Estimation of reference curves for brain atrophy and analysis of robustness to machine effects

Neurodegenerative diseases like Alzheimer's are difficult to diagnose due to brain complexity and imaging variability. However, volumetric analysis tools, using reference curves, help detect abnormal brain atrophy and support diagnosis and monitoring. This study evaluates the robustness of three segmentation algorithms, AssemblyNet, FastSurfer and FreeSurfer, in constructing brain volume reference curves and detecting hippocampal atrophy. Using data from 3,730 cognitively normal subjects, we built reference curves and assessed robustness to magnetic field strength (1.5T vs. 3T) using four error metrics (sMAPE, sMSPE, wMAPE, sMdAPE) with bootstrap validation. We evaluated classification performance using hippocampal atrophy rates and HAVAs scores (Hippocampal-Amygdalo-Ventricular Atrophy scores). AssemblyNet shows the lowest errors across all robustness metrics. In contrast, FastSurfer and FreeSurfer exhibit greater deviations, indicating higher sensitivity to field strength variability. AssemblyNet provides consistent hippocampal atrophy rates across all reference models, despite slightly lower sensitivity, while FastSurfer and FreeSurfer display greater variability. Specificity ranges from 0.87 to 0.91 for AssemblyNet, compared to 0.76-0.93 for FastSurfer and 0.86-0.93 for FreeSurfer. Using the HAVAs score, all methods detect high atrophy rates in Alzheimer's patients. FastSurfer achieves the highest sensitivity (0.98), while AssemblyNet reaches the best specificity (0.95) and the highest balanced accuracy (0.91). This study underscores the importance of algorithm choice for reliable brain volumetric analysis in heterogeneous imaging environments. Among the methods tested, AssemblyNet stands out as both sensitive to Alzheimer's-related atrophy and robust to acquisition variability, making it a strong candidate when analyzing hippocampal volumes in large, multi-site datasets.

Authors

Institutions

Publication Details

Journal
Scientific Reports
Published
2025-10-03
DOI
https://doi.org/10.1038/s41598-025-18073-z
Primary Topic
Dementia and Cognitive Impairment Research
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Estimation of reference curves for brain atrophy and analysis of robustness to machine effects

Evan Fletcher, Stephen Salloway, Masandra Weintraub, Effie Mitsis et al.
Scientific Reports
Dementia and Cognitive Impairment Research
article

Estimation of reference curves for brain atrophy and analysis of robustness to machine effects

Evan Fletcher, Stephen Salloway, Masandra Weintraub, Effie Mitsis, Pierre N. Tariot, Lawrence S. Honig, Laura A. Flashman, Clifford R. Jack, Brandy R. Matthews, Magdalena Korecka, Robert B. Santulli, John Olichney, Kristine Lipowski, Li Shen, Gregory A. Jicha, Maria T. Greig, Donna Munic, Dana M. Pogorelec, Dana Nguyen, Irina Rachisky, Robert Bartha, Lean Tha, Richard E. Carson, Ann Marie Hake, Maria Carroll, Karen L. Bell, John Brockington, Stephen Pasternak, Teresa Villena, William M. Brooks, Nadira Trncic, Susan K. Schultz, Stephen Pasternack, Balebail Ashok Raj, Parianne Fatica, Nick C. Fox, Kaycee M. Sink, Adam L. Boxer, Raina Carter, Gus Jiminez, Karen Crawford, Carl Sadowsky, Michelle Rainka, Raj C. Shah, Judith L. Heidebrink, Neil W. Kowall, Peggy Roberts, Karen Ekstam Smith, Henry Rusinek, Allan I. Levey, Jerome A. Yesavage, Daniel D’Agostino, John Q. Trojanowki, Kristin Fargher, Iris Sim, Antero Sarrael, Lidia Glodzik, Andrew J. Saykin, Dzintra Celmins, Erin Franklin, Kenneth Spicer, Howard Rosen, Randall Griffith, Chester A. Mathis, Heather Johnson, Sonia Pawluczyk, Helen Vanderswag, Michael Donohue, Tom Montine, Yaakov Stern, Erin Householder, Joanne Lord, MaryAnn Oakley, Ronald Killiany, Elodie Piot, Sherye A. Sirrel, Susan De Santi, Anton P. Porsteinsson, Irina Rachinsky, Norman Relkin, Henry Querfurth, John Hsiao, Dick Trost, Barton Lane, Alexander Norbash, Marissa Natelson Love, Nunzio Pomara, Andrew E. Budson, Mauricio Beccera, Marc Seltzer, Adrian Preda, Danielle Harvey, Kathleen Tingus, Charles DeCarli, Kim Martin, Leslie Gordineer, Daniel Silverman, Ronald G. Thomas, Jeffrey Kaye, Sandra E. Black, Liberty Teodoro, Sandra A. Jacobson, Cynthia Hunt, Bojana Stefanovic, Janet S. Cella, David A. Wolk, Bonnie S. Goldstein, Prashanthi Vemuri, Norbert Schuff, Charles Bernick, Borna Bonakdarpour, Michal Figurski, Kris Johnson, Howard Bergman, Bruce L. Miller, Chiadi U. Onyike, Peter Davies, Eric M. Reiman, Maria Kataki, Ellen Woo, Russell Howard Swerdlow, Christina A. Michel, Matthew L. Senjem, Javier Villanueva‐Meyer, Saba Wolday, Jeff D. Williamson, Sara S. Mason, Brian R. Ott, Bret Borowski, Liana Apostolova, Geoffrey Tremont, Franz Hefti, Colleen S. Albers, Owen Carmichael, Karen Blank, Kelly M. Makino, Gloria Chaing, Pauline Maillard, Oscar L. Lopez, Kyle Womack, Hillel Grossman, Curtis Tatsuoka, Partha Sinha, Lon S. Schneider, David S. Knopman, Laurel Beckett, Brigid Reynolds, Christine M. Belden, David C. Perry, Alexandre Krainik, Ging‐Yuek Robin Hsiung, Maria Luisa Mandelli, Alan J. Lerner, David Geldmacher, P. Hardy, Sarah Walter, Stephen Correia, Stephanie Reeder, Gad Marshall, Martha G. MacAvoy, Virginia M.‐Y. Lee, Jeffrey Murray Burns, Chris Hosein, Michael Borrie, Mary L. Creech, Lisa Ravdin, Rachelle S. Doody, Sara Dolen, Daniel Varón, Paula Ogrocki, Arnaud Attyé, Joy L. Taylor, Howard J. Rosen, Nigel J. Cairns, Elizabeth Finger, Stacy Schneider, Thomas O. Obisesan, Raymundo Hernando, Davie Holtzman, Benita Mudge, George Bartzokis, Howard Chertkow, Michael Lin, Maria Carrillo, Jennifer Mason, Beau M. Ances, Jared R. Brosch, Mark A. Mintun, Kejal Kantarci, Scott Herring, Terence Z. Wong, Félix Renard, Greg Sorensen, Francine Parfitt, Brittany Cerbone, Salvador Borges‐Neto, James J. Lah, Jeffrey R. Petrella, Connie Brand, Christopher H. van Dyck, Betty Lind, Tracy Kendall, Erik D. Roberson, Robert A. Koeppe, Paul Thompson, James E. Galvin, Elizabeth Oates, Joseph F. Quinn, Ranjan Duara, Matt A. Bernstein, Diana Kerwin, Peter J. Snyder, Bradford C. Dickerson, Lisa C. Silbert, Zaven S. Khachaturian, Raymond Scott Turner, Steven Edward Arnold, Archana B. Balasubramanian, Daniel Marson, Paul Malloy, David Bachman, Neill R. Graff‐Radford, John R. Morris, Douglas W. Scharre, Kewei Chen, Sung Eun Kim, David G. Clark, Steven M. Paul, Karen Anderson, Pradeep Garg, Robert C. Green, Marilyn Albert, Howard Fillit, David Jones, Amanda Smith, Laura L. Boles Ponto, Stephanie Kielb, Mony J. de Leon, Marwan N. Sabbagh, M. Saleem Ismail, Maria-Luisa Gorno-Tempini, Bryan M. Spann, Andrew Kertesz, Jason H. Karlawish, Norm Foster, Vernice Bates, Arthur W. Toga, Jared Tinklenberg, Kimoko Domoto-Reilly, Charles D. Smith, Tamie Sather, M. Marcel Mesulam, Mary L. Hynes, Paul Aisen, Sterling C. Johnson, Curtis Caldwell, Patricia Lynn Johnson, Tatiana M. Foroud, William Jagust, Steven Potkin, Sanjay Asthana, Michael W. Weiner, James Brewer, Angela Oliver, Keith A. Johnson, William C. Potter, Scott Neu, Joanne Allard, Smita Kittur, Alice D. Brown, Allyson Rosen, Godfrey D. Pearlson, Dick Drost, Ronald Petersen, Anahita Adeli, John Rogers, R.T. Frank, Mary Quiceno, Po H. Lu, Cynthia M. Carlsson, Scott McGinnis, Marc Raichle, Kathleen A. Johnson, T.-Y. Lee, Hyungsub Shim, Martin R. Farlow, John Kornak, Nancy Johnson, Dana Mathews, Edward Coleman, P. Murali Doraiswamy, Munir Chowdhury, Howard J. Feldman, Horacio Capote, Gary Conrad, Susan Rountree, William W. Seeley, Lisa Taylor‐Reinwald, Catherine Mc-Adams-Ortiz, Anna Burke, Chad Ward, Reisa A. Sperling, Earl A. Zimmerman, Lew Kuller, Donna M. Simpson, Pradeep Varma, Jeff Gunter, Michele Assaly, Ruth A. Mulnard, Leyla deToledo‐Morrell, Leslie M. Shaw, Jacobo Mintzer, Franklin Watkins, Chuang-Kuo Wu, Gaby Thai, Mimi Dang, Bradley F. Boeve, Kelley Faber, Susan M. Landau
article en

Abstract

Neurodegenerative diseases like Alzheimer's are difficult to diagnose due to brain complexity and imaging variability. However, volumetric analysis tools, using reference curves, help detect abnormal brain atrophy and support diagnosis and monitoring. This study evaluates the robustness of three segmentation algorithms, AssemblyNet, FastSurfer and FreeSurfer, in constructing brain volume reference curves and detecting hippocampal atrophy. Using data from 3,730 cognitively normal subjects, we built reference curves and assessed robustness to magnetic field strength (1.5T vs. 3T) using four error metrics (sMAPE, sMSPE, wMAPE, sMdAPE) with bootstrap validation. We evaluated classification performance using hippocampal atrophy rates and HAVAs scores (Hippocampal-Amygdalo-Ventricular Atrophy scores). AssemblyNet shows the lowest errors across all robustness metrics. In contrast, FastSurfer and FreeSurfer exhibit greater deviations, indicating higher sensitivity to field strength variability. AssemblyNet provides consistent hippocampal atrophy rates across all reference models, despite slightly lower sensitivity, while FastSurfer and FreeSurfer display greater variability. Specificity ranges from 0.87 to 0.91 for AssemblyNet, compared to 0.76-0.93 for FastSurfer and 0.86-0.93 for FreeSurfer. Using the HAVAs score, all methods detect high atrophy rates in Alzheimer's patients. FastSurfer achieves the highest sensitivity (0.98), while AssemblyNet reaches the best specificity (0.95) and the highest balanced accuracy (0.91). This study underscores the importance of algorithm choice for reliable brain volumetric analysis in heterogeneous imaging environments. Among the methods tested, AssemblyNet stands out as both sensitive to Alzheimer's-related atrophy and robust to acquisition variability, making it a strong candidate when analyzing hippocampal volumes in large, multi-site datasets.

Scientific ReportsVol. 15(1)
Boston University (US), Northwestern University (US), University of London (GB), Nathan Kline Institute for Psychiatric Research (US), University of Iowa (US), Alzheimer's Association (US), Brigham and Women's Hospital (US), Rush University Medical Center (US), Dartmouth–Hitchcock Medical Center (US), Centre National de la Recherche Scientifique (FR), Albert Einstein College of Medicine (US), Hartford Hospital (US), Cleveland Clinic (US), Sunnybrook Health Science Centre (CA), Siemens (Germany) (DE), Alzheimer's Drug Discovery Foundation (US), LAC+USC Medical Center (US), Berkeley College (US), Harvard University (US), Howard University (US), University of Kentucky (US), Johns Hopkins University (US), Emory University (US), Medical University of South Carolina (US), Inserm (FR), Oregon Health & Science University (US), University of Pittsburgh (US), University of California, San Francisco (US), Baylor College of Medicine (US), Georgetown University (US), Yeshiva University (US), University of Washington (US), University of California, Irvine (US), Washington University in St. Louis (US), Cornell University (US), University of Utah (US), University of South Florida (US), Brown University (US), University of Michigan (US), Columbia University Irving Medical Center (US), Georgetown University Medical Center (US), Lou Ruvo Brain Institute (US), Centre Hospitalier Universitaire de Grenoble (FR), UCLA Medical Center (US), Jacksonville College (US), Jewish General Hospital (CA), WinnMed (US), University of Rochester Medical Center (US), University of Alabama at Birmingham (US), Yale University (US), University of California San Diego (US), FIT Consulting (Italy) (IT), Massachusetts General Hospital (US), St Joseph's Health Centre (CA), Dent Neurologic Institute (US), Palm Beach Neurology (US), Indiana University Bloomington (US), Mayo Clinic in Arizona (US), Los Angeles Biomedical Research Institute (US), Premiere Research Institute (US), Duke Medical Center (US), University of Kansas Medical Center (US), Prevent Alzheimer’s Disease 2020 (US), National Institute on Aging (US), Laboratoire d'Écologie Alpine (FR), USF Health Byrd Alzheimer's Institute (US), Mayo Clinic in Florida (US), Grenoble Institute of Neurosciences (FR), St Joseph's Health Care (CA), Neurological Surgery (US), National Institute of Mental Health (US), Acumen Pharmaceuticals (United States) (US), Banner Alzheimer’s Institute, Wake Forest University (US), McGill University (CA), The Ohio State University (US), New York University (US), Seattle University (US), Case Western Reserve University (US), Albany Medical Center Hospital (US), University of Pennsylvania (US), University of California, Davis (US), Université Grenoble Alpes (FR), University of California, Berkeley (US), Stanford University (US), Icahn School of Medicine at Mount Sinai (US)
University of Southern California, Alzheimer's Disease Neuroimaging Initiative, Northern California Institute for Research and Education, Association Nationale de la Recherche et de la Technologie, National Institutes of Health, Engineering and Physical Sciences Research Council
Openalex Percentile: Top 67%
Dementia and Cognitive Impairment Research
DOIOpen Access PDFOpenAlex
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.