Literature DB >> 33902457

MADGAN: unsupervised medical anomaly detection GAN using multiple adjacent brain MRI slice reconstruction.

Changhee Han1, Leonardo Rundo2,3, Kohei Murao4, Tomoyuki Noguchi5, Yuki Shimahara6, Zoltán Ádám Milacski7, Saori Koshino8, Evis Sala2,3, Hideki Nakayama9,10, Shin'ichi Satoh4.   

Abstract

BACKGROUND: Unsupervised learning can discover various unseen abnormalities, relying on large-scale unannotated medical images of healthy subjects. Towards this, unsupervised methods reconstruct a 2D/3D single medical image to detect outliers either in the learned feature space or from high reconstruction loss. However, without considering continuity between multiple adjacent slices, they cannot directly discriminate diseases composed of the accumulation of subtle anatomical anomalies, such as Alzheimer's disease (AD). Moreover, no study has shown how unsupervised anomaly detection is associated with either disease stages, various (i.e., more than two types of) diseases, or multi-sequence magnetic resonance imaging (MRI) scans.
RESULTS: We propose unsupervised medical anomaly detection generative adversarial network (MADGAN), a novel two-step method using GAN-based multiple adjacent brain MRI slice reconstruction to detect brain anomalies at different stages on multi-sequence structural MRI: (Reconstruction) Wasserstein loss with Gradient Penalty + 100 [Formula: see text] loss-trained on 3 healthy brain axial MRI slices to reconstruct the next 3 ones-reconstructs unseen healthy/abnormal scans; (Diagnosis) Average [Formula: see text] loss per scan discriminates them, comparing the ground truth/reconstructed slices. For training, we use two different datasets composed of 1133 healthy T1-weighted (T1) and 135 healthy contrast-enhanced T1 (T1c) brain MRI scans for detecting AD and brain metastases/various diseases, respectively. Our self-attention MADGAN can detect AD on T1 scans at a very early stage, mild cognitive impairment (MCI), with area under the curve (AUC) 0.727, and AD at a late stage with AUC 0.894, while detecting brain metastases on T1c scans with AUC 0.921.
CONCLUSIONS: Similar to physicians' way of performing a diagnosis, using massive healthy training data, our first multiple MRI slice reconstruction approach, MADGAN, can reliably predict the next 3 slices from the previous 3 ones only for unseen healthy images. As the first unsupervised various disease diagnosis, MADGAN can reliably detect the accumulation of subtle anatomical anomalies and hyper-intense enhancing lesions, such as (especially late-stage) AD and brain metastases on multi-sequence MRI scans.

Entities:  

Keywords:  Brain MRI reconstruction; Generative adversarial networks; Self-attention; Unsupervised anomaly detection; Various disease diagnosis

Year:  2021        PMID: 33902457     DOI: 10.1186/s12859-020-03936-1

Source DB:  PubMed          Journal:  BMC Bioinformatics        ISSN: 1471-2105            Impact factor:   3.169


  19 in total

1.  Unsupervised pathology detection in medical images using conditional variational autoencoders.

Authors:  Hristina Uzunova; Sandra Schultz; Heinz Handels; Jan Ehrhardt
Journal:  Int J Comput Assist Radiol Surg       Date:  2018-12-12       Impact factor: 2.924

2.  Texture descriptors and voxels for the early diagnosis of Alzheimer's disease.

Authors:  Loris Nanni; Sheryl Brahnam; Christian Salvatore; Isabella Castiglioni
Journal:  Artif Intell Med       Date:  2019-05-18       Impact factor: 5.326

Review 3.  Whole-genome analyses of whole-brain data: working within an expanded search space.

Authors:  Sarah E Medland; Neda Jahanshad; Benjamin M Neale; Paul M Thompson
Journal:  Nat Neurosci       Date:  2014-05-27       Impact factor: 24.884

Review 4.  Diagnosis and Management of Dementia: Review.

Authors:  Zoe Arvanitakis; Raj C Shah; David A Bennett
Journal:  JAMA       Date:  2019-10-22       Impact factor: 56.272

5.  A parameter-efficient deep learning approach to predict conversion from mild cognitive impairment to Alzheimer's disease.

Authors:  Simeon Spasov; Luca Passamonti; Andrea Duggento; Pietro Liò; Nicola Toschi
Journal:  Neuroimage       Date:  2019-01-14       Impact factor: 6.556

6.  Deep ensemble learning of sparse regression models for brain disease diagnosis.

Authors:  Heung-Il Suk; Seong-Whan Lee; Dinggang Shen
Journal:  Med Image Anal       Date:  2017-01-24       Impact factor: 8.545

Review 7.  Not-so-supervised: A survey of semi-supervised, multi-instance, and transfer learning in medical image analysis.

Authors:  Veronika Cheplygina; Marleen de Bruijne; Josien P W Pluim
Journal:  Med Image Anal       Date:  2019-03-29       Impact factor: 8.545

8.  Temporoparietal MR imaging measures of atrophy in subjects with mild cognitive impairment that predict subsequent diagnosis of Alzheimer disease.

Authors:  R S Desikan; H J Cabral; B Fischl; C R G Guttmann; D Blacker; B T Hyman; M S Albert; R J Killiany
Journal:  AJNR Am J Neuroradiol       Date:  2008-12-26       Impact factor: 3.825

9.  Identify the Atrophy of Alzheimer's Disease, Mild Cognitive Impairment and Normal Aging Using Morphometric MRI Analysis.

Authors:  Xiangyu Ma; Zhaoxia Li; Bin Jing; Han Liu; Dan Li; Haiyun Li
Journal:  Front Aging Neurosci       Date:  2016-10-18       Impact factor: 5.750

10.  Magnetic resonance imaging biomarkers for the early diagnosis of Alzheimer's disease: a machine learning approach.

Authors:  Christian Salvatore; Antonio Cerasa; Petronilla Battista; Maria C Gilardi; Aldo Quattrone; Isabella Castiglioni
Journal:  Front Neurosci       Date:  2015-09-01       Impact factor: 4.677

View more
  15 in total

1.  BrainGAN: Brain MRI Image Generation and Classification Framework Using GAN Architectures and CNN Models.

Authors:  Halima Hamid N Alrashedy; Atheer Fahad Almansour; Dina M Ibrahim; Mohammad Ali A Hammoudeh
Journal:  Sensors (Basel)       Date:  2022-06-06       Impact factor: 3.847

2.  Longitudinal clustering analysis and prediction of Parkinson's disease progression using radiomics and hybrid machine learning.

Authors:  Mohammad R Salmanpour; Mojtaba Shamsaei; Ghasem Hajianfar; Hamid Soltanian-Zadeh; Arman Rahmim
Journal:  Quant Imaging Med Surg       Date:  2022-02

3.  The effect of deep feature concatenation in the classification problem: An approach on COVID-19 disease detection.

Authors:  Emine Cengil; Ahmet Çınar
Journal:  Int J Imaging Syst Technol       Date:  2021-10-10       Impact factor: 2.177

4.  Pediatric Otoscopy Video Screening With Shift Contrastive Anomaly Detection.

Authors:  Weiyao Wang; Aniruddha Tamhane; Christine Santos; John R Rzasa; James H Clark; Therese L Canares; Mathias Unberath
Journal:  Front Digit Health       Date:  2022-02-10

5.  Automated detection of pulmonary embolism from CT-angiograms using deep learning.

Authors:  Heidi Huhtanen; Mikko Nyman; Tarek Mohsen; Arho Virkki; Antti Karlsson; Jussi Hirvonen
Journal:  BMC Med Imaging       Date:  2022-03-14       Impact factor: 1.930

6.  Explanation-Driven Deep Learning Model for Prediction of Brain Tumour Status Using MRI Image Data.

Authors:  Loveleen Gaur; Mohan Bhandari; Tanvi Razdan; Saurav Mallik; Zhongming Zhao
Journal:  Front Genet       Date:  2022-03-14       Impact factor: 4.599

Review 7.  Generative Adversarial Networks in Brain Imaging: A Narrative Review.

Authors:  Maria Elena Laino; Pierandrea Cancian; Letterio Salvatore Politi; Matteo Giovanni Della Porta; Luca Saba; Victor Savevski
Journal:  J Imaging       Date:  2022-03-23

8.  Diagnostic Performance of Generative Adversarial Network-Based Deep Learning Methods for Alzheimer's Disease: A Systematic Review and Meta-Analysis.

Authors:  Changxing Qu; Yinxi Zou; Yingqiao Ma; Qin Chen; Jiawei Luo; Huiyong Fan; Zhiyun Jia; Qiyong Gong; Taolin Chen
Journal:  Front Aging Neurosci       Date:  2022-04-21       Impact factor: 5.750

9.  Automatic Liver Segmentation in CT Images with Enhanced GAN and Mask Region-Based CNN Architectures.

Authors:  Xiaoqin Wei; Xiaowen Chen; Ce Lai; Yuanzhong Zhu; Hanfeng Yang; Yong Du
Journal:  Biomed Res Int       Date:  2021-12-16       Impact factor: 3.411

10.  Impact of GAN-based lesion-focused medical image super-resolution on the robustness of radiomic features.

Authors:  Mauro Castelli; Leonardo Rundo; Erick Costa de Farias; Christian di Noia; Changhee Han; Evis Sala
Journal:  Sci Rep       Date:  2021-11-01       Impact factor: 4.379

View more

北京卡尤迪生物科技股份有限公司 © 2022-2023.