MICCAI-CDMRI 2023 QuantConn Challenge Findings on Achieving Robust Quantitative Connectivity through Harmonized Preprocessing of Diffusion MRI

Nancy R. Newlin1, Kurt Schilling2, Serge Koudoro3, Bramsh Qamar Chandio4, Praitayini Kanakaraj1, Daniel Moyer1, Claire E. Kelly5,6,7, Sila Genc5,8, Jian Chen5, Joseph Yuan-Mou Yang5,8,9,10, Ye Wu11, Yifei He11, Jiawei Zhang12, Qingrun Zeng13, Fan Zhang13, Nagesh Adluru14, Vishwesh Nath15, Sudhir Pathak16, Walter Schneider16, Anurag Gade17, Yogesh Rathi18, Tom Hendriks19, Anna Vilanova19, Maxime Chamberland19, Tomasz Pieciak20,21, Dominika Ciupek21, Antonio Tristán-Vega20, Santiago Aja-Fernández20, Maciej Malawski21, Gani Ouedraogo22, Julia Machnio21, Christian Ewert23, Paul M. Thompson4, Neda Jahanshad4, Eleftherios Garyfallidis3, Bennett A. Landman1,2,24,25
1: Department of Computer Science, Vanderbilt University, Nashville, TN, 2: Department of Radiology and Radiological Sciences, Vanderbilt University Medical Center, 3: Indiana University Bloomington, Bloomington, IN, 4: Mark and Mary Stevens Neuroimaging and Informatics Institute, Keck School of Medicine of USC, Los Angeles, CA, 5: Developmental Imaging, Murdoch Children’s Research Institute, Melbourne, Australia, 6: Victorian Infant Brain Studies (VIBeS), Murdoch Children’s Research Institute, Melbourne, Australia, 7: Turner Institute for Brain and Mental Health, School of Psychological Sciences, Monash University, Melbourne, Australia, 8: Neuroscience Advanced Clinical Imaging Service (NACIS), Department of Neurosurgery, Royal Children’s Hospital, Melbourne, Australia, 9: Neuroscience Research, Murdoch Children’s Research Institute, Melbourne, Australia, 10: Department of Pediatrics, University of Melbourne, Melbourne, Australia, 11: School of Computer Science and Technology, Nanjing University of Science and Technology, Nanjing, China, 12: College of Information Engineering, Zhejiang University of Technology, Hangzhou, China, 13: School of Information and Communication Engineering, University of Electronic Science and Technology of China, Chengdu, China, 14: Waisman Center, Department of Radiology, University of Wisconsin, Madison, 15: NVIDIA, Nashville, TN, USA, 16: Learning Research and Development Center, University of Pittsburgh, 17: Brigham and Women’s Hospital, Boston, 18: Brigham and Women’s Hospital, Harvard Medical School, Boston, 19: Department of Computer Science and Mathematics, Eindhoven University of Technology, Netherlands, 20: LPI, ETSI Telecomunicación, Universidad de Valladolid, Castilla y León, Spain, 21: Sano Centre for Computational Medicine, 30-054 Kraków, Poland, 22: Aix-Marseille Université, Marseille, France, 23: AI in Medical Imaging, German Center for Neurodegenerative Diseases (DZNE), 24: Vanderbilt University Institute of Imaging Science, Vanderbilt University, Nashville, TN, USA, 25: Department of Electrical and Computer Engineering, Vanderbilt University, Nashville, TN, USA
Publication date: 2024/08/31
https://doi.org/10.59275/j.melba.2024-9c68
PDF · Code

Abstract

White matter alterations are increasingly implicated in neurological diseases and their progression. International-scale studies use diffusion-weighted magnetic resonance imaging (DW-MRI) to qualitatively identify changes in white matter microstructure and connectivity. Yet, quantitative analysis of DW-MRI data is hindered by inconsistencies stemming from varying acquisition protocols. Specifically, there is a pressing need to harmonize the preprocessing of DW-MRI datasets to ensure the derivation of robust quantitative diffusion metrics across acquisitions. In the MICCAI-CDMRI 2023 QuantConn challenge, participants were provided raw data from the same individuals collected on the same scanner but with two different acquisitions and tasked with preprocessing the DW-MRI to minimize acquisition differences while retaining biological variation. Harmonized submissions are evaluated on the reproducibility and comparability of cross-acquisition bundlewise microstructure measures, bundle shape features, and connectomics. The key innovations of the QuantConn challenge are that (1) we assess bundles and tractography in the context of harmonization for the first time, (2) we assess connectomics in the context of harmonization for the first time, and (3) we have 10x additional subjects over prior harmonization challenge, MUSHAC and 100x over SuperMUDI. We find that bundle surface area, fractional anisotropy, connectome assortativity, betweenness centrality, edge count, modularity, nodal strength, and participation coefficient measures are most biased by acquisition and that machine learning voxel-wise correction, RISH mapping, and NeSH methods effectively reduce these biases. In addition, microstructure measures AD, MD, RD, bundle length, connectome density, efficiency, and path length are least biased by these acquisition differences. A machine learning approach that learned voxelwise cross-acquisition relationships was the most effective at harmonizing connectomic, microstructure, and macrostructure features, but requires the same subject be scanned at each site co-registered. NeSH, a spatial and angular resampling method, was also effective and has generalizable framework not reliant co-registration. Our code is available at https://github.com/nancynewlin-masi/QuantConn/

Keywords

Diffusion MRI · harmonization · tractometry · tractography · connectomics · image processing

Bibtex @article{melba:2024:019:newlin, title = "MICCAI-CDMRI 2023 QuantConn Challenge Findings on Achieving Robust Quantitative Connectivity through Harmonized Preprocessing of Diffusion MRI", author = "Newlin, Nancy R. and Schilling, Kurt and Koudoro, Serge and Chandio, Bramsh Qamar and Kanakaraj, Praitayini and Moyer, Daniel and Kelly, Claire E. and Genc, Sila and Chen, Jian and Yang, Joseph Yuan-Mou and Wu, Ye and He, Yifei and Zhang, Jiawei and Zeng, Qingrun and Zhang, Fan and Adluru, Nagesh and Nath, Vishwesh and Pathak, Sudhir and Schneider, Walter and Gade, Anurag and Rathi, Yogesh and Hendriks, Tom and Vilanova, Anna and Chamberland, Maxime and Pieciak, Tomasz and Ciupek, Dominika and Tristán-Vega, Antonio and Aja-Fernández, Santiago and Malawski, Maciej and Ouedraogo, Gani and Machnio, Julia and Ewert, Christian and Thompson, Paul M. and Jahanshad, Neda and Garyfallidis, Eleftherios and Landman, Bennett A.", journal = "Machine Learning for Biomedical Imaging", volume = "2", issue = "August 2024 issue", year = "2024", pages = "1083--1105", issn = "2766-905X", doi = "https://doi.org/10.59275/j.melba.2024-9c68", url = "https://melba-journal.org/2024:019" }
RISTY - JOUR AU - Newlin, Nancy R. AU - Schilling, Kurt AU - Koudoro, Serge AU - Chandio, Bramsh Qamar AU - Kanakaraj, Praitayini AU - Moyer, Daniel AU - Kelly, Claire E. AU - Genc, Sila AU - Chen, Jian AU - Yang, Joseph Yuan-Mou AU - Wu, Ye AU - He, Yifei AU - Zhang, Jiawei AU - Zeng, Qingrun AU - Zhang, Fan AU - Adluru, Nagesh AU - Nath, Vishwesh AU - Pathak, Sudhir AU - Schneider, Walter AU - Gade, Anurag AU - Rathi, Yogesh AU - Hendriks, Tom AU - Vilanova, Anna AU - Chamberland, Maxime AU - Pieciak, Tomasz AU - Ciupek, Dominika AU - Tristán-Vega, Antonio AU - Aja-Fernández, Santiago AU - Malawski, Maciej AU - Ouedraogo, Gani AU - Machnio, Julia AU - Ewert, Christian AU - Thompson, Paul M. AU - Jahanshad, Neda AU - Garyfallidis, Eleftherios AU - Landman, Bennett A. PY - 2024 TI - MICCAI-CDMRI 2023 QuantConn Challenge Findings on Achieving Robust Quantitative Connectivity through Harmonized Preprocessing of Diffusion MRI T2 - Machine Learning for Biomedical Imaging VL - 2 IS - August 2024 issue SP - 1083 EP - 1105 SN - 2766-905X DO - https://doi.org/10.59275/j.melba.2024-9c68 UR - https://melba-journal.org/2024:019 ER -
2024:019 cover