework for brain vasculature,which consists of deep learning-based vascular segmentation followed by *** paves the way for future modeling and analysis of hemodynamic response at much greater scales that were inaccessible before.