Literature DB >> 35767314

Logistic Regression-Based Model Is More Efficient Than U-Net Model for Reliable Whole Brain Magnetic Resonance Imaging Segmentation.

Henry Dieckhaus1, Rozanna Meijboom2, Serhat Okar3, Tianxia Wu4, Prasanna Parvathaneni3, Yair Mina5,6, Siddharthan Chandran2, Adam D Waldman2, Daniel S Reich3, Govind Nair1.   

Abstract

OBJECTIVES: Automated whole brain segmentation from magnetic resonance images is of great interest for the development of clinically relevant volumetric markers for various neurological diseases. Although deep learning methods have demonstrated remarkable potential in this area, they may perform poorly in nonoptimal conditions, such as limited training data availability. Manual whole brain segmentation is an incredibly tedious process, so minimizing the data set size required for training segmentation algorithms may be of wide interest. The purpose of this study was to compare the performance of the prototypical deep learning segmentation architecture (U-Net) with a previously published atlas-free traditional machine learning method, Classification using Derivative-based Features (C-DEF) for whole brain segmentation, in the setting of limited training data.
MATERIALS AND METHODS: C-DEF and U-Net models were evaluated after training on manually curated data from 5, 10, and 15 participants in 2 research cohorts: (1) people living with clinically diagnosed HIV infection and (2) relapsing-remitting multiple sclerosis, each acquired at separate institutions, and between 5 and 295 participants' data using a large, publicly available, and annotated data set of glioblastoma and lower grade glioma (brain tumor segmentation). Statistics was performed on the Dice similarity coefficient using repeated-measures analysis of variance and Dunnett-Hsu pairwise comparison.
RESULTS: C-DEF produced better segmentation than U-Net in lesion (29.2%-38.9%) and cerebrospinal fluid (5.3%-11.9%) classes when trained with data from 15 or fewer participants. Unlike C-DEF, U-Net showed significant improvement when increasing the size of the training data (24%-30% higher than baseline). In the brain tumor segmentation data set, C-DEF produced equivalent or better segmentations than U-Net for enhancing tumor and peritumoral edema regions across all training data sizes explored. However, U-Net was more effective than C-DEF for segmentation of necrotic/non-enhancing tumor when trained on 10 or more participants, probably because of the inconsistent signal intensity of the tissue class.
CONCLUSIONS: These results demonstrate that classical machine learning methods can produce more accurate brain segmentation than the far more complex deep learning methods when only small or moderate amounts of training data are available (n ≤ 15). The magnitude of this advantage varies by tissue and cohort, while U-Net may be preferable for deep gray matter and necrotic/non-enhancing tumor segmentation, particularly with larger training data sets (n ≥ 20). Given that segmentation models often need to be retrained for application to novel imaging protocols or pathology, the bottleneck associated with large-scale manual annotation could be avoided with classical machine learning algorithms, such as C-DEF.
Copyright © 2022 Written work prepared by employees of the Federal Government as part of their official duties is, under the U.S. Copyright Act, a “work of the United States Government” for which copyright protection under Title 17 of the United States Code is not available. As such, copyright does not extend to the contributions of employees of the Federal Government.

Entities:  

Mesh:

Year:  2022        PMID: 35767314      PMCID: PMC9258518          DOI: 10.1097/RMR.0000000000000296

Source DB:  PubMed          Journal:  Top Magn Reson Imaging        ISSN: 0899-3459


  27 in total

Review 1.  Artificial Intelligence and Machine Learning in Radiology: Current State and Considerations for Routine Clinical Implementation.

Authors:  Julian L Wichmann; Martin J Willemink; Carlo N De Cecco
Journal:  Invest Radiol       Date:  2020-09       Impact factor: 6.016

2.  UNet++: Redesigning Skip Connections to Exploit Multiscale Features in Image Segmentation.

Authors:  Zongwei Zhou; Md Mahfuzur Rahman Siddiquee; Nima Tajbakhsh; Jianming Liang
Journal:  IEEE Trans Med Imaging       Date:  2019-12-13       Impact factor: 10.048

Review 3.  A Practical Guide to Artificial Intelligence-Based Image Analysis in Radiology.

Authors:  Thomas Weikert; Joshy Cyriac; Shan Yang; Ivan Nesic; Victor Parmar; Bram Stieltjes
Journal:  Invest Radiol       Date:  2020-01       Impact factor: 6.016

4.  MultiResUNet : Rethinking the U-Net architecture for multimodal biomedical image segmentation.

Authors:  Nabil Ibtehaz; M Sohel Rahman
Journal:  Neural Netw       Date:  2019-09-04

5.  Advancing The Cancer Genome Atlas glioma MRI collections with expert segmentation labels and radiomic features.

Authors:  Spyridon Bakas; Hamed Akbari; Aristeidis Sotiras; Michel Bilello; Martin Rozycki; Justin S Kirby; John B Freymann; Keyvan Farahani; Christos Davatzikos
Journal:  Sci Data       Date:  2017-09-05       Impact factor: 6.444

Review 6.  Assessing treatment outcomes in multiple sclerosis trials and in the clinical setting.

Authors:  Carmen Tur; Marcello Moccia; Frederik Barkhof; Jeremy Chataway; Jaume Sastre-Garriga; Alan J Thompson; Olga Ciccarelli
Journal:  Nat Rev Neurol       Date:  2018-01-12       Impact factor: 42.937

7.  Comparison of multiple sclerosis lesions at 1.5 and 3.0 Tesla.

Authors:  Nancy L Sicotte; Rhonda R Voskuhl; Seth Bouvier; Rochelle Klutch; Mark S Cohen; John C Mazziotta
Journal:  Invest Radiol       Date:  2003-07       Impact factor: 6.016

8.  Diagnostic criteria for multiple sclerosis: 2010 revisions to the McDonald criteria.

Authors:  Chris H Polman; Stephen C Reingold; Brenda Banwell; Michel Clanet; Jeffrey A Cohen; Massimo Filippi; Kazuo Fujihara; Eva Havrdova; Michael Hutchinson; Ludwig Kappos; Fred D Lublin; Xavier Montalban; Paul O'Connor; Magnhild Sandberg-Wollheim; Alan J Thompson; Emmanuelle Waubant; Brian Weinshenker; Jerry S Wolinsky
Journal:  Ann Neurol       Date:  2011-02       Impact factor: 10.422

Review 9.  Neuroimaging Biomarkers for Alzheimer's Disease.

Authors:  Freddie Márquez; Michael A Yassa
Journal:  Mol Neurodegener       Date:  2019-06-07       Impact factor: 14.195

Review 10.  Variability and Standardization of Quantitative Imaging: Monoparametric to Multiparametric Quantification, Radiomics, and Artificial Intelligence.

Authors:  Akifumi Hagiwara; Shohei Fujita; Yoshiharu Ohno; Shigeki Aoki
Journal:  Invest Radiol       Date:  2020-09       Impact factor: 10.065

View more

北京卡尤迪生物科技股份有限公司 © 2022-2023.