1
|
Théberge A, Desrosiers C, Boré A, Descoteaux M, Jodoin PM. What matters in reinforcement learning for tractography. Med Image Anal 2024; 93:103085. [PMID: 38219499 DOI: 10.1016/j.media.2024.103085] [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: 05/21/2023] [Revised: 12/15/2023] [Accepted: 01/05/2024] [Indexed: 01/16/2024]
Abstract
Recently, deep reinforcement learning (RL) has been proposed to learn the tractography procedure and train agents to reconstruct the structure of the white matter without manually curated reference streamlines. While the performances reported were competitive, the proposed framework is complex, and little is still known about the role and impact of its multiple parts. In this work, we thoroughly explore the different components of the proposed framework, such as the choice of the RL algorithm, seeding strategy, the input signal and reward function, and shed light on their impact. Approximately 7,400 models were trained for this work, totalling nearly 41,000 h of GPU time. Our goal is to guide researchers eager to explore the possibilities of deep RL for tractography by exposing what works and what does not work with the category of approach. As such, we ultimately propose a series of recommendations concerning the choice of RL algorithm, the input to the agents, the reward function and more to help future work using reinforcement learning for tractography. We also release the open source codebase, trained models, and datasets for users and researchers wanting to explore reinforcement learning for tractography.
Collapse
Affiliation(s)
- Antoine Théberge
- Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada, J1K 2R1.
| | - Christian Desrosiers
- Département de génie logiciel et des TI, École de technologie supérieure, Montréal, QC, Canada, H3C 1K3
| | - Arnaud Boré
- Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada, J1K 2R1
| | - Maxime Descoteaux
- Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada, J1K 2R1
| | - Pierre-Marc Jodoin
- Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, Canada, J1K 2R1
| |
Collapse
|
2
|
Jin R, Cai Y, Zhang S, Yang T, Feng H, Jiang H, Zhang X, Hu Y, Liu J. Computational approaches for the reconstruction of optic nerve fibers along the visual pathway from medical images: a comprehensive review. Front Neurosci 2023; 17:1191999. [PMID: 37304011 PMCID: PMC10250625 DOI: 10.3389/fnins.2023.1191999] [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: 03/22/2023] [Accepted: 05/09/2023] [Indexed: 06/13/2023] Open
Abstract
Optic never fibers in the visual pathway play significant roles in vision formation. Damages of optic nerve fibers are biomarkers for the diagnosis of various ophthalmological and neurological diseases; also, there is a need to prevent the optic nerve fibers from getting damaged in neurosurgery and radiation therapy. Reconstruction of optic nerve fibers from medical images can facilitate all these clinical applications. Although many computational methods are developed for the reconstruction of optic nerve fibers, a comprehensive review of these methods is still lacking. This paper described both the two strategies for optic nerve fiber reconstruction applied in existing studies, i.e., image segmentation and fiber tracking. In comparison to image segmentation, fiber tracking can delineate more detailed structures of optic nerve fibers. For each strategy, both conventional and AI-based approaches were introduced, and the latter usually demonstrates better performance than the former. From the review, we concluded that AI-based methods are the trend for optic nerve fiber reconstruction and some new techniques like generative AI can help address the current challenges in optic nerve fiber reconstruction.
Collapse
Affiliation(s)
- Richu Jin
- Research Institute of Trustworthy Autonomous Systems, Southern University of Science and Technology, Shenzhen, China
- Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
| | - Yongning Cai
- Research Institute of Trustworthy Autonomous Systems, Southern University of Science and Technology, Shenzhen, China
| | - Shiyang Zhang
- Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
| | - Ting Yang
- Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
| | - Haibo Feng
- Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
| | - Hongyang Jiang
- Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
| | - Xiaoqing Zhang
- Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
| | - Yan Hu
- Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
| | - Jiang Liu
- Research Institute of Trustworthy Autonomous Systems, Southern University of Science and Technology, Shenzhen, China
- Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
- Guangdong Provincial Key Laboratory of Brain-inspired Intelligent Computation, Department of Computer Science and Engineering, Southern University of Science and Technology, Shenzhen, China
| |
Collapse
|
3
|
Renauld E, Théberge A, Petit L, Houde JC, Descoteaux M. Validate your white matter tractography algorithms with a reappraised ISMRM 2015 Tractography Challenge scoring system. Sci Rep 2023; 13:2347. [PMID: 36759653 PMCID: PMC9911766 DOI: 10.1038/s41598-023-28560-w] [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: 12/24/2022] [Accepted: 01/20/2023] [Indexed: 02/11/2023] Open
Abstract
Since 2015, research groups have sought to produce the ne plus ultra of tractography algorithms using the ISMRM 2015 Tractography Challenge as evaluation. In particular, since 2017, machine learning has made its entrance into the tractography world. The ISMRM 2015 Tractography Challenge is the most used phantom during tractography validation, although it contains limitations. Here, we offer a new scoring system for this phantom, where segmentation of the bundles is now based on manually defined regions of interest rather than on bundle recognition. Bundles are now more reliably segmented, offering more representative metrics for future users. New code is available online. Scores of the initial 96 submissions to the challenge are updated. Overall, conclusions from the 2015 challenge are confirmed with the new scoring, but individual tractogram scores have changed, and the data is much improved at the bundle- and streamline-level. This work also led to the production of a ground truth tractogram with less broken or looping streamlines and of an example of processed data, all available on the Tractometer website. This enhanced scoring system and new data should continue helping researchers develop and evaluate the next generation of tractography techniques.
Collapse
Affiliation(s)
- Emmanuelle Renauld
- Sherbrooke Connectivity Imaging Laboratory (SCIL), Computer Sciences Department, Université de Sherbrooke, Sherbrooke, Canada.
| | - Antoine Théberge
- Sherbrooke Connectivity Imaging Laboratory (SCIL), Computer Sciences Department, Université de Sherbrooke, Sherbrooke, Canada
| | - Laurent Petit
- Université de Bordeaux, CNRS, CEA, IMN, GIN, UMR 5293, 33000, Bordeaux, France
| | | | - Maxime Descoteaux
- Sherbrooke Connectivity Imaging Laboratory (SCIL), Computer Sciences Department, Université de Sherbrooke, Sherbrooke, Canada.,Imeka Solutions Inc, Sherbrooke, QC, Canada
| |
Collapse
|
4
|
Yan Y, Fan G, Liao X, Zhao X. Research trends and hotspots on connectomes from 2005 to 2021: A bibliometric and latent Dirichlet allocation application study. Front Neurosci 2022; 16:1046562. [PMID: 36620450 PMCID: PMC9814013 DOI: 10.3389/fnins.2022.1046562] [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: 09/16/2022] [Accepted: 12/06/2022] [Indexed: 12/24/2022] Open
Abstract
Background This study aimed to conduct a bibliometric analysis of publications on connectomes and illustrate its trends and hotspots using a machine-learning-based text mining algorithm. Methods Documents were retrieved from the Web of Science Core Collection (WoSCC) and Scopus databases and analyzed in Rstudio 1.3.1. Through quantitative and qualitative methods, the most productive and impactful academic journals in the field of connectomes were compared in terms of the total number of publications and h-index over time. Meanwhile, the countries/regions and institutions involved in connectome research were compared, as well as their scientific collaboration. The study analyzed topics and research trends by R package "bibliometrix." The major topics of connectomes were classified by Latent Dirichlet allocation (LDA). Results A total of 14,140 publications were included in the study. NEUROIMAGE ranked first in terms of publication volume (1,427 articles) and impact factor (h-index:122) among all the relevant journals. The majority of articles were published by developed countries, with the United States having the most. Harvard Medical School and the University of Pennsylvania were the two most productive institutions. Neuroimaging analysis technology and brain functions and diseases were the two major topics of connectome research. The application of machine learning, deep learning, and graph theory analysis in connectome research has become the current trend, while an increasing number of studies were concentrating on dynamic functional connectivity. Meanwhile, researchers have begun investigating alcohol use disorders and migraine in terms of brain connectivity in the past 2 years. Conclusion This study illustrates a comprehensive overview of connectome research and provides researchers with critical information for understanding the recent trends and hotspots of connectomes.
Collapse
Affiliation(s)
- Yangye Yan
- Tongji University School of Medicine, Shanghai Eastern Hospital Affiliated to Tongji University, Shanghai, China
| | - Guoxin Fan
- Department of Pain Medicine, Huazhong University of Science and Technology Union Shenzhen Hospital, Shenzhen, China,School of Biomedical Engineering, School of Medicine, Guangdong Key Laboratory for Biomedical Measurements and Ultrasound Imaging, Shenzhen University, Shenzhen, China,Department of Spine Surgery, Third Affiliated Hospital, Sun Yat-sen University, Guangzhou, China
| | - Xiang Liao
- Department of Pain Medicine, Huazhong University of Science and Technology Union Shenzhen Hospital, Shenzhen, China,School of Biomedical Engineering, School of Medicine, Guangdong Key Laboratory for Biomedical Measurements and Ultrasound Imaging, Shenzhen University, Shenzhen, China,*Correspondence: Xiang Liao,
| | - Xudong Zhao
- Clinical Research Center for Mental Disorders, Chinese-German Institute of Mental Health, Shanghai Pudong New Area Mental Health Center, School of Medicine, Tongji University, Shanghai, China,Xudong Zhao,
| |
Collapse
|
5
|
TractoInferno - A large-scale, open-source, multi-site database for machine learning dMRI tractography. Sci Data 2022; 9:725. [PMID: 36433966 PMCID: PMC9700736 DOI: 10.1038/s41597-022-01833-1] [Citation(s) in RCA: 11] [Impact Index Per Article: 5.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/21/2022] [Accepted: 11/09/2022] [Indexed: 11/26/2022] Open
Abstract
TractoInferno is the world's largest open-source multi-site tractography database, including both research- and clinical-like human acquisitions, aimed specifically at machine learning tractography approaches and related ML algorithms. It provides 284 samples acquired from 3 T scanners across 6 different sites. Available data includes T1-weighted images, single-shell diffusion MRI (dMRI) acquisitions, spherical harmonics fitted to the dMRI signal, fiber ODFs, and reference streamlines for 30 delineated bundles generated using 4 tractography algorithms, as well as masks needed to run tractography algorithms. Manual quality control was additionally performed at multiple steps of the pipeline. We showcase TractoInferno by benchmarking the learn2track algorithm and 5 variations of the same recurrent neural network architecture. Creating the TractoInferno database required approximately 20,000 CPU-hours of processing power, 200 man-hours of manual QC, 3,000 GPU-hours of training baseline models, and 4 Tb of storage, to produce a final database of 350 Gb. By providing a standardized training dataset and evaluation protocol, TractoInferno is an excellent tool to address common issues in machine learning tractography.
Collapse
|
6
|
Karimi D, Gholipour A. Diffusion tensor estimation with transformer neural networks. Artif Intell Med 2022; 130:102330. [PMID: 35809969 PMCID: PMC9675900 DOI: 10.1016/j.artmed.2022.102330] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/05/2021] [Revised: 03/23/2022] [Accepted: 05/29/2022] [Indexed: 11/02/2022]
Abstract
Diffusion tensor imaging (DTI) is a widely used method for studying brain white matter development and degeneration. However, standard DTI estimation methods depend on a large number of high-quality measurements. This would require long scan times and can be particularly difficult to achieve with certain patient populations such as neonates. Here, we propose a method that can accurately estimate the diffusion tensor from only six diffusion-weighted measurements. Our method achieves this by learning to exploit the relationships between the diffusion signals and tensors in neighboring voxels. Our model is based on transformer networks, which represent the state of the art in modeling the relationship between signals in a sequence. In particular, our model consists of two such networks. The first network estimates the diffusion tensor based on the diffusion signals in a neighborhood of voxels. The second network provides more accurate tensor estimations by learning the relationships between the diffusion signals as well as the tensors estimated by the first network in neighboring voxels. Our experiments with three datasets show that our proposed method achieves highly accurate estimations of the diffusion tensor and is significantly superior to three competing methods. Estimations produced by our method with six diffusion-weighted measurements are comparable with those of standard estimation methods with 30-88 diffusion-weighted measurements. Hence, our method promises shorter scan times and more reliable assessment of brain white matter, particularly in non-cooperative patients such as neonates and infants.
Collapse
|
7
|
Zhang F, Daducci A, He Y, Schiavi S, Seguin C, Smith RE, Yeh CH, Zhao T, O'Donnell LJ. Quantitative mapping of the brain's structural connectivity using diffusion MRI tractography: A review. Neuroimage 2022; 249:118870. [PMID: 34979249 PMCID: PMC9257891 DOI: 10.1016/j.neuroimage.2021.118870] [Citation(s) in RCA: 84] [Impact Index Per Article: 42.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/15/2021] [Revised: 12/03/2021] [Accepted: 12/31/2021] [Indexed: 12/13/2022] Open
Abstract
Diffusion magnetic resonance imaging (dMRI) tractography is an advanced imaging technique that enables in vivo reconstruction of the brain's white matter connections at macro scale. It provides an important tool for quantitative mapping of the brain's structural connectivity using measures of connectivity or tissue microstructure. Over the last two decades, the study of brain connectivity using dMRI tractography has played a prominent role in the neuroimaging research landscape. In this paper, we provide a high-level overview of how tractography is used to enable quantitative analysis of the brain's structural connectivity in health and disease. We focus on two types of quantitative analyses of tractography, including: 1) tract-specific analysis that refers to research that is typically hypothesis-driven and studies particular anatomical fiber tracts, and 2) connectome-based analysis that refers to research that is more data-driven and generally studies the structural connectivity of the entire brain. We first provide a review of methodology involved in three main processing steps that are common across most approaches for quantitative analysis of tractography, including methods for tractography correction, segmentation and quantification. For each step, we aim to describe methodological choices, their popularity, and potential pros and cons. We then review studies that have used quantitative tractography approaches to study the brain's white matter, focusing on applications in neurodevelopment, aging, neurological disorders, mental disorders, and neurosurgery. We conclude that, while there have been considerable advancements in methodological technologies and breadth of applications, there nevertheless remains no consensus about the "best" methodology in quantitative analysis of tractography, and researchers should remain cautious when interpreting results in research and clinical applications.
Collapse
Affiliation(s)
- Fan Zhang
- Brigham and Women's Hospital, Harvard Medical School, Boston, USA.
| | | | - Yong He
- State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China; Beijing Key Laboratory of Brain Imaging and Connectomics, Beijing Normal University, Beijing, China; IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, China; Chinese Institute for Brain Research, Beijing, China
| | - Simona Schiavi
- Department of Computer Science, University of Verona, Verona, Italy
| | - Caio Seguin
- Melbourne Neuropsychiatry Centre, University of Melbourne and Melbourne Health, Melbourne, Australia; The University of Sydney, School of Biomedical Engineering, Sydney, Australia
| | - Robert E Smith
- The Florey Institute of Neuroscience and Mental Health, Melbourne, Australia; Florey Department of Neuroscience and Mental Health, The University of Melbourne, Melbourne, Australia
| | - Chun-Hung Yeh
- Institute for Radiological Research, Chang Gung University, Taoyuan, Taiwan; Department of Psychiatry, Chang Gung Memorial Hospital at Linkou, Taoyuan, Taiwan
| | - Tengda Zhao
- State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University, Beijing, China; Beijing Key Laboratory of Brain Imaging and Connectomics, Beijing Normal University, Beijing, China; IDG/McGovern Institute for Brain Research, Beijing Normal University, Beijing, China
| | | |
Collapse
|
8
|
Sarwar T, Ramamohanarao K, Zalesky A. A critical review of connectome validation studies. NMR IN BIOMEDICINE 2021; 34:e4605. [PMID: 34516016 DOI: 10.1002/nbm.4605] [Citation(s) in RCA: 7] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Subscribe] [Scholar Register] [Received: 04/12/2021] [Revised: 07/22/2021] [Accepted: 08/01/2021] [Indexed: 06/13/2023]
Abstract
Diffusion MRI tractography is the most widely used macroscale method for mapping connectomes in vivo. However, tractography is prone to various errors and biases, and thus tractography-derived connectomes require careful validation. Here, we critically review studies that have developed or utilized phantoms and tracer maps to validate tractography-derived connectomes, either quantitatively or qualitatively. We identify key factors impacting connectome reconstruction accuracy, including streamline seeding, propagation and filtering methods, and consider the strengths and limitations of state-of-the-art connectome phantoms and associated validation studies. These studies demonstrate the inherent limitations of current fiber orientation models and tractography algorithms and their impact on connectome reconstruction accuracy. Reconstructing connectomes with both high sensitivity and high specificity is challenging, given that some tractography methods can generate an abundance of spurious connections, while others can overlook genuine fiber bundles. We argue that streamline filtering can minimize spurious connections and potentially improve the biological plausibility of connectomes derived from tractography. We find that algorithmic choices such as the tractography seeding methodology, angular threshold, and streamline propagation method can substantially impact connectome reconstruction accuracy. Hence, careful application of tractography is necessary to reconstruct accurate connectomes. Improvements in diffusion MRI acquisition techniques will not necessarily overcome current tractography limitations without accompanying modeling and algorithmic advances.
Collapse
Affiliation(s)
- Tabinda Sarwar
- School of Computing Technologies, RMIT University, Melbourne, Victoria, Australia
| | - Kotagiri Ramamohanarao
- Department of Computing and Information Systems, The University of Melbourne, Melbourne, Victoria, Australia
| | - Andrew Zalesky
- Melbourne Neuropsychiatry Centre, The University of Melbourne, Melbourne, Victoria, Australia
- Department of Biomedical Engineering, The University of Melbourne, Melbourne, Victoria, Australia
| |
Collapse
|
9
|
Drobnjak I, Neher P, Poupon C, Sarwar T. Physical and digital phantoms for validating tractography and assessing artifacts. Neuroimage 2021; 245:118704. [PMID: 34748954 DOI: 10.1016/j.neuroimage.2021.118704] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/12/2021] [Revised: 10/01/2021] [Accepted: 11/01/2021] [Indexed: 11/17/2022] Open
Abstract
Fiber tractography is widely used to non-invasively map white-matter bundles in vivo using diffusion-weighted magnetic resonance imaging (dMRI). As it is the case for all scientific methods, proper validation is a key prerequisite for the successful application of fiber tractography, be it in the area of basic neuroscience or in a clinical setting. It is well-known that the indirect estimation of the fiber tracts from the local diffusion signal is highly ambiguous and extremely challenging. Furthermore, the validation of fiber tractography methods is hampered by the lack of a real ground truth, which is caused by the extremely complex brain microstructure that is not directly observable non-invasively and that is the basis of the huge network of long-range fiber connections in the brain that are the actual target of fiber tractography methods. As a substitute for in vivo data with a real ground truth that could be used for validation, a widely and successfully employed approach is the use of synthetic phantoms. In this work, we are providing an overview of the state-of-the-art in the area of physical and digital phantoms, answering the following guiding questions: "What are dMRI phantoms and what are they good for?", "What would the ideal phantom for validation fiber tractography look like?" and "What phantoms, phantom datasets and tools used for their creation are available to the research community?". We will further discuss the limitations and opportunities that come with the use of dMRI phantoms, and what future direction this field of research might take.
Collapse
Affiliation(s)
- Ivana Drobnjak
- Center for Medical Image Computing, Department of Computer Science, University College London, UK.
| | - Peter Neher
- Division of Medical Image Computing, German Cancer Research Center (DKFZ), Heidelberg, Germany
| | - Cyril Poupon
- BAOBAB, NeuroSpin, Commissariat à l'Energie Atomique, Institut des Sciences du Vivant Frédéric Joliot, Gif-sur-Yvette, France
| | - Tabinda Sarwar
- School of Computing Technologies, RMIT University, Australia
| |
Collapse
|
10
|
Karimi D, Jaimes C, Machado-Rivas F, Vasung L, Khan S, Warfield SK, Gholipour A. Deep learning-based parameter estimation in fetal diffusion-weighted MRI. Neuroimage 2021; 243:118482. [PMID: 34455242 PMCID: PMC8573718 DOI: 10.1016/j.neuroimage.2021.118482] [Citation(s) in RCA: 7] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/10/2021] [Revised: 08/03/2021] [Accepted: 08/17/2021] [Indexed: 11/24/2022] Open
Abstract
Diffusion-weighted magnetic resonance imaging (DW-MRI) of fetal brain is challenged by frequent fetal motion and signal to noise ratio that is much lower than non-fetal imaging. As a result, accurate and robust parameter estimation in fetal DW-MRI remains an open problem. Recently, deep learning techniques have been successfully used for DW-MRI parameter estimation in non-fetal subjects. However, none of those prior works has addressed the fetal brain because obtaining reliable fetal training data is challenging. To address this problem, in this work we propose a novel methodology that utilizes fetal scans as well as scans from prematurely-born infants. High-quality newborn scans are used to estimate accurate maps of the parameter of interest. These parameter maps are then used to generate DW-MRI data that match the measurement scheme and noise distribution that are characteristic of fetal data. In order to demonstrate the effectiveness and reliability of the proposed data generation pipeline, we used the generated data to train a convolutional neural network (CNN) to estimate color fractional anisotropy (CFA). We evaluated the trained CNN on independent sets of fetal data in terms of reconstruction accuracy, precision, and expert assessment of reconstruction quality. Results showed significantly lower reconstruction error (n=100,p<0.001) and higher reconstruction precision (n=20,p<0.001) for the proposed machine learning pipeline compared with standard estimation methods. Expert assessments on 20 fetal test scans showed significantly better overall reconstruction quality (p<0.001) and more accurate reconstruction of 11 regions of interest (p<0.001) with the proposed method.
Collapse
Affiliation(s)
- Davood Karimi
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA.
| | - Camilo Jaimes
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| | - Fedel Machado-Rivas
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| | - Lana Vasung
- Department of Pediatrics at Boston Children's Hospital, and Harvard Medical School, Boston, Massachusetts, USA
| | - Shadab Khan
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| | - Simon K Warfield
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| | - Ali Gholipour
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| |
Collapse
|
11
|
Karimi D, Vasung L, Jaimes C, Machado-Rivas F, Warfield SK, Gholipour A. Learning to estimate the fiber orientation distribution function from diffusion-weighted MRI. Neuroimage 2021; 239:118316. [PMID: 34182101 PMCID: PMC8385546 DOI: 10.1016/j.neuroimage.2021.118316] [Citation(s) in RCA: 6] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/16/2020] [Revised: 05/20/2021] [Accepted: 06/25/2021] [Indexed: 02/06/2023] Open
Abstract
Estimation of white matter fiber orientation distribution function (fODF) is the essential first step for reliable brain tractography and connectivity analysis. Most of the existing fODF estimation methods rely on sub-optimal physical models of the diffusion signal or mathematical simplifications, which can impact the estimation accuracy. In this paper, we propose a data-driven method that avoids some of these pitfalls. Our proposed method is based on a multilayer perceptron that learns to map the diffusion-weighted measurements, interpolated onto a fixed spherical grid in the q space, to the target fODF. Importantly, we also propose methods for synthesizing reliable simulated training data. We show that the model can be effectively trained with simulated or real training data. Our phantom experiments show that the proposed method results in more accurate fODF estimation and tractography than several competing methods including the multi-tensor model, Bayesian estimation, spherical deconvolution, and two other machine learning techniques. On real data, we compare our method with other techniques in terms of accuracy of estimating the ground-truth fODF. The results show that our method is more accurate than other methods, and that it performs better than the competing methods when applied to under-sampled diffusion measurements. We also compare our method with the Sparse Fascicle Model in terms of expert ratings of the accuracy of reconstruction of several commissural, projection, association, and cerebellar tracts. The results show that the tracts reconstructed with the proposed method are rated significantly higher by three independent experts. Our study demonstrates the potential of data-driven methods for improving the accuracy and robustness of fODF estimation.
Collapse
Affiliation(s)
- Davood Karimi
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA.
| | - Lana Vasung
- Department of Pediatrics, Boston Children's Hospital, and Harvard Medical School, Boston, MA, USA
| | - Camilo Jaimes
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| | - Fedel Machado-Rivas
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| | - Simon K Warfield
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| | - Ali Gholipour
- Computational Radiology Laboratory (CRL), Department of Radiology, Boston Children's Hospital, and Harvard Medical School, USA
| |
Collapse
|
12
|
Moody JF, Adluru N, Alexander AL, Field AS. The Connectomes: Methods of White Matter Tractography and Contributions of Resting State fMRI. Semin Ultrasound CT MR 2021; 42:507-522. [PMID: 34537118 DOI: 10.1053/j.sult.2021.07.007] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/18/2022]
Abstract
A comprehensive mapping of the structural and functional circuitry of the brain is a major unresolved problem in contemporary neuroimaging research. Diffusion-weighted and functional MRI have provided investigators with the capability to assess structural and functional connectivity in-vivo, driven primarily by methods of white matter tractography and resting-state fMRI, respectively. These techniques have paved the way for the construction of the functional and structural connectomes, which are quantitative representations of brain architecture as neural networks, comprised of nodes and edges. The connectomes, typically depicted as matrices or graphs, possess topological properties that inherently characterize the strength, efficiency, and organization of the connections between distinct brain regions. Graph theory, a general mathematical framework for analyzing networks, can be implemented to derive metrics from the connectomes that are sensitive to changes in brain connectivity associated with age, sex, cognitive function, and disease. These quantities can be assessed at either the global (whole brain) or local levels, allowing for the identification of distinct regional connectivity hubs and associated localized brain networks, which together serve crucial roles in establishing the structural and functional architecture of the brain. As a result, structural and functional connectomes have each been employed to study the brain circuitry underlying early brain development, neuroplasticity, developmental disorders, psychopathology, epilepsy, aging, neurodegenerative disorders, and traumatic brain injury. While these studies have yielded important insights into brain structure, function, and pathology, a precise description of the innate relationship between functional and structural networks across the brain remains unachieved. To date, connectome research has merely scratched the surface of potential clinical applications and related characterizations of brain-wide connectivity. Continued advances in diffusion and functional MRI acquisition, the delineation of functional and structural networks, and the quantification of neural network properties in specific brain regions, will be invaluable to future progress in neuroimaging science.
Collapse
Affiliation(s)
- Jason F Moody
- Department of Medical Physics, University of Wisconsin-Madison, Madison, WI; Waisman Center, University of Wisconsin-Madison, Madison, WI
| | - Nagesh Adluru
- Waisman Center, University of Wisconsin-Madison, Madison, WI; Department of Radiology, University of Wisconsin-Madison, Madison, WI
| | - Andrew L Alexander
- Department of Medical Physics, University of Wisconsin-Madison, Madison, WI; Department of Psychiatry, University of Wisconsin-Madison, Madison, WI; Waisman Center, University of Wisconsin-Madison, Madison, WI
| | - Aaron S Field
- Department of Radiology, University of Wisconsin-Madison, Madison, WI.
| |
Collapse
|
13
|
Karimi D, Vasung L, Jaimes C, Machado-Rivas F, Khan S, Warfield SK, Gholipour A. A machine learning-based method for estimating the number and orientations of major fascicles in diffusion-weighted magnetic resonance imaging. Med Image Anal 2021; 72:102129. [PMID: 34182203 PMCID: PMC8320341 DOI: 10.1016/j.media.2021.102129] [Citation(s) in RCA: 7] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/01/2020] [Revised: 05/26/2021] [Accepted: 05/28/2021] [Indexed: 12/29/2022]
Abstract
Accurate modeling of diffusion-weighted magnetic resonance imaging measurements is necessary for accurate brain connectivity analysis. Existing methods for estimating the number and orientations of fascicles in an imaging voxel either depend on non-convex optimization techniques that are sensitive to initialization and measurement noise, or are prone to predicting spurious fascicles. In this paper, we propose a machine learning-based technique that can accurately estimate the number and orientations of fascicles in a voxel. Our method can be trained with either simulated or real diffusion-weighted imaging data. Our method estimates the angle to the closest fascicle for each direction in a set of discrete directions uniformly spread on the unit sphere. This information is then processed to extract the number and orientations of fascicles in a voxel. On realistic simulated phantom data with known ground truth, our method predicts the number and orientations of crossing fascicles more accurately than several classical and machine learning methods. It also leads to more accurate tractography. On real data, our method is better than or compares favorably with other methods in terms of robustness to measurement down-sampling and also in terms of expert quality assessment of tractography results.
Collapse
Affiliation(s)
- Davood Karimi
- Department of Radiology, Boston Children's Hospital, Harvard Medical School, Boston, MA, USA.
| | - Lana Vasung
- Department of Pediatrics at Boston Children's Hospital, and Harvard Medical School, Boston, Massachusetts, USA
| | - Camilo Jaimes
- Department of Radiology, Boston Children's Hospital, Harvard Medical School, Boston, MA, USA
| | - Fedel Machado-Rivas
- Department of Radiology, Boston Children's Hospital, Harvard Medical School, Boston, MA, USA
| | - Shadab Khan
- Department of Radiology, Boston Children's Hospital, Harvard Medical School, Boston, MA, USA
| | - Simon K Warfield
- Department of Radiology, Boston Children's Hospital, Harvard Medical School, Boston, MA, USA
| | - Ali Gholipour
- Department of Radiology, Boston Children's Hospital, Harvard Medical School, Boston, MA, USA
| |
Collapse
|
14
|
Lucena O, Vos SB, Vakharia V, Duncan J, Ashkan K, Sparks R, Ourselin S. Enhancing the estimation of fiber orientation distributions using convolutional neural networks. Comput Biol Med 2021; 135:104643. [PMID: 34280774 DOI: 10.1016/j.compbiomed.2021.104643] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/15/2021] [Revised: 07/07/2021] [Accepted: 07/07/2021] [Indexed: 11/17/2022]
Abstract
Local fiber orientation distributions (FODs) can be computed from diffusion magnetic resonance imaging (dMRI). The accuracy and ability of FODs to resolve complex fiber configurations benefits from acquisition protocols that sample a high number of gradient directions, a high maximum b-value, and multiple b-values. However, acquisition time and scanners that follow these standards are limited in clinical settings, often resulting in dMRI acquired at a single shell (single b-value). In this work, we learn improved FODs from clinically acquired dMRI. We evaluate patch-based 3D convolutional neural networks (CNNs) on their ability to regress multi-shell FODs from single-shell FODs, using constrained spherical deconvolution (CSD). We evaluate U-Net and High-Resolution Network (HighResNet) 3D CNN architectures on data from the Human Connectome Project and an in-house dataset. We evaluate how well each CNN can resolve FODs 1) when training and testing on datasets with the same dMRI acquisition protocol; 2) when testing on a dataset with a different dMRI acquisition protocol than used to train the CNN; and 3) when testing on a dataset with a fewer number of gradient directions than used to train the CNN. This work is a step towards more accurate FOD estimation in time- and resource-limited clinical environments.
Collapse
Affiliation(s)
- Oeslle Lucena
- School of Biomedical Engineering and Imaging Sciences, King's College London, UK.
| | - Sjoerd B Vos
- Centre for Medical Image Computing, Department of Computer Sciences, University College London, London, UK; Neuroradiological Academic Unit, University College London Queen Square Institute of Neurology, University College London, London, UK
| | - Vejay Vakharia
- Department of Clinical and Experimental Epilepsy, University College London, UK
| | - John Duncan
- Department of Clinical and Experimental Epilepsy, University College London, UK; National Hospital for Neurology and Neurosurgery, Queen Square, UK
| | | | - Rachel Sparks
- School of Biomedical Engineering and Imaging Sciences, King's College London, UK
| | - Sebastien Ourselin
- School of Biomedical Engineering and Imaging Sciences, King's College London, UK
| |
Collapse
|
15
|
Théberge A, Desrosiers C, Descoteaux M, Jodoin PM. Track-to-Learn: A general framework for tractography with deep reinforcement learning. Med Image Anal 2021; 72:102093. [PMID: 34023562 DOI: 10.1016/j.media.2021.102093] [Citation(s) in RCA: 6] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/15/2020] [Revised: 04/15/2021] [Accepted: 04/22/2021] [Indexed: 11/26/2022]
Abstract
Diffusion MRI tractography is currently the only non-invasive tool able to assess the white-matter structural connectivity of a brain. Since its inception, it has been widely documented that tractography is prone to producing erroneous tracks while missing true positive connections. Recently, supervised learning algorithms have been proposed to learn the tracking procedure implicitly from data, without relying on anatomical priors. However, these methods rely on curated streamlines that are very hard to obtain. To remove the need for such data but still leverage the expressiveness of neural networks, we introduce Track-To-Learn: A general framework to pose tractography as a deep reinforcement learning problem. Deep reinforcement learning is a type of machine learning that does not depend on ground-truth data but rather on the concept of "reward". We implement and train algorithms to maximize returns from a reward function based on the alignment of streamlines with principal directions extracted from diffusion data. We show competitive results on known data and little loss of performance when generalizing to new, unseen data, compared to prior machine learning-based tractography algorithms. To the best of our knowledge, this is the first successful use of deep reinforcement learning for tractography.
Collapse
Affiliation(s)
- Antoine Théberge
- Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, CA, J1K 2R1.
| | - Christian Desrosiers
- Département de génie logiciel et des TI, École de technologie supérieure, Montréal, QC, CA, H3C 1K3
| | - Maxime Descoteaux
- Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, CA, J1K 2R1
| | - Pierre-Marc Jodoin
- Faculté des Sciences, Université de Sherbrooke, Sherbrooke, QC, CA, J1K 2R1
| |
Collapse
|
16
|
Reymbaut A, Critchley J, Durighel G, Sprenger T, Sughrue M, Bryskhe K, Topgaard D. Toward nonparametric diffusion- T1 characterization of crossing fibers in the human brain. Magn Reson Med 2021; 85:2815-2827. [PMID: 33301195 PMCID: PMC7898694 DOI: 10.1002/mrm.28604] [Citation(s) in RCA: 16] [Impact Index Per Article: 5.3] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/14/2020] [Revised: 10/26/2020] [Accepted: 10/27/2020] [Indexed: 12/24/2022]
Abstract
PURPOSE To estimate T 1 for each distinct fiber population within voxels containing multiple brain tissue types. METHODS A diffusion- T 1 correlation experiment was carried out in an in vivo human brain using tensor-valued diffusion encoding and multiple repetition times. The acquired data were inverted using a Monte Carlo algorithm that retrieves nonparametric distributions P ( D , R 1 ) of diffusion tensors and longitudinal relaxation rates R 1 = 1 / T 1 . Orientation distribution functions (ODFs) of the highly anisotropic components of P ( D , R 1 ) were defined to visualize orientation-specific diffusion-relaxation properties. Finally, Monte Carlo density-peak clustering (MC-DPC) was performed to quantify fiber-specific features and investigate microstructural differences between white matter fiber bundles. RESULTS Parameter maps corresponding to P ( D , R 1 ) 's statistical descriptors were obtained, exhibiting the expected R 1 contrast between brain tissue types. Our ODFs recovered local orientations consistent with the known anatomy and indicated differences in R 1 between major crossing fiber bundles. These differences, confirmed by MC-DPC, were in qualitative agreement with previous model-based works but seem biased by the limitations of our current experimental setup. CONCLUSIONS Our Monte Carlo framework enables the nonparametric estimation of fiber-specific diffusion- T 1 features, thereby showing potential for characterizing developmental or pathological changes in T 1 within a given fiber bundle, and for investigating interbundle T 1 differences.
Collapse
Affiliation(s)
- Alexis Reymbaut
- Department of Physical ChemistryLund UniversityLundSweden
- Random Walk Imaging ABLundSweden
| | | | | | - Tim Sprenger
- Karolinska InstituteStockholmSweden
- GE HealthcareStockholmSweden
| | | | | | - Daniel Topgaard
- Department of Physical ChemistryLund UniversityLundSweden
- Random Walk Imaging ABLundSweden
| |
Collapse
|
17
|
Abstract
Tractography is an important technique that allows the in vivo reconstruction of structural connections in the brain using diffusion MRI. Although tracking algorithms have improved during the last two decades, results of validation studies and international challenges warn about the reliability of tractography and point out the need for improved algorithms. In propagation-based tracking, connections have traditionally been modeled as piece-wise linear segments. In this work, we propose a novel propagation-based tracker that is capable of generating geometrically smooth ( C1 ) curves using parallel transport frames. Notably, our approach does not increase the complexity of the propagation problem that remains two-dimensional. Moreover, our tracker has a novel mechanism to reduce noise related propagation errors by incorporating topographic regularity of connections, a neuroanatomic property of many brain pathways. We ran extensive experiments and compared our approach against deterministic and other probabilistic algorithms. Our experiments on FiberCup and ISMRM 2015 challenge datasets as well as on 56 subjects of the Human Connectome Project show highly promising results both visually and quantitatively. Open-source implementations of the algorithm are shared publicly.
Collapse
|
18
|
Blesa M, Galdi P, Cox SR, Sullivan G, Stoye DQ, Lamb GJ, Quigley AJ, Thrippleton MJ, Escudero J, Bastin ME, Smith KM, Boardman JP. Hierarchical Complexity of the Macro-Scale Neonatal Brain. Cereb Cortex 2020; 31:2071-2084. [PMID: 33280008 PMCID: PMC7945030 DOI: 10.1093/cercor/bhaa345] [Citation(s) in RCA: 7] [Impact Index Per Article: 1.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/03/2023] Open
Abstract
The human adult structural connectome has a rich nodal hierarchy, with highly diverse connectivity patterns aligned to the diverse range of functional specializations in the brain. The emergence of this hierarchical complexity in human development is unknown. Here, we substantiate the hierarchical tiers and hierarchical complexity of brain networks in the newborn period, assess correspondences with hierarchical complexity in adulthood, and investigate the effect of preterm birth, a leading cause of atypical brain development and later neurocognitive impairment, on hierarchical complexity. We report that neonatal and adult structural connectomes are both composed of distinct hierarchical tiers and that hierarchical complexity is greater in term born neonates than in preterms. This is due to diversity of connectivity patterns of regions within the intermediate tiers, which consist of regions that underlie sensorimotor processing and its integration with cognitive information. For neonates and adults, the highest tier (hub regions) is ordered, rather than complex, with more homogeneous connectivity patterns in structural hubs. This suggests that the brain develops first a more rigid structure in hub regions allowing for the development of greater and more diverse functional specialization in lower level regions, while connectivity underpinning this diversity is dysmature in infants born preterm.
Collapse
Affiliation(s)
- Manuel Blesa
- MRC Centre for Reproductive Health, University of Edinburgh, Edinburgh EH16 4TJ, UK
| | - Paola Galdi
- MRC Centre for Reproductive Health, University of Edinburgh, Edinburgh EH16 4TJ, UK
| | - Simon R Cox
- Lothian Birth Cohorts Group, Department of Psychology, University of Edinburgh, Edinburgh EH8 9JZ, UK
| | - Gemma Sullivan
- MRC Centre for Reproductive Health, University of Edinburgh, Edinburgh EH16 4TJ, UK
| | - David Q Stoye
- MRC Centre for Reproductive Health, University of Edinburgh, Edinburgh EH16 4TJ, UK
| | - Gillian J Lamb
- MRC Centre for Reproductive Health, University of Edinburgh, Edinburgh EH16 4TJ, UK
| | - Alan J Quigley
- Department of Radiology, Royal Hospital for Sick Children, Edinburgh EH9 1LF, UK
| | - Michael J Thrippleton
- Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh EH16 4SB, UK.,Edinburgh Imaging, University of Edinburgh, Edinburgh EH16 4SB, UK
| | - Javier Escudero
- School of Engineering, Institute for Digital Communications, University of Edinburgh, Edinburgh EH9 3FG, UK
| | - Mark E Bastin
- Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh EH16 4SB, UK
| | - Keith M Smith
- Usher Institute, University of Edinburgh, Edinburgh EH16 4UX, UK.,Health Data Research UK, London NW1 2BE, UK
| | - James P Boardman
- MRC Centre for Reproductive Health, University of Edinburgh, Edinburgh EH16 4TJ, UK.,Centre for Clinical Brain Sciences, University of Edinburgh, Edinburgh EH16 4SB, UK
| |
Collapse
|
19
|
Wegmayr V, Buhmann JM. Entrack: Probabilistic Spherical Regression with Entropy Regularization for Fiber Tractography. Int J Comput Vis 2020. [DOI: 10.1007/s11263-020-01384-1] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/23/2022]
Abstract
AbstractWhite matter tractography, based on diffusion-weighted magnetic resonance images, is currently the only available in vivo method to gather information on the structural brain connectivity. The low resolution of diffusion MRI data suggests to employ probabilistic methods for streamline reconstruction, i.e., for fiber crossings. We propose a general probabilistic model for spherical regression based on the Fisher-von-Mises distribution, which efficiently estimates maximum entropy posteriors of local streamline directions with machine learning methods. The optimal precision of posteriors for streamlines is determined by an information-theoretic technique, the expected log-posterior agreement concept. It relies on the requirement that the posterior distributions of streamlines, inferred on retest measurements of the same subject, should yield stable results within the precision determined by the noise level of the data source.
Collapse
|
20
|
Touska P, Connor S. Imaging of the temporal bone. Clin Radiol 2020; 75:658-674. [DOI: 10.1016/j.crad.2020.06.013] [Citation(s) in RCA: 6] [Impact Index Per Article: 1.5] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/01/2020] [Accepted: 06/22/2020] [Indexed: 12/20/2022]
|
21
|
Sarwar T, Seguin C, Ramamohanarao K, Zalesky A. Towards deep learning for connectome mapping: A block decomposition framework. Neuroimage 2020; 212:116654. [PMID: 32068163 DOI: 10.1016/j.neuroimage.2020.116654] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/29/2019] [Revised: 02/09/2020] [Accepted: 02/13/2020] [Indexed: 12/13/2022] Open
Abstract
We propose a new framework to map structural connectomes using deep learning and diffusion MRI. We show that our framework not only enables connectome mapping with a convolutional neural network (CNN), but can also be straightforwardly incorporated into conventional connectome mapping pipelines to enhance accuracy. Our framework involves decomposing the entire brain volume into overlapping blocks. Blocks are sufficiently small to ensure that a CNN can be efficiently trained to predict each block's internal connectivity architecture. We develop a block stitching algorithm to rebuild the full brain volume from these blocks and thereby map end-to-end connectivity matrices. To evaluate our block decomposition and stitching (BDS) framework independent of CNN performance, we first map each block's internal connectivity using conventional streamline tractography. Performance is evaluated using simulated diffusion MRI data generated from numerical connectome phantoms with known ground truth connectivity. Due to the redundancy achieved by allowing blocks to overlap, we find that our block decomposition and stitching steps per se can enhance the accuracy of probabilistic and deterministic tractography algorithms by up to 20-30%. Moreover, we demonstrate that our framework can improve the strength of structure-function coupling between in vivo diffusion and functional MRI data. We find that structural brain networks mapped with deep learning correlate more strongly with functional brain networks (r = 0.45) than those mapped with conventional tractography (r = 0.36). In conclusion, our BDS framework not only enables connectome mapping with deep learning, but its two constituent steps can be straightforwardly incorporated as part of conventional connectome mapping pipelines to enhance accuracy.
Collapse
Affiliation(s)
- Tabinda Sarwar
- Department of Computing and Information Systems, The University of Melbourne, Victoria, 3010, Australia.
| | - Caio Seguin
- Melbourne Neuropsychiatry Centre, The University of Melbourne, Victoria, 3010, Australia
| | - Kotagiri Ramamohanarao
- Department of Computing and Information Systems, The University of Melbourne, Victoria, 3010, Australia
| | - Andrew Zalesky
- Department of Biomedical Engineering, The University of Melbourne, Victoria, 3010, Australia; Melbourne Neuropsychiatry Centre, The University of Melbourne, Victoria, 3010, Australia.
| |
Collapse
|
22
|
Jacquesson T, Yeh FC, Panesar S, Barrios J, Attyé A, Frindel C, Cotton F, Gardner P, Jouanneau E, Fernandez-Miranda JC. Full tractography for detecting the position of cranial nerves in preoperative planning for skull base surgery: technical note. J Neurosurg 2020; 132:1642-1652. [PMID: 31003214 DOI: 10.3171/2019.1.jns182638] [Citation(s) in RCA: 11] [Impact Index Per Article: 2.8] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/11/2018] [Accepted: 01/28/2019] [Indexed: 11/06/2022]
Abstract
OBJECTIVE Diffusion imaging tractography has allowed the in vivo description of brain white matter. One of its applications is preoperative planning for brain tumor resection. Due to a limited spatial and angular resolution, it is difficult for fiber tracking to delineate fiber crossing areas and small-scale structures, in particular brainstem tracts and cranial nerves. New methods are being developed but these involve extensive multistep tractography pipelines including the patient-specific design of multiple regions of interest (ROIs). The authors propose a new practical full tractography method that could be implemented in routine presurgical planning for skull base surgery. METHODS A Philips MRI machine provided diffusion-weighted and anatomical sequences for 2 healthy volunteers and 2 skull base tumor patients. Tractography of the full brainstem, the cerebellum, and cranial nerves was performed using the software DSI Studio, generalized-q-sampling reconstruction, orientation distribution function (ODF) of fibers, and a quantitative anisotropy-based generalized deterministic algorithm. No ROI or extensive manual filtering of spurious fibers was used. Tractography rendering was displayed in a tridimensional space with directional color code. This approach was also tested on diffusion data from the Human Connectome Project (HCP) database. RESULTS The brainstem, the cerebellum, and the cisternal segments of most cranial nerves were depicted in all participants. In cases of skull base tumors, the tridimensional rendering permitted the visualization of the whole anatomical environment and cranial nerve displacement, thus helping the surgical strategy. CONCLUSIONS As opposed to classical ROI-based methods, this novel full tractography approach could enable routine enhanced surgical planning or brain imaging for skull base tumors.
Collapse
Affiliation(s)
- Timothee Jacquesson
- 1Department of Neurological Surgery, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
- 2Skull Base Multi-Disciplinary Unit, Department of Neurosurgery B, Neurological Hospital Pierre Wertheimer, Hospices Civils de Lyon
- 3CREATIS Laboratory CNRS UMR5220, Inserm U1206, INSA-Lyon, University of Lyon 1
| | - Fang-Chang Yeh
- 1Department of Neurological Surgery, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
| | - Sandip Panesar
- 4Department of Neurosurgery, Stanford University Medical Center, Stanford, California
| | - Jessica Barrios
- 1Department of Neurological Surgery, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
| | - Arnaud Attyé
- 5Department of Neuroradiology and MRI, Grenoble University Hospital, Grenoble, France; and
| | - Carole Frindel
- 3CREATIS Laboratory CNRS UMR5220, Inserm U1206, INSA-Lyon, University of Lyon 1
| | - Francois Cotton
- 3CREATIS Laboratory CNRS UMR5220, Inserm U1206, INSA-Lyon, University of Lyon 1
- 6Department of Radiology, Lyon Sud Hospital, Hospices Civils de Lyon, Lyon
| | - Paul Gardner
- 1Department of Neurological Surgery, University of Pittsburgh Medical Center, Pittsburgh, Pennsylvania
| | - Emmanuel Jouanneau
- 2Skull Base Multi-Disciplinary Unit, Department of Neurosurgery B, Neurological Hospital Pierre Wertheimer, Hospices Civils de Lyon
| | | |
Collapse
|
23
|
Vanderweyen DC, Theaud G, Sidhu J, Rheault F, Sarubbo S, Descoteaux M, Fortin D. The role of diffusion tractography in refining glial tumor resection. Brain Struct Funct 2020; 225:1413-1436. [PMID: 32180019 DOI: 10.1007/s00429-020-02056-z] [Citation(s) in RCA: 22] [Impact Index Per Article: 5.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/08/2019] [Accepted: 02/28/2020] [Indexed: 12/14/2022]
Abstract
Primary brain tumors are notoriously hard to resect surgically. Due to their infiltrative nature, finding the optimal resection boundary without damaging healthy tissue can be challenging. One potential tool to help make this decision is diffusion-weighted magnetic resonance imaging (dMRI) tractography. dMRI exploits the diffusion of water molecule along axons to generate a 3D modelization of the white matter bundles in the brain. This feature is particularly useful to visualize how a tumor affects its surrounding white matter and plan a surgical path. This paper reviews the different ways in which dMRI can be used to improve brain tumor resection, its benefits and also its limitations. We expose surgical tools that can be paired with dMRI to improve its impact on surgical outcome, such as loading the 3D tractography in the neuronavigation system and direct electrical stimulation to validate the position of the white matter bundles of interest. We also review articles validating dMRI findings using other anatomical investigation techniques, such as postmortem dissections, manganese-enhanced MRI, electrophysiological stimulations, and phantom studies with known ground truth. We will be discussing the areas of the brain where dMRI performs well and where the future challenges are. We will conclude this review with suggestions and take home messages for neurosurgeons, tractographers, and vendors for advancing the field and on how to benefit from tractography's use in clinical practice.
Collapse
Affiliation(s)
- Davy Charles Vanderweyen
- Department of Surgery, Division of Neurosurgery, Faculty of Medicine, University of Sherbrooke, 3001 12 Ave N, Sherbrooke, QC, J1H 5H3, Canada.
| | - Guillaume Theaud
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, University of Sherbrooke, 2500 Boulevard Université, Sherbrooke, QC, J1K2R1, Canada
| | - Jasmeen Sidhu
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, University of Sherbrooke, 2500 Boulevard Université, Sherbrooke, QC, J1K2R1, Canada
| | - François Rheault
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, University of Sherbrooke, 2500 Boulevard Université, Sherbrooke, QC, J1K2R1, Canada
| | - Silvio Sarubbo
- Division of Neurosurgery, Emergency Area, Structural and Functional Connectivity Lab Project, "S. Chiara" Hospital, Azienda Provinciale Per I Servizi Sanitari (APSS), Trento, Italy
| | - Maxime Descoteaux
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, University of Sherbrooke, 2500 Boulevard Université, Sherbrooke, QC, J1K2R1, Canada
| | - David Fortin
- Department of Surgery, Division of Neurosurgery, Faculty of Medicine, University of Sherbrooke, 3001 12 Ave N, Sherbrooke, QC, J1H 5H3, Canada
| |
Collapse
|
24
|
Witulla B, Goerig N, Putz F, Frey B, Engelhorn T, Dörfler A, Uder M, Fietkau R, Bert C, Laun FB. On PTV definition for glioblastoma based on fiber tracking of diffusion tensor imaging data. PLoS One 2020; 15:e0227146. [PMID: 31905221 PMCID: PMC6944332 DOI: 10.1371/journal.pone.0227146] [Citation(s) in RCA: 4] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/16/2019] [Accepted: 12/11/2019] [Indexed: 01/20/2023] Open
Abstract
Radiotherapy (RT) is commonly applied for the treatment of glioblastoma multiforme (GBM). Following the planning target volume (PTV) definition procedure standardized in guidelines, a 20% risk of missing non-local recurrences is present. Purpose of this study was to evaluate whether diffusion tensor imaging (DTI)-based fiber tracking may be beneficial for PTV definition taking into account the prediction of distant recurrences. 56 GBM patients were examined with magnetic resonance imaging (MRI) including DTI performed before RT after resection of the primary tumor. Follow-up MRIs were acquired in three month intervals. For the seven patients with a distant recurrence, fiber tracking was performed with three algorithms and it was evaluated whether connections existed from the primary tumor region to the distant recurrence. It depended strongly on the used tracking algorithm and the used tracking parameters whether a connection was observed. Most of the connections were weak and thus not usable for PTV definition. Only in one of the seven patients with a recurring tumor, a clear connection was present. It seems unlikely that DTI-based fiber tracking can be beneficial for predicting distant recurrences in the planning of PTVs for glioblastoma multiforme.
Collapse
Affiliation(s)
- Barbara Witulla
- Department of Radiation Oncology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Nicole Goerig
- Department of Radiation Oncology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Florian Putz
- Department of Radiation Oncology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Benjamin Frey
- Department of Radiation Oncology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Tobias Engelhorn
- Department of Neuroradiology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Arnd Dörfler
- Department of Neuroradiology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Michael Uder
- Institute of Radiology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Rainer Fietkau
- Department of Radiation Oncology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| | - Christoph Bert
- Department of Radiation Oncology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
- * E-mail:
| | - Frederik Bernd Laun
- Institute of Radiology, Universitätsklinikum Erlangen, Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
| |
Collapse
|
25
|
Tractography and machine learning: Current state and open challenges. Magn Reson Imaging 2019; 64:37-48. [PMID: 31078615 DOI: 10.1016/j.mri.2019.04.013] [Citation(s) in RCA: 32] [Impact Index Per Article: 6.4] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/04/2019] [Revised: 04/29/2019] [Accepted: 04/29/2019] [Indexed: 12/28/2022]
|
26
|
Schilling KG, Daducci A, Maier-Hein K, Poupon C, Houde JC, Nath V, Anderson AW, Landman BA, Descoteaux M. Challenges in diffusion MRI tractography - Lessons learned from international benchmark competitions. Magn Reson Imaging 2019; 57:194-209. [PMID: 30503948 PMCID: PMC6331218 DOI: 10.1016/j.mri.2018.11.014] [Citation(s) in RCA: 70] [Impact Index Per Article: 14.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/08/2018] [Accepted: 11/17/2018] [Indexed: 12/13/2022]
Abstract
Diffusion MRI (dMRI) fiber tractography has become a pillar of the neuroimaging community due to its ability to noninvasively map the structural connectivity of the brain. Despite widespread use in clinical and research domains, these methods suffer from several potential drawbacks or limitations. Thus, validating the accuracy and reproducibility of techniques is critical for sound scientific conclusions and effective clinical outcomes. Towards this end, a number of international benchmark competitions, or "challenges", has been organized by the diffusion MRI community in order to investigate the reliability of the tractography process by providing a platform to compare algorithms and results in a fair manner, and evaluate common and emerging algorithms in an effort to advance the state of the field. In this paper, we summarize the lessons from a decade of challenges in tractography, and give perspective on the past, present, and future "challenges" that the field of diffusion tractography faces.
Collapse
Affiliation(s)
- Kurt G Schilling
- Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States of America.
| | | | - Klaus Maier-Hein
- Division of Medical Image Computing, German Cancer Research Center (DKFZ), Heidelberg 69120, Germany
| | - Cyril Poupon
- Neurospin, Frédéric Joliot Life Sciences Institute, CEA, Gif-sur-Yvette, France
| | - Jean-Christophe Houde
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, Université de Sherbrooke, Québec, Canada
| | - Vishwesh Nath
- Electrical Engineering & Computer Science, Vanderbilt University, Nashville, TN, United States of America
| | - Adam W Anderson
- Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States of America; Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States of America
| | - Bennett A Landman
- Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, United States of America; Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, United States of America; Department of Electrical Engineering, Vanderbilt University, Nashville, TN, United States of America
| | - Maxime Descoteaux
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, Université de Sherbrooke, Québec, Canada
| |
Collapse
|
27
|
Straathof M, Sinke MRT, Dijkhuizen RM, Otte WM. A systematic review on the quantitative relationship between structural and functional network connectivity strength in mammalian brains. J Cereb Blood Flow Metab 2019; 39:189-209. [PMID: 30375267 PMCID: PMC6360487 DOI: 10.1177/0271678x18809547] [Citation(s) in RCA: 61] [Impact Index Per Article: 12.2] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 05/02/2018] [Accepted: 09/28/2018] [Indexed: 12/29/2022]
Abstract
The mammalian brain is composed of densely connected and interacting regions, which form structural and functional networks. An improved understanding of the structure-function relation is crucial to understand the structural underpinnings of brain function and brain plasticity after injury. It is currently unclear how functional connectivity strength relates to structural connectivity strength. We obtained an overview of recent papers that report on correspondences between quantitative functional and structural connectivity measures in the mammalian brain. We included network studies in which functional connectivity was measured with resting-state fMRI, and structural connectivity with either diffusion-weighted MRI or neuronal tract tracers. Twenty-seven of the 28 included studies showed a positive structure-function relationship. Large inter-study variations were found comparing functional connectivity strength with either quantitative diffusion-based (correlation coefficient (r) ranges: 0.18-0.82) or neuronal tracer-based structural connectivity measures (r = 0.24-0.74). Two functional datasets demonstrated lower structure-function correlations with neuronal tracer-based (r = 0.22 and r = 0.30) than with diffusion-based measures (r = 0.49 and r = 0.65). The robust positive quantitative structure-function relationship supports the hypothesis that structural connectivity provides the hardware from which functional connectivity emerges. However, methodological differences between the included studies complicate the comparison across studies, which emphasize the need for validation and standardization in brain structure-function studies.
Collapse
Affiliation(s)
- Milou Straathof
- Biomedical MR Imaging and Spectroscopy Group, Center for Image Sciences, University Medical Center Utrecht and Utrecht University, Utrecht, the Netherlands
| | - Michel RT Sinke
- Biomedical MR Imaging and Spectroscopy Group, Center for Image Sciences, University Medical Center Utrecht and Utrecht University, Utrecht, the Netherlands
| | - Rick M Dijkhuizen
- Biomedical MR Imaging and Spectroscopy Group, Center for Image Sciences, University Medical Center Utrecht and Utrecht University, Utrecht, the Netherlands
| | - Willem M Otte
- Biomedical MR Imaging and Spectroscopy Group, Center for Image Sciences, University Medical Center Utrecht and Utrecht University, Utrecht, the Netherlands
- Department of Pediatric Neurology, Brain Center Rudolf Magnus, University Medical Center Utrecht, Utrecht, the Netherlands
| |
Collapse
|
28
|
Ravi D, Ghavami N, Alexander DC, Ianus A. Current Applications and Future Promises of Machine Learning in Diffusion MRI. COMPUTATIONAL DIFFUSION MRI 2019. [DOI: 10.1007/978-3-030-05831-9_9] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 12/12/2022]
|
29
|
Schilling KG, Nath V, Hansen C, Parvathaneni P, Blaber J, Gao Y, Neher P, Aydogan DB, Shi Y, Ocampo-Pineda M, Schiavi S, Daducci A, Girard G, Barakovic M, Rafael-Patino J, Romascano D, Rensonnet G, Pizzolato M, Bates A, Fischi E, Thiran JP, Canales-Rodríguez EJ, Huang C, Zhu H, Zhong L, Cabeen R, Toga AW, Rheault F, Theaud G, Houde JC, Sidhu J, Chamberland M, Westin CF, Dyrby TB, Verma R, Rathi Y, Irfanoglu MO, Thomas C, Pierpaoli C, Descoteaux M, Anderson AW, Landman BA. Limits to anatomical accuracy of diffusion tractography using modern approaches. Neuroimage 2018; 185:1-11. [PMID: 30317017 DOI: 10.1016/j.neuroimage.2018.10.029] [Citation(s) in RCA: 144] [Impact Index Per Article: 24.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/13/2018] [Revised: 09/14/2018] [Accepted: 10/09/2018] [Indexed: 12/12/2022] Open
Abstract
Diffusion MRI fiber tractography is widely used to probe the structural connectivity of the brain, with a range of applications in both clinical and basic neuroscience. Despite widespread use, tractography has well-known pitfalls that limits the anatomical accuracy of this technique. Numerous modern methods have been developed to address these shortcomings through advances in acquisition, modeling, and computation. To test whether these advances improve tractography accuracy, we organized the 3-D Validation of Tractography with Experimental MRI (3D-VoTEM) challenge at the ISBI 2018 conference. We made available three unique independent tractography validation datasets - a physical phantom and two ex vivo brain specimens - resulting in 176 distinct submissions from 9 research groups. By comparing results over a wide range of fiber complexities and algorithmic strategies, this challenge provides a more comprehensive assessment of tractography's inherent limitations than has been reported previously. The central results were consistent across all sub-challenges in that, despite advances in tractography methods, the anatomical accuracy of tractography has not dramatically improved in recent years. Taken together, our results independently confirm findings from decades of tractography validation studies, demonstrate inherent limitations in reconstructing white matter pathways using diffusion MRI data alone, and highlight the need for alternative or combinatorial strategies to accurately map the fiber pathways of the brain.
Collapse
Affiliation(s)
- Kurt G Schilling
- Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, USA.
| | - Vishwesh Nath
- Electrical Engineering & Computer Science, Vanderbilt University, Nashville, TN, USA
| | - Colin Hansen
- Department of Electrical Engineering, Vanderbilt University, Nashville, TN, USA
| | | | - Justin Blaber
- Department of Electrical Engineering, Vanderbilt University, Nashville, TN, USA
| | - Yurui Gao
- Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, USA
| | - Peter Neher
- Division of Medical Image Computing, German Cancer Research Center, Heidelberg, Germany
| | - Dogu Baran Aydogan
- Laboratory of Neuro Imaging (LONI), USC Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA
| | - Yonggang Shi
- Laboratory of Neuro Imaging (LONI), USC Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA
| | | | - Simona Schiavi
- Computer Science Department, University of Verona, Verona, Italy
| | | | - Gabriel Girard
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
| | - Muhamed Barakovic
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
| | - Jonathan Rafael-Patino
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
| | - David Romascano
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
| | - Gaëtan Rensonnet
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
| | - Marco Pizzolato
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
| | - Alice Bates
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
| | - Elda Fischi
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland
| | - Jean-Philippe Thiran
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland; Radiology Department, Centre Hospitalier Universitaire Vaudois and University of Lausanne, Lausanne, Switzerland
| | - Erick J Canales-Rodríguez
- Signal Processing Lab (LTS5), École Polytechnique Fédérale de Lausanne, Lausanne, Switzerland; Radiology Department, Centre Hospitalier Universitaire Vaudois and University of Lausanne, Lausanne, Switzerland
| | - Chao Huang
- Department of Biostatistics, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
| | - Hongtu Zhu
- Department of Biostatistics, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
| | - Liming Zhong
- Guangdong Provincial Key Laboratory of Medical Image Processing, School of Biomedical Engineering, Southern Medical University, Guangzhou, China
| | - Ryan Cabeen
- Laboratory of Neuro Imaging (LONI), USC Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA
| | - Arthur W Toga
- Laboratory of Neuro Imaging (LONI), USC Stevens Neuroimaging and Informatics Institute, Keck School of Medicine, University of Southern California, Los Angeles, CA, USA
| | - Francois Rheault
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, Université de Sherbrooke, Canada
| | - Guillaume Theaud
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, Université de Sherbrooke, Canada
| | - Jean-Christophe Houde
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, Université de Sherbrooke, Canada
| | - Jasmeen Sidhu
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, Université de Sherbrooke, Canada
| | - Maxime Chamberland
- Cardiff University, Brain Research Imaging Centre, School of Psychology, Cardiff, UK
| | | | - Tim B Dyrby
- Danish Research Centre for Magnetic Resonance, Centre for Functional and Diagnostic Imaging and Research, Copenhagen University Hospital Hvidovre, Hvidovre, Denmark; Department of Applied Mathematics and Computer Science, Technical University of Denmark, Kongens Lyngby, Denmark
| | - Ragini Verma
- Center for Biomedical Image Computing and Analytics, University of Pennsylvania, Philadelphia, PA, USA
| | - Yogesh Rathi
- Brigham and Women's Hospital, Harvard Medical School, USA
| | - M Okan Irfanoglu
- National Institute of Biomedical Imaging and Bioengineering, NIH, Bethesda, MD, USA
| | - Cibu Thomas
- Section on Learning and Plasticity, Laboratory of Brain and Cognition, NIMH, Bethesda, MD, USA
| | - Carlo Pierpaoli
- National Institute of Biomedical Imaging and Bioengineering, NIH, Bethesda, MD, USA
| | - Maxime Descoteaux
- Sherbrooke Connectivity Imaging Lab (SCIL), Computer Science Department, Université de Sherbrooke, Canada
| | - Adam W Anderson
- Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, USA; Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, USA
| | - Bennett A Landman
- Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, USA; Department of Biomedical Engineering, Vanderbilt University, Nashville, TN, USA; Department of Electrical Engineering, Vanderbilt University, Nashville, TN, USA
| |
Collapse
|
30
|
Maier-Hein KH, Neher PF, Houde JC, Côté MA, Garyfallidis E, Zhong J, Chamberland M, Yeh FC, Lin YC, Ji Q, Reddick WE, Glass JO, Chen DQ, Feng Y, Gao C, Wu Y, Ma J, He R, Li Q, Westin CF, Deslauriers-Gauthier S, González JOO, Paquette M, St-Jean S, Girard G, Rheault F, Sidhu J, Tax CMW, Guo F, Mesri HY, Dávid S, Froeling M, Heemskerk AM, Leemans A, Boré A, Pinsard B, Bedetti C, Desrosiers M, Brambati S, Doyon J, Sarica A, Vasta R, Cerasa A, Quattrone A, Yeatman J, Khan AR, Hodges W, Alexander S, Romascano D, Barakovic M, Auría A, Esteban O, Lemkaddem A, Thiran JP, Cetingul HE, Odry BL, Mailhe B, Nadar MS, Pizzagalli F, Prasad G, Villalon-Reina JE, Galvis J, Thompson PM, Requejo FDS, Laguna PL, Lacerda LM, Barrett R, Dell'Acqua F, Catani M, Petit L, Caruyer E, Daducci A, Dyrby TB, Holland-Letz T, Hilgetag CC, Stieltjes B, Descoteaux M. The challenge of mapping the human connectome based on diffusion tractography. Nat Commun 2017; 8:1349. [PMID: 29116093 PMCID: PMC5677006 DOI: 10.1038/s41467-017-01285-x] [Citation(s) in RCA: 721] [Impact Index Per Article: 103.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/21/2016] [Accepted: 09/01/2017] [Indexed: 01/14/2023] Open
Abstract
Tractography based on non-invasive diffusion imaging is central to the study of human brain connectivity. To date, the approach has not been systematically validated in ground truth studies. Based on a simulated human brain data set with ground truth tracts, we organized an open international tractography challenge, which resulted in 96 distinct submissions from 20 research groups. Here, we report the encouraging finding that most state-of-the-art algorithms produce tractograms containing 90% of the ground truth bundles (to at least some extent). However, the same tractograms contain many more invalid than valid bundles, and half of these invalid bundles occur systematically across research groups. Taken together, our results demonstrate and confirm fundamental ambiguities inherent in tract reconstruction based on orientation information alone, which need to be considered when interpreting tractography and connectivity results. Our approach provides a novel framework for estimating reliability of tractography and encourages innovation to address its current limitations. Though tractography is widely used, it has not been systematically validated. Here, authors report results from 20 groups showing that many tractography algorithms produce both valid and invalid bundles.
Collapse
Affiliation(s)
- Klaus H Maier-Hein
- Division of Medical Image Computing, German Cancer Research Center (DKFZ), Heidelberg, 69120, Germany.
| | - Peter F Neher
- Division of Medical Image Computing, German Cancer Research Center (DKFZ), Heidelberg, 69120, Germany
| | - Jean-Christophe Houde
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada
| | - Marc-Alexandre Côté
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada
| | - Eleftherios Garyfallidis
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada.,Department of Intelligent Systems Engineering, School of Informatics and Computing, Indiana University, Bloomington, IN 47408, USA
| | - Jidan Zhong
- Krembil Research Institute, University Health Network, Toronto, Canada, M5G 2C4
| | - Maxime Chamberland
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada
| | - Fang-Cheng Yeh
- Department of Neurological Surgery, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213, USA
| | - Ying-Chia Lin
- IMT-Institute for Advanced Studies, Lucca, 55100, Italy
| | - Qing Ji
- Department of Diagnostic Imaging, St. Jude Children's Research Hospital, Memphis, TN 38105, USA
| | - Wilburn E Reddick
- Department of Diagnostic Imaging, St. Jude Children's Research Hospital, Memphis, TN 38105, USA
| | - John O Glass
- Department of Diagnostic Imaging, St. Jude Children's Research Hospital, Memphis, TN 38105, USA
| | - David Qixiang Chen
- University of Toronto Institute of Medical Science, Toronto, Canada, M5S 1A8
| | - Yuanjing Feng
- Institute of Information Processing and Automation, Zhejiang University of Technology, Hangzhou, 310023, Zhejiang, China
| | - Chengfeng Gao
- Institute of Information Processing and Automation, Zhejiang University of Technology, Hangzhou, 310023, Zhejiang, China
| | - Ye Wu
- Institute of Information Processing and Automation, Zhejiang University of Technology, Hangzhou, 310023, Zhejiang, China
| | - Jieyan Ma
- United Imaging Healthcare Co., Shanghai, 201807, China
| | - Renjie He
- United Imaging Healthcare Co., Shanghai, 201807, China
| | - Qiang Li
- United Imaging Healthcare Co., Shanghai, 201807, China.,Shanghai Advanced Research Institute, Shanghai, 201210, China
| | - Carl-Fredrik Westin
- Laboratory of Mathematics in Imaging, Harvard Medical School, Boston, MA, 02215, USA
| | | | | | - Michael Paquette
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada
| | - Samuel St-Jean
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada
| | - Gabriel Girard
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada
| | - François Rheault
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada
| | - Jasmeen Sidhu
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada
| | - Chantal M W Tax
- PROVIDI Lab, Image Sciences Institute, University Medical Center Utrecht, Utrecht, 3508, The Netherlands.,Cardiff University Brain Research Imaging Centre, School of Psychology, Cardiff University, Maindy Road, Cardiff, CF24 4HQ, UK
| | - Fenghua Guo
- PROVIDI Lab, Image Sciences Institute, University Medical Center Utrecht, Utrecht, 3508, The Netherlands
| | - Hamed Y Mesri
- PROVIDI Lab, Image Sciences Institute, University Medical Center Utrecht, Utrecht, 3508, The Netherlands
| | - Szabolcs Dávid
- PROVIDI Lab, Image Sciences Institute, University Medical Center Utrecht, Utrecht, 3508, The Netherlands
| | - Martijn Froeling
- Department of Radiology, University Medical Center Utrecht, Utrecht, 3508, The Netherlands
| | - Anneriet M Heemskerk
- PROVIDI Lab, Image Sciences Institute, University Medical Center Utrecht, Utrecht, 3508, The Netherlands
| | - Alexander Leemans
- PROVIDI Lab, Image Sciences Institute, University Medical Center Utrecht, Utrecht, 3508, The Netherlands
| | - Arnaud Boré
- Centre de recherche institut universitaire de geriatrie de Montreal (CRIUGM), Université de Montréal, Montreal, QC, Canada, H3W 1W5
| | - Basile Pinsard
- Centre de recherche institut universitaire de geriatrie de Montreal (CRIUGM), Université de Montréal, Montreal, QC, Canada, H3W 1W5.,Sorbonne Universités, UPMC Univ Paris 06, CNRS, INSERM, Laboratoire d'Imagerie Biomédicale (LIB), 75013, Paris, France
| | - Christophe Bedetti
- Centre de recherche institut universitaire de geriatrie de Montreal (CRIUGM), Université de Montréal, Montreal, QC, Canada, H3W 1W5.,Center for Advanced Research in Sleep Medicine, Hôpital du Sacré-Coeur de Montréal, Montreal, Canada, H4J 1C5
| | - Matthieu Desrosiers
- Centre de recherche institut universitaire de geriatrie de Montreal (CRIUGM), Université de Montréal, Montreal, QC, Canada, H3W 1W5
| | - Simona Brambati
- Centre de recherche institut universitaire de geriatrie de Montreal (CRIUGM), Université de Montréal, Montreal, QC, Canada, H3W 1W5
| | - Julien Doyon
- Centre de recherche institut universitaire de geriatrie de Montreal (CRIUGM), Université de Montréal, Montreal, QC, Canada, H3W 1W5
| | - Alessia Sarica
- Neuroimaging Unit, Institute of Bioimaging and Molecular Physiology (IBFM), National Research Council (CNR), Policlinico Magna Graecia, Germaneto, 88100, CZ, Italy
| | - Roberta Vasta
- Neuroimaging Unit, Institute of Bioimaging and Molecular Physiology (IBFM), National Research Council (CNR), Policlinico Magna Graecia, Germaneto, 88100, CZ, Italy
| | - Antonio Cerasa
- Neuroimaging Unit, Institute of Bioimaging and Molecular Physiology (IBFM), National Research Council (CNR), Policlinico Magna Graecia, Germaneto, 88100, CZ, Italy
| | - Aldo Quattrone
- Neuroimaging Unit, Institute of Bioimaging and Molecular Physiology (IBFM), National Research Council (CNR), Policlinico Magna Graecia, Germaneto, 88100, CZ, Italy.,Institute of Neurology, University Magna Graecia, Germaneto, 88100, CZ, Italy
| | - Jason Yeatman
- Institute for Learning & Brain Sciences and Department of Speech & Hearing Sciences, University of Washington, Seattle, WA, 98195, USA
| | - Ali R Khan
- Departments of Medical Biophysics & Medical Imaging, Schulich School of Medicine and Dentistry, Western University, 1151 Richmond St N, London, ON, Canada, N6A 5C1
| | - Wes Hodges
- Synaptive Medical Inc., MaRS Discovery District, 101 College Street, Suite 200, Toronto, ON, Canada, M5V 3B1
| | - Simon Alexander
- Synaptive Medical Inc., MaRS Discovery District, 101 College Street, Suite 200, Toronto, ON, Canada, M5V 3B1
| | - David Romascano
- Signal Processing Lab (LTS5), Ecole Polytechnique Federale de Lausanne, Lausanne, 1015, Switzerland
| | - Muhamed Barakovic
- Signal Processing Lab (LTS5), Ecole Polytechnique Federale de Lausanne, Lausanne, 1015, Switzerland
| | - Anna Auría
- Signal Processing Lab (LTS5), Ecole Polytechnique Federale de Lausanne, Lausanne, 1015, Switzerland
| | - Oscar Esteban
- Biomedical Image Technologies (BIT), ETSI Telecom., U. Politécnica de Madrid and CIBER-BBN, Madrid, 28040, Spain
| | - Alia Lemkaddem
- Signal Processing Lab (LTS5), Ecole Polytechnique Federale de Lausanne, Lausanne, 1015, Switzerland
| | - Jean-Philippe Thiran
- Signal Processing Lab (LTS5), Ecole Polytechnique Federale de Lausanne, Lausanne, 1015, Switzerland.,Department of Radiology, University Hospital Center (CHUV) and University of Lausanne (UNIL), Lausanne, 1011, Switzerland
| | - H Ertan Cetingul
- Medical Imaging Technologies, Siemens Healthcare, Princeton, NJ, 08540, USA
| | - Benjamin L Odry
- Medical Imaging Technologies, Siemens Healthcare, Princeton, NJ, 08540, USA
| | - Boris Mailhe
- Medical Imaging Technologies, Siemens Healthcare, Princeton, NJ, 08540, USA
| | - Mariappan S Nadar
- Medical Imaging Technologies, Siemens Healthcare, Princeton, NJ, 08540, USA
| | - Fabrizio Pizzagalli
- Imaging Genetics Center, Stevens Neuro Imaging and Informatics Institute, Keck School of Medicine of USC, Marina del Rey, CA, 90033, USA
| | - Gautam Prasad
- Imaging Genetics Center, Stevens Neuro Imaging and Informatics Institute, Keck School of Medicine of USC, Marina del Rey, CA, 90033, USA
| | - Julio E Villalon-Reina
- Imaging Genetics Center, Stevens Neuro Imaging and Informatics Institute, Keck School of Medicine of USC, Marina del Rey, CA, 90033, USA
| | - Justin Galvis
- Imaging Genetics Center, Stevens Neuro Imaging and Informatics Institute, Keck School of Medicine of USC, Marina del Rey, CA, 90033, USA
| | - Paul M Thompson
- Imaging Genetics Center, Stevens Neuro Imaging and Informatics Institute, Keck School of Medicine of USC, Marina del Rey, CA, 90033, USA
| | | | - Pedro Luque Laguna
- NatBrainLab, Institute of Psychiatry, Psychology & Neuroscience, King's College London, London, SE5 8AF, UK
| | - Luis Miguel Lacerda
- NatBrainLab, Institute of Psychiatry, Psychology & Neuroscience, King's College London, London, SE5 8AF, UK
| | - Rachel Barrett
- NatBrainLab, Institute of Psychiatry, Psychology & Neuroscience, King's College London, London, SE5 8AF, UK
| | - Flavio Dell'Acqua
- NatBrainLab, Institute of Psychiatry, Psychology & Neuroscience, King's College London, London, SE5 8AF, UK
| | - Marco Catani
- NatBrainLab, Institute of Psychiatry, Psychology & Neuroscience, King's College London, London, SE5 8AF, UK
| | - Laurent Petit
- Groupe d'imagerie Neurofonctionnelle-Institut des Maladies Neurodégénératives (GIN-IMN), UMR5293 CNRS, CEA, University of Bordeaux, Bordeaux, 33000, France
| | - Emmanuel Caruyer
- Centre national de la recherche scientifique (CNRS), Institute for Research in IT and Random Systems (IRISA), UMR 6074 VISAGES Project-Team, Rennes, 35042, France
| | - Alessandro Daducci
- Signal Processing Lab (LTS5), Ecole Polytechnique Federale de Lausanne, Lausanne, 1015, Switzerland.,Department of Radiology, University Hospital Center (CHUV) and University of Lausanne (UNIL), Lausanne, 1011, Switzerland
| | - Tim B Dyrby
- Danish Research Centre for Magnetic Resonance, Center for Functional and Diagnostic Imaging and Research, Copenhagen University Hospital Hvidovre, Hvidovre, 2650, Denmark.,Department of Applied Mathematics and Computer Science, Technical University of Denmark, Kongens Lyngby, 2800, Denmark
| | - Tim Holland-Letz
- Division of Biostatistics, German Cancer Research Center (DKFZ), Heidelberg, 69120, Germany
| | - Claus C Hilgetag
- Department of Computational Neuroscience, University Medical Center Eppendorf, Hamburg, 20246, Germany
| | - Bram Stieltjes
- University Hospital Basel, Radiology & Nuclear Medicine Clinic, Basel, 4031, Switzerland
| | - Maxime Descoteaux
- Sherbrooke Connectivity Imaging Lab (SCIL), Université de Sherbrooke, Sherbrooke, QC J1K 0A5, QC, Canada.
| |
Collapse
|
31
|
Abstract
The ability to map brain networks in living individuals is fundamental in efforts to chart the relation between human behavior, health and disease. Advances in network neuroscience may benefit from developing new frameworks for mapping brain connectomes. We present a framework to encode structural brain connectomes and diffusion-weighted magnetic resonance (dMRI) data using multidimensional arrays. The framework integrates the relation between connectome nodes, edges, white matter fascicles and diffusion data. We demonstrate the utility of the framework for in vivo white matter mapping and anatomical computing by evaluating 1,490 connectomes, thirteen tractography methods, and three data sets. The framework dramatically reduces storage requirements for connectome evaluation methods, with up to 40x compression factors. Evaluation of multiple, diverse datasets demonstrates the importance of spatial resolution in dMRI. We measured large increases in connectome resolution as function of data spatial resolution (up to 52%). Moreover, we demonstrate that the framework allows performing anatomical manipulations on white matter tracts for statistical inference and to study the white matter geometrical organization. Finally, we provide open-source software implementing the method and data to reproduce the results.
Collapse
Affiliation(s)
- Cesar F Caiafa
- Department of Psychological and, Brain Sciences Indiana University Bloomington, IN, 47405, USA
- Instituto Argentino de Radioastronomía (IAR), CONICET CCT, La Plata Villa Elisa, 1894, Argentina
- Facultad de Ingeniería - Departamento de Computación, UBA Buenos Aires, C1063ACV, Argentina
| | - Franco Pestilli
- Department of Psychological and, Brain Sciences Indiana University Bloomington, IN, 47405, USA.
- Department of Intelligent Systems, Engineering Indiana University Bloomington, IN, 47405, USA.
- Department of Computer Science, Indiana University Bloomington, IN, 47405, USA.
- Program in Neuroscience Indiana University Bloomington, IN, 47405, USA.
- Program in Cognitive Science Indiana University Bloomington, IN, 47405, USA.
- School of Optometry Indiana University Bloomington, IN, 47405, USA.
- Indiana Network Science Institute Indiana University Bloomington, IN, 47405, USA.
| |
Collapse
|