A Feature Point Based Registration of Brain Microscopic Images to the Reference Brain Atlas

Hong Ni,Shangbin Chen,Hui Gong,Qingming Luo,Anan Li

DOI: https://doi.org/10.1364/pibm.2017.w3a.84

2017-01-01

Abstract:Recently, we have developed a Micro-Optical Sectioning Tomography system which enables single neuron resolution. By using a flexible strategy, we have aligned our MOST dataset to a reference brain atlas from Allen from Allen Institute.

What problem does this paper attempt to address?

Micro-Optical Sectioning Tomography to Obtain a High-Resolution Atlas of the Mouse Brain

Anan Li,Hui Gong,Bin Zhang,Qingdi Wang,Cheng Yan,Jingpeng Wu,Qian Liu,Shaoqun Zeng,Qingming Luo

DOI: https://doi.org/10.1126/science.1191776

IF: 56.9

2010-01-01

Science

Abstract:The neuroanatomical architecture is considered to be the basis for understanding brain function and dysfunction. However, existing imaging tools have limitations for brainwide mapping of neural circuits at a mesoscale level. We developed a micro-optical sectioning tomography (MOST) system that can provide micrometer-scale tomography of a centimeter-sized whole mouse brain. Using MOST, we obtained a three-dimensional structural data set of a Golgi-stained whole mouse brain at the neurite level. The morphology and spatial locations of neurons and traces of neurites could be clearly distinguished. We found that neighboring Purkinje cells stick to each other.
Optical Brain Imaging: A Powerful Tool for Neuroscience

Xinpei Zhu,Yanfang Xia,Xuecen Wang,Ke Si,Wei Gong

DOI: https://doi.org/10.1007/s12264-016-0053-6

2017-01-01

Abstract:As the control center of organisms,the brain remains little understood due to its complexity. Taking advantage of imaging methods,scientists have found an accessible approach to unraveling the mystery of neuroscience. Among these methods,optical imaging techniques are widely used due to their high molecular specificity and single-molecule sensitivity. Here,we overview several optical imaging techniques in neuroscience of recent years,including brain clearing,the micro-optical sectioning tomography system,and deep tissue imaging.
An Image Registration Approach Based on 3D Geometric Projection Similarity of the Human Head

Jun Liu,Yun Peng,Hao Chen,Thomas Potter,Yingchun Zhang

DOI: https://doi.org/10.1007/s40846-018-0395-9

IF: 2

2018-01-01

Journal of Medical and Biological Engineering

Abstract:This study aims to develop a novel brain image registration algorithm, the geometry projection similarity (GPS) algorithm, which utilizes the anatomical feature of human head and simplifies the objective function for optimization and search for registration parameters in a step-by-step fashion, largely reducing the solution space and improving the efficiency. The algorithm was tested and compared with existing algorithms in simulation environment and tested on computed tomography/magnetic resonance imaging (CT/MRI) clinical datasets pre- and post-craniotomy surgery. Compared with the maximal mutual information algorithm, the GPS algorithm showed significantly smaller errors (1.2 ± 1.1 vs. 4.5 ± 3.5, 1.0 ± 1.0 vs. 3.5 ± 2.5, 0.8 ± 1.1 vs. 2.8 ± 2.4 pixels, at cluster size of 100, 150 and 200, respectively) and shorter computational durations. Compared with the iterative closest point algorithm, the GPS algorithm showed a greater robustness to the perturbed initial mismatch between registering images (registration failure rate 0 vs. 29 out of 200 trials), while keeping a satisfactorily low registration error (0.71 ± 0.83 pixels). Overall results demonstrated that registration accuracy, efficiency, and robustness of this new algorithm surpass those of existing methods. Pilot test on CT/MRI clinical datasets also showed a satisfactory outcome, demonstrating great potentials for future applications of registration of brain images across modalities.
Cross-Modality Coherent Registration of Whole Mouse Brains

Hanchuan Peng,Lei Qu,Yuanyuan Li,Peng Xie,Lijuan Liu,Yimin Wang,Jun Wu,Yu Liu,Tao Wang,Longfei Li,Kaixuan Guo,Wan Wan,Lei Ouyang,Feng Xiong,Yefeng Zheng,Fang Xu,Hui Gong,Hongwei Dong,Qingming Luo,Guo-Qiang Bi,Michael Hawrylycz,Hongkui Zeng

DOI: https://doi.org/10.21203/rs.3.rs-321118/v1

2021-01-01

Abstract:Abstract Recent whole brain mapping projects are collecting large-scale 3D images using powerful and informative modalities, such as STPT, fMOST, VISoR, or MRI. Registration of these multi-dimensional whole-brain images onto a standard atlas is essential for characterizing neuron types and constructing brain wiring diagrams. However, cross-modality image registration is challenging due to intrinsic variations of brain anatomy and artifacts resulted from different sample preparation methods and imaging modalities. We introduced a cross-modality registration method, called mBrainAligner, which uses coherent landmark mapping as well as deep neural networks to align whole mouse brain images to the standard Allen Common Coordinate Framework atlas. We also built a single cell resolution atlas using the fMOST modality, and used our method to generate whole brain map of 3D full single neuron morphology and neuron cell types.
Cross-modal coherent registration of whole mouse brains

Lei Qu,Yuanyuan Li,Peng Xie,Lijuan Liu,Yimin Wang,Jun Wu,Yu Liu,Tao Wang,Longfei Li,Kaixuan Guo,Wan Wan,Lei Ouyang,Feng Xiong,Anna C. Kolstad,Zhuhao Wu,Fang Xu,Yefeng Zheng,Hui Gong,Qingming Luo,Guoqiang Bi,Hongwei Dong,Michael Hawrylycz,Hongkui Zeng,Hanchuan Peng

DOI: https://doi.org/10.1038/s41592-021-01334-w

IF: 48

2021-12-09

Nature Methods

Abstract:Recent whole-brain mapping projects are collecting large-scale three-dimensional images using modalities such as serial two-photon tomography, fluorescence micro-optical sectioning tomography, light-sheet fluorescence microscopy, volumetric imaging with synchronous on-the-fly scan and readout or magnetic resonance imaging. Registration of these multi-dimensional whole-brain images onto a standard atlas is essential for characterizing neuron types and constructing brain wiring diagrams. However, cross-modal image registration is challenging due to intrinsic variations of brain anatomy and artifacts resulting from different sample preparation methods and imaging modalities. We introduce a cross-modal registration method, mBrainAligner, which uses coherent landmark mapping and deep neural networks to align whole mouse brain images to the standard Allen Common Coordinate Framework atlas. We build a brain atlas for the fluorescence micro-optical sectioning tomography modality to facilitate single-cell mapping, and used our method to generate a whole-brain map of three-dimensional single-neuron morphology and neuron cell types.

biochemical research methods
DeepMapi: a Fully Automatic Registration Method for Mesoscopic Optical Brain Images Using Convolutional Neural Networks

Hong Ni,Zhao Feng,Yue Guan,Xueyan Jia,Wu Chen,Tao Jiang,Qiuyuan Zhong,Jing Yuan,Miao Ren,Xiangning Li,Hui Gong,Qingming Luo,Anan Li

DOI: https://doi.org/10.1007/s12021-020-09483-7

IF: 2.864

2021-01-01

Neuroinformatics

Abstract:The extreme complexity of mammalian brains requires a comprehensive deconstruction of neuroanatomical structures. Scientists normally use a brain stereotactic atlas to determine the locations of neurons and neuronal circuits. However, different brain images are normally not naturally aligned even when they are imaged with the same setup, let alone under the differing resolutions and dataset sizes used in mesoscopic imaging. As a result, it is difficult to achieve high-throughput automatic registration without manual intervention. Here, we propose a deep learning-based registration method called DeepMapi to predict a deformation field used to register mesoscopic optical images to an atlas. We use a self-feedback strategy to address the problem of imbalanced training sets (sampling at a fixed step size in nonuniform brains of structures and deformations) and use a dual-hierarchical network to capture the large and small deformations. By comparing DeepMapi with other registration methods, we demonstrate its superiority over a set of ground truth images, including both optical and MRI images. DeepMapi achieves fully automatic registration of mesoscopic micro-optical images, even macroscopic MRI datasets, in minutes, with an accuracy comparable to those of manual annotations by anatomists.
A Novel Registration Method Of Brain Pet Images And Talairach Atlas

Tan Wenjun,Kang Ying,Sun Xiaohan,Guo Chan

DOI: https://doi.org/10.1109/CCDC.2017.7978552

2017-01-01

Abstract:In recent years, the PET as an important clinical examination imaging technology of radionuclide imaging has become the indispensable tool of for cancer and neurological diseases diagnosis. For the defect of poor resolution and brain structural features are not clear on PET image, it is difficult to select the points accurately when brain PET image register for the brain atlas. The CT image has good spatial resolution and high density resolution. So a novel registration algorithm of PET and standard brain atlas is presented in this paper by transforming PET-CT-atlas multimodality image based on marking the feature points on the CT images. Firstly, the brain tissue, the minimum oriented bounding box and the mid-sagital plane are extracted from brain CT images. Secondly, the brain cortical landmarks is manually marked from CT images and transformed into PET images with mutual information matrix of PET-CT images. Finally, the automatic registration of PET-Talairach atlas is realized with the spatial affine transformation matrix, which is calculated by using the brain cortical landmarks of PET image and Talairach atlas. The experimental results show that the method of multi modality registration of PET-CT-T atlas greatly reduces the error of manual intervention, and is good accuracy and lower time complexity.
A Robust Image Registration Interface for Large Volume Brain Atlas.

Hong Ni,Chaozhen Tan,Zhao Feng,Shangbin Chen,Zoutao Zhang,Wenwei Li,Yue Guan,Hui Gong,Qingming Luo,Anan Li

DOI: https://doi.org/10.1038/s41598-020-59042-y

IF: 4.6

2020-01-01

Scientific Reports

Abstract:Accurately mapping brain structures in three-dimensions is critical for an in-depth understanding of brain functions. Using the brain atlas as a hub, mapping detected datasets into a standard brain space enables efficient use of various datasets. However, because of the heterogeneous and nonuniform brain structure characteristics at the cellular level introduced by recently developed high-resolution whole-brain microscopy techniques, it is difficult to apply a single standard to robust registration of various large-volume datasets. In this study, we propose a robust Brain Spatial Mapping Interface (BrainsMapi) to address the registration of large-volume datasets by introducing extracted anatomically invariant regional features and a large-volume data transformation method. By performing validation on model data and biological images, BrainsMapi achieves accurate registration on intramodal, individual, and multimodality datasets and can also complete the registration of large-volume datasets (approximately 20 TB) within 1 day. In addition, it can register and integrate unregistered vectorized datasets into a common brain space. BrainsMapi will facilitate the comparison, reuse and integration of a variety of brain datasets.
Co-registration of Diffusion Tensor Imaging and Micro-Optical Imaging Based on Ants

Hong Ni,Miao Ren,Qiuyuan Zhong,Hui Gong,Qingming Luo,Shangbin Chen

DOI: https://doi.org/10.1049/cp.2015.0762

2015-01-01

Abstract:Diffusion tensor imaging (DTI) can provide important macroscopic structural information for mouse brain but is limited by the available imaging resolution and inferentia tractography. To validate DTI tractography, it is paramount to merge DTI images and microscopic images with true representation of fibers. Recently, we have developed a micro-optical sectioning tomography (MOST) system which enables neurite level resolution. By using Advanced Normalization Tools, we have aligned our MOST dataset to the template dataset of DTI from Duke University. To optimize the computing efficacy, lower resolution of two different modal images has been used to get a displacement field of diffeomorphism. Then the displacement field is extended to the deformation of full resolution images. This has been a flexible strategy for 3D nonlinear mouse brain registration across different modalities and different resolutions. Under this kind of co-registration, we can show very fine neural fiber architectures as foreground (MOST) on the background (DTI) of the fibre density map. By comparing the derived maps between DTI and MOST, we have realized that DTI technique encounters pitfall for resolving complex neural fibres. Image fusion of mouse DTI and MOST dataset should promote both neural circuits study and DTI applications.
Visible brain-wide networks at single-neuron resolution with Micro-Optical Sectioning Tomography

Qingming Luo

DOI: https://doi.org/10.1364/cleo_at.2014.af1b.5

2014-01-01

Abstract:We report a protocol combining a novel resin-embedding method for maintaining fluorescence, an automated fluorescence MOST system for long-term stable imaging, and a digital reconstruction-registration-annotation pipeline for tracing the axonal pathways individually in the mouse brain.
[Chinese Brain Template Built on 3D High Resolution MR Imaging at 3.0T].

Dong Zeng,Tijiang Zhang,Lijun Jiang,Dongming Li,Wei Deng,Xiuli Li,Hehan Tang,Ling Zou,Su Lu,Xiaoqi Huang,Tao Li,Dezhong Yao,Qiyong Gong

2010-01-01

Abstract:Brain atlas provides a spatial reference system on which other images can be interpreted in a consistent way, and it is essential for the brain imaging research. However, because of the differences in structure between occidental and oriental brains, the brain atlas based on Western populations, e. g., the International Consortium for Brain Mapping's 154 T1 Weighted Average Atlas, may not be appropriate for other ethnic groups. Therefore, in the present study, we produce an average brain atlas which is based on the data collected from 100 healthy Chinese volunteers. The differences in brains between the Chinese population and the Western population were also investigated. Comparatively,Chinese brains are wider and shorter in size, and smaller in volume.
A Target-Oriented Segmentation Method for Specific Tissues in MRI Images of the Brain

Enming Song,Yuejing Qian,Hong Liu,Meng Yan,Huimin Song,Chih-Cheng Hung

DOI: https://doi.org/10.1007/s11042-017-5484-1

IF: 2.577

2017-01-01

Multimedia Tools and Applications

Abstract:The multi-atlas based segmentation method can achieve the accurate segmentation of specific tissues of the human brain in the magnetic resonance imaging (MRI). The correct image registration and fusion scheme used in this method have an impact on the accuracy of segmentation. Similar to any traditional rigid registration method, we use the same method in our proposed target-oriented registration for the coarse registration between the target image and atlas image. However, to improve the registration accuracy in the area to be segmented, we propose a target-oriented image registration method for the refinement. We employ the distribution probability of the tissue (to be segmented) in the sparse patch-based label fusion process. Our aim is to determine if the proposed registration method can contribute the segmentation accuracy and which label fusion method is a good fit with this target-oriented registration. To evaluate the efficiency of our proposed method, we compare the performance of the majority voting method (MV), the nonlocal patch-based method (Nonlocal-PBM) and the sparse patch-based method (Sparse-PBM). Experimental results show that more accurate segmentation results can be obtained with the proposed registration method in this study. This result can provide more accurate clinical diagnosis information.
High-throughput Dual-Color Precision Imaging for Brain-Wide Mapping of the Connectome with Cytoarchitectonic Landmarks at the Cellular Level (conference Presentation)

Qingming Luo,Hui Gong,Jing Yuan,Xiangning Li,Anan Li,Tonghui Xu

DOI: https://doi.org/10.1117/12.2251624

IF: 16.6

2016-01-01

Nature Communications

Abstract:Deciphering the fine morphology and precise location of neurons and neural circuits are crucial to enhance our understanding of brain function and diseases. Traditionally, we have to map brain images to coarse axial-sampling planar reference atlases to orient neural structures. However, this means might fail to orient neural projections at single-cell resolution due to position errors resulting from individual differences at the cellular level. Here, we present a high-throughput imaging method that can automatically obtain the fine morphologies and precise locations of both neurons and circuits, employing wide-field large-volume tomography to acquire three-dimensional images of thick tissue and implementing real-time soma counterstaining to obtain cytoarchitectonic landmarks during the imaging process. The reconstruction and orientation of brain-wide neural circuits at single-neuron resolution can be accomplished for the same mouse brain without additional counterstains or image registration. Using our method, mouse brain imaging datasets of multiple type-specific neurons and circuits were successfully acquired, demonstrating the versatility. The results show that the simultaneous acquisition of labeled neural structures and cytoarchitecture reference at single-neuron resolution in the same brain greatly facilitates precise tracing of long-range projections and accurate locating of nuclei. Our method provides a novel and effective tool for application in studies on genetic dissection, brain function and the pathology of the nervous system.
Automated Brain Structures Segmentation from PET/CT Images Based on Landmark-Constrained Dual-Modality Atlas Registration.

Zhaofeng Chen,Tianshuang Qiu,Yang Tian,Hongbo Feng,Yanjun Zhang,Hongkai Wang

DOI: https://doi.org/10.1088/1361-6560/abf201

2021-01-01

Abstract:Automated brain structures segmentation in positron emission tomography (PET) images has been widely investigated to help brain disease diagnosis and follow-up. To relieve the burden of a manual definition of volume of interest (VOI), automated atlas-based VOI definition algorithms were developed, but these algorithms mostly adopted a global optimization strategy which may not be particularly accurate for local small structures (especially the deep brain structures). This paper presents a PET/CT-based brain VOI segmentation algorithm combining anatomical atlas, local landmarks, and dual-modality information. The method incorporates local deep brain landmarks detected by the Deep Q-Network (DQN) to constrain the atlas registration process. Dual-modality PET/CT image information is also combined to improve the registration accuracy of the extracerebral contour. We compare our algorithm with the representative brain atlas registration methods based on 86 clinical PET/CT images. The proposed algorithm obtained accurate delineation of brain VOIs with an average Dice similarity score of 0.79, an average surface distance of 0.97 mm (sub-pixel level), and a volume recovery coefficient close to 1. The main advantage of our method is that it optimizes both global-scale brain matching and local-scale small structure alignment around the key landmarks, it is fully automated and produces high-quality parcellation of the brain structures from brain PET/CT images.
Opportunities and challenges for mouse brainwide neuroanatomical connectivity with micrometer resolution

Qingming Luo,Li Anan,Bin Zhang,H. Gong,A. Li,H. Gong,Bin Zhang,Qingdi Wang,Cheng Yan,Jinpeng Wu,Qian Liu,S. Zeng,Zhongqin Yang,Jingpeng Wu,Qingming Luo

2012-01-01

Abstract:We obtained a three-dimensional structural data set of a Golgi-stained whole mouse brain at the neurite level with our developed Micro-Optical Sectioning Tomography (MOST) system, which directly demonstrates the whole-brain structural connectivity at the neurite level in a standard format. Our results demonstrated a resolution of micro meter in three dimensions across the whole brain, which was not provided by other techniques.
NeuroAtlas: An Artificial Intelligence-based Framework for Annotation, Segmentation and Registration of Large Scale Biomedical Imaging Data

Hassan Mahmood,Farah Nawar,Syed Mohammed Shamsul Islam,Asim Iqbal

DOI: https://doi.org/10.1101/2024.08.24.609507

2024-08-26

Abstract:With increasing neuroimaging modalities and data diversity, mapping brain regions to a standard atlas template has become a challenging problem. Machine learning in general and deep learning, in particular, have been providing robust solutions for several neuroimaging tasks, including brain image registration and segmentation. However, these methods require a large amount of data for ground-truth labels, annotated by human experts, which is time-consuming. In this work, we introduce NeuroAtlas, an AI-based framework for atlas generation and brain region segmentation. We showcase an end-to-end solution for brain registration and segmentation by providing i) a deep learning modeling suite with a variety of high-performing model architectures to map a brain atlas onto the input brain section and ii) a Graphical User Interface (GUI)-based plugin for large-scale data annotation with a feature of modifying the predicted labels for active learning. We demonstrate a robust performance of our framework on the human brains, captured through various imaging modalities and age groups, and demonstrate its application for mouse brains as well. NeuroAtlas tool will be open-sourced and entirely compatible with both local as well as cloud-based computing so that users can easily adapt to their neuroimaging custom datasets.

Neuroscience
Deep-learning-based Whole-Brain Imaging at Single-Neuron Resolution

Kefu Ning,Xiaoyu Zhang,Xuefei Gao,Tao Jiang,He Wang,Siqi Chen,Anan Li,Jing Yuan

DOI: https://doi.org/10.1364/boe.393081

IF: 3.4

2020-01-01

Biomedical Optics Express

Abstract:Obtaining fine structures of neurons is necessary for understanding brain function. Simple and effective methods for large-scale 3D imaging at optical resolution are still lacking. Here, we proposed a deep-learning-based fluorescence micro-optical sectioning tomography (DL-fMOST) method for high-throughput, high-resolution whole-brain imaging. We utilized a wide-field microscope for imaging, a U-net convolutional neural network for real-time optical sectioning, and histological sectioning for exceeding the imaging depth limit. A 3D dataset of a mouse brain with a voxel size of 0.32 × 0.32 × 2 µm was acquired in 1.5 days. We demonstrated the robustness of DL-fMOST for mouse brains with labeling of different types of neurons.
3D Affine Registration of Large Image Stack for High-Resolution Brain Imaging Techniques

Chaozhen Tan,Zhao Feng,Hong Ni,Yuxin Li,Hui Gong,Qingming Luo,Anan Li

DOI: https://doi.org/10.1364/pibm.2017.w3a.101

2017-01-01

Abstract:To solve the stereotaxic deviation and sample deformation in Micro-Optical brain images at high resolution, we develop a high performance 3D affine registration method for large image stack to correct the stereotaxic space and morphology.
Micro-Optical Sectioning Tomography for Visualizing the Mouse Brainwide Neuroanatomical Connectivity

Qingming Luo

DOI: https://doi.org/10.1109/foi.2011.6154830

2011-01-01

Abstract:Neuroanatomical architecture is considered the basis for understanding brain functions and dysfunctions. Several imaging approaches have been made for brainwide mapping of neural circuits at a mesoscopic scale, including our developed Micro-Optical Sectioning Tomography (MOST) system that can provide submicron tomography of a centimeter-sized whole mouse brain. In this talk, we will clarify the unique features of the MOST in comparison to other techniques, discuss the challenges for visualizing the mouse brainwide neuroanatomical connectivity at the neurite level. Furthermore, we will present the new developed fMOST for imaging the fluorescent proteins or other fluorescent markers in the mouse brain.
An Atlas Based Level Set Method for 3D Brain Structures Segmentation

Junyi Yan,Jinzhu Yang,Wenjun Tan,Peng Cao,Lili Dou,Dazhe Zhao

DOI: https://doi.org/10.1145/3451421.3451444

2020-01-01

Abstract:Brain structures segmentation in Magnetic Resonance Imaging (MRI) is a challenging task because of the large anatomical variability in both shape and size between individuals. In this work, an atlas based level set method is proposed for brain tissues segmentation. First, a skull stripping method is utilized to remove the skull region of brain. A registration algorithm is then applied to obtain the initial six brain structure labels (thalamus, hippocampus, amygdala, putamen, pallidum, caudate) with a manually labeled atlas, which can be used as the initial level set contours. The final label is achieved by minimizing the energy function which consists of three terms: an image data term, a length regularization term and an intensity constrained term. Experimental results demonstrate that our method can obtain satisfactory results compared with other state-of-art methods.

A Feature Point Based Registration of Brain Microscopic Images to the Reference Brain Atlas

Micro-Optical Sectioning Tomography to Obtain a High-Resolution Atlas of the Mouse Brain

Optical Brain Imaging: A Powerful Tool for Neuroscience

An Image Registration Approach Based on 3D Geometric Projection Similarity of the Human Head

Cross-Modality Coherent Registration of Whole Mouse Brains

Cross-modal coherent registration of whole mouse brains

DeepMapi: a Fully Automatic Registration Method for Mesoscopic Optical Brain Images Using Convolutional Neural Networks

A Novel Registration Method Of Brain Pet Images And Talairach Atlas

A Robust Image Registration Interface for Large Volume Brain Atlas.

Co-registration of Diffusion Tensor Imaging and Micro-Optical Imaging Based on Ants

Visible brain-wide networks at single-neuron resolution with Micro-Optical Sectioning Tomography

[Chinese Brain Template Built on 3D High Resolution MR Imaging at 3.0T].

A Target-Oriented Segmentation Method for Specific Tissues in MRI Images of the Brain

High-throughput Dual-Color Precision Imaging for Brain-Wide Mapping of the Connectome with Cytoarchitectonic Landmarks at the Cellular Level (conference Presentation)

Automated Brain Structures Segmentation from PET/CT Images Based on Landmark-Constrained Dual-Modality Atlas Registration.

Opportunities and challenges for mouse brainwide neuroanatomical connectivity with micrometer resolution

NeuroAtlas: An Artificial Intelligence-based Framework for Annotation, Segmentation and Registration of Large Scale Biomedical Imaging Data

Deep-learning-based Whole-Brain Imaging at Single-Neuron Resolution

3D Affine Registration of Large Image Stack for High-Resolution Brain Imaging Techniques

Micro-Optical Sectioning Tomography for Visualizing the Mouse Brainwide Neuroanatomical Connectivity

An Atlas Based Level Set Method for 3D Brain Structures Segmentation