1
|
Zehtabian A, Fuchs J, Eickholt BJ, Ewers H. Automated Analysis of Neuronal Morphology in 2D Fluorescence Micrographs through an Unsupervised Semantic Segmentation of Neurons. Neuroscience 2024; 551:333-344. [PMID: 38838980 DOI: 10.1016/j.neuroscience.2024.05.024] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/19/2023] [Revised: 05/17/2024] [Accepted: 05/20/2024] [Indexed: 06/07/2024]
Abstract
Brain function emerges from a highly complex network of specialized cells that are interlinked by billions of synapses. The synaptic connectivity between neurons is established between the elongated processes of their axons and dendrites or, together, neurites. To establish these connections, cellular neurites have to grow in highly specialized, cell-type dependent patterns covering extensive distances and connecting with thousands of other neurons. The outgrowth and branching of neurites are tightly controlled during development and are a commonly used functional readout of imaging in the neurosciences. Manual analysis of neuronal morphology from microscopy images, however, is very time intensive and prone to bias. Most automated analyses of neurons rely on reconstruction of the neuron as a whole without a semantic analysis of each neurite. A fully-automated classification of all neurites still remains unavailable in open-source software. Here we present a standalone, GUI-based software for batch-quantification of neuronal morphology in two-dimensional fluorescence micrographs of cultured neurons with minimal requirements for user interaction. Single neurons are first reconstructed into binarized images using a Hessian-based segmentation algorithm to detect thin neurite structures combined with intensity- and shape-based reconstruction of the cell body. Neurites are then classified into axon, dendrites and their branches of increasing order using a geodesic distance transform of the cell skeleton. The software was benchmarked against a published dataset and reproduced the phenotype observed after manual annotation. Our tool promises accelerated and improved morphometric studies of neuronal morphology by allowing for consistent and automated analysis of large datasets.
Collapse
Affiliation(s)
- Amin Zehtabian
- Institute for Chemistry and Biochemistry, Freie Universität Berlin, Thielallee 63, 14195 Berlin, Germany.
| | - Joachim Fuchs
- Charité - Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Institute of Molecular Biology and Biochemistry, Virchowweg 6, 10117 Berlin, Germany
| | - Britta J Eickholt
- Charité - Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Institute of Molecular Biology and Biochemistry, Virchowweg 6, 10117 Berlin, Germany
| | - Helge Ewers
- Institute for Chemistry and Biochemistry, Freie Universität Berlin, Thielallee 63, 14195 Berlin, Germany
| |
Collapse
|
2
|
Caznok Silveira AC, Antunes ASLM, Athié MCP, da Silva BF, Ribeiro dos Santos JV, Canateli C, Fontoura MA, Pinto A, Pimentel-Silva LR, Avansini SH, de Carvalho M. Between neurons and networks: investigating mesoscale brain connectivity in neurological and psychiatric disorders. Front Neurosci 2024; 18:1340345. [PMID: 38445254 PMCID: PMC10912403 DOI: 10.3389/fnins.2024.1340345] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/17/2023] [Accepted: 01/29/2024] [Indexed: 03/07/2024] Open
Abstract
The study of brain connectivity has been a cornerstone in understanding the complexities of neurological and psychiatric disorders. It has provided invaluable insights into the functional architecture of the brain and how it is perturbed in disorders. However, a persistent challenge has been achieving the proper spatial resolution, and developing computational algorithms to address biological questions at the multi-cellular level, a scale often referred to as the mesoscale. Historically, neuroimaging studies of brain connectivity have predominantly focused on the macroscale, providing insights into inter-regional brain connections but often falling short of resolving the intricacies of neural circuitry at the cellular or mesoscale level. This limitation has hindered our ability to fully comprehend the underlying mechanisms of neurological and psychiatric disorders and to develop targeted interventions. In light of this issue, our review manuscript seeks to bridge this critical gap by delving into the domain of mesoscale neuroimaging. We aim to provide a comprehensive overview of conditions affected by aberrant neural connections, image acquisition techniques, feature extraction, and data analysis methods that are specifically tailored to the mesoscale. We further delineate the potential of brain connectivity research to elucidate complex biological questions, with a particular focus on schizophrenia and epilepsy. This review encompasses topics such as dendritic spine quantification, single neuron morphology, and brain region connectivity. We aim to showcase the applicability and significance of mesoscale neuroimaging techniques in the field of neuroscience, highlighting their potential for gaining insights into the complexities of neurological and psychiatric disorders.
Collapse
Affiliation(s)
- Ana Clara Caznok Silveira
- National Laboratory of Biosciences, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
- School of Electrical and Computer Engineering, University of Campinas, Campinas, Brazil
| | | | - Maria Carolina Pedro Athié
- National Laboratory of Biosciences, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
| | - Bárbara Filomena da Silva
- National Laboratory of Biosciences, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
| | | | - Camila Canateli
- National Laboratory of Biosciences, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
| | - Marina Alves Fontoura
- National Laboratory of Biosciences, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
| | - Allan Pinto
- Brazilian Synchrotron Light Laboratory, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
| | | | - Simoni Helena Avansini
- National Laboratory of Biosciences, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
| | - Murilo de Carvalho
- National Laboratory of Biosciences, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
- Brazilian Synchrotron Light Laboratory, Brazilian Center for Research in Energy and Materials, Campinas, Brazil
| |
Collapse
|
3
|
Ma R, Hao L, Tao Y, Mendoza X, Khodeiry M, Liu Y, Shyu ML, Lee RK. RGC-Net: An Automatic Reconstruction and Quantification Algorithm for Retinal Ganglion Cells Based on Deep Learning. Transl Vis Sci Technol 2023; 12:7. [PMID: 37140906 PMCID: PMC10166122 DOI: 10.1167/tvst.12.5.7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/01/2022] [Accepted: 03/31/2023] [Indexed: 05/05/2023] Open
Abstract
Purpose The purpose of this study was to develop a deep learning-based fully automated reconstruction and quantification algorithm which automatically delineates the neurites and somas of retinal ganglion cells (RGCs). Methods We trained a deep learning-based multi-task image segmentation model, RGC-Net, that automatically segments the neurites and somas in RGC images. A total of 166 RGC scans with manual annotations from human experts were used to develop this model, whereas 132 scans were used for training, and the remaining 34 scans were reserved as testing data. Post-processing techniques removed speckles or dead cells in soma segmentation results to further improve the robustness of the model. Quantification analyses were also conducted to compare five different metrics obtained by our automated algorithm and manual annotations. Results Quantitatively, our segmentation model achieves average foreground accuracy, background accuracy, overall accuracy, and dice similarity coefficient of 0.692, 0.999, 0.997, and 0.691 for the neurite segmentation task, and 0.865, 0.999, 0.997, and 0.850 for the soma segmentation task, respectively. Conclusions The experimental results demonstrate that RGC-Net can accurately and reliably reconstruct neurites and somas in RGC images. We also demonstrate our algorithm is comparable to human manually curated annotations in quantification analyses. Translational Relevance Our deep learning model provides a new tool that can trace and analyze the RGC neurites and somas efficiently and faster than manual analysis.
Collapse
Affiliation(s)
- Rui Ma
- Department of Electrical and Computer Engineering, University of Miami, Coral Gables, FL, USA
| | - Lili Hao
- Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
- Department of Ophthalmology, The First Affiliated Hospital of Jinan University, Guangzhou, China
| | - Yudong Tao
- Department of Electrical and Computer Engineering, University of Miami, Coral Gables, FL, USA
| | - Ximena Mendoza
- Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
| | - Mohamed Khodeiry
- Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
| | - Yuan Liu
- Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
| | - Mei-Ling Shyu
- School of Science and Engineering, University of Missouri-Kansas City, Kansas City, MO, USA
| | - Richard K. Lee
- Department of Electrical and Computer Engineering, University of Miami, Coral Gables, FL, USA
- Bascom Palmer Eye Institute, University of Miami Miller School of Medicine, Miami, FL, USA
| |
Collapse
|
4
|
Analysis of the Expression and Subcellular Distribution of eEF1A1 and eEF1A2 mRNAs during Neurodevelopment. Cells 2022; 11:cells11121877. [PMID: 35741005 PMCID: PMC9220863 DOI: 10.3390/cells11121877] [Citation(s) in RCA: 4] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 05/12/2022] [Revised: 05/30/2022] [Accepted: 05/31/2022] [Indexed: 12/04/2022] Open
Abstract
Neurodevelopment is accompanied by a precise change in the expression of the translation elongation factor 1A variants from eEF1A1 to eEF1A2. These are paralogue genes that encode 92% identical proteins in mammals. The switch in the expression of eEF1A variants has been well studied in mouse motor neurons, which solely express eEF1A2 by four weeks of postnatal development. However, changes in the subcellular localization of eEF1A variants during neurodevelopment have not been studied in detail in other neuronal types because antibodies lack perfect specificity, and immunofluorescence has a low sensitivity. In hippocampal neurons, eEF1A is related to synaptic plasticity and memory consolidation, and decreased eEF1A expression is observed in the hippocampus of Alzheimer's patients. However, the specific variant involved in these functions is unknown. To distinguish eEF1A1 from eEF1A2 expression, we have designed single-molecule fluorescence in-situ hybridization probes to detect either eEF1A1 or eEF1A2 mRNAs in cultured primary hippocampal neurons and brain tissues. We have developed a computational framework, ARLIN (analysis of RNA localization in neurons), to analyze and compare the subcellular distribution of eEF1A1 and eEF1A2 mRNAs at specific developmental stages and in mature neurons. We found that eEF1A1 and eEF1A2 mRNAs differ in expression and subcellular localization over neurodevelopment, and eEF1A1 mRNAs localize in dendrites and synapses during dendritogenesis and synaptogenesis. Interestingly, mature hippocampal neurons coexpress both variant mRNAs, and eEF1A1 remains the predominant variant in dendrites.
Collapse
|
5
|
Wu X, Tao Y, He G, Liu D, Fan M, Yang S, Gong H, Xiao R, Chen S, Huang J. Boosting Multilabel Semantic Segmentation for Somata and Vessels in Mouse Brain. Front Neurosci 2021; 15:610122. [PMID: 33912000 PMCID: PMC8071950 DOI: 10.3389/fnins.2021.610122] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/25/2020] [Accepted: 03/22/2021] [Indexed: 11/17/2022] Open
Abstract
Deep convolutional neural networks (DCNNs) are widely utilized for the semantic segmentation of dense nerve tissues from light and electron microscopy (EM) image data; the goal of this technique is to achieve efficient and accurate three-dimensional reconstruction of the vasculature and neural networks in the brain. The success of these tasks heavily depends on the amount, and especially the quality, of the human-annotated labels fed into DCNNs. However, it is often difficult to acquire the gold standard of human-annotated labels for dense nerve tissues; human annotations inevitably contain discrepancies or even errors, which substantially impact the performance of DCNNs. Thus, a novel boosting framework consisting of a DCNN for multilabel semantic segmentation with a customized Dice-logarithmic loss function, a fusion module combining the annotated labels and the corresponding predictions from the DCNN, and a boosting algorithm to sequentially update the sample weights during network training iterations was proposed to systematically improve the quality of the annotated labels; this framework eventually resulted in improved segmentation task performance. The microoptical sectioning tomography (MOST) dataset was then employed to assess the effectiveness of the proposed framework. The result indicated that the framework, even trained with a dataset including some poor-quality human-annotated labels, achieved state-of-the-art performance in the segmentation of somata and vessels in the mouse brain. Thus, the proposed technique of artificial intelligence could advance neuroscience research.
Collapse
Affiliation(s)
- Xinglong Wu
- School of Computer Science & Engineering, Wuhan Institute of Technology, Wuhan, China
| | - Yuhang Tao
- Engineering Research Center of Hubei Province for Clothing Information, School of Mathematics and Computer Science, Wuhan Textile University, Wuhan, China
| | - Guangzhi He
- Engineering Research Center of Hubei Province for Clothing Information, School of Mathematics and Computer Science, Wuhan Textile University, Wuhan, China
| | - Dun Liu
- Engineering Research Center of Hubei Province for Clothing Information, School of Mathematics and Computer Science, Wuhan Textile University, Wuhan, China
| | - Meiling Fan
- Engineering Research Center of Hubei Province for Clothing Information, School of Mathematics and Computer Science, Wuhan Textile University, Wuhan, China
| | - Shuo Yang
- Engineering Research Center of Hubei Province for Clothing Information, School of Mathematics and Computer Science, Wuhan Textile University, Wuhan, China
| | - Hui Gong
- Britton Chance Center for Biomedical Photonics, Wuhan National Laboratory for Optoelectronics-Huazhong University of Science and Technology, Wuhan, China
| | - Rong Xiao
- Britton Chance Center for Biomedical Photonics, Wuhan National Laboratory for Optoelectronics-Huazhong University of Science and Technology, Wuhan, China
| | - Shangbin Chen
- Britton Chance Center for Biomedical Photonics, Wuhan National Laboratory for Optoelectronics-Huazhong University of Science and Technology, Wuhan, China
| | - Jin Huang
- Engineering Research Center of Hubei Province for Clothing Information, School of Mathematics and Computer Science, Wuhan Textile University, Wuhan, China
| |
Collapse
|