Literature DB >> 31659587

Performance Comparison of Individual and Ensemble CNN Models for the Classification of Brain 18F-FDG-PET Scans.

Tomomi Nobashi1, Claudia Zacharias2, Jason K Ellis3, Valentina Ferri1, Mary Ellen Koran1, Benjamin L Franc1, Andrei Iagaru1, Guido A Davidzon4.   

Abstract

The high-background glucose metabolism of normal gray matter on [18F]-fluoro-2-D-deoxyglucose (FDG) positron emission tomography (PET) of the brain results in a low signal-to-background ratio, potentially increasing the possibility of missing important findings in patients with intracranial malignancies. To explore the strategy of using a deep learning classifier to aid in distinguishing normal versus abnormal findings on PET brain images, this study evaluated the performance of a two-dimensional convolutional neural network (2D-CNN) to classify FDG PET brain scans as normal (N) or abnormal (A).
METHODS: Two hundred eighty-nine brain FDG-PET scans (N; n = 150, A; n = 139) resulting in a total of 68,260 images were included. Nine individual 2D-CNN models with three different window settings for axial, coronal, and sagittal axes were trained and validated. The performance of these individual and ensemble models was evaluated and compared using a test dataset. Odds ratio, Akaike's information criterion (AIC), and area under curve (AUC) on receiver-operative-characteristic curve, accuracy, and standard deviation (SD) were calculated.
RESULTS: An optimal window setting to classify normal and abnormal scans was different for each axis of the individual models. An ensembled model using different axes with an optimized window setting (window-triad) showed better performance than ensembled models using the same axis and different windows settings (axis-triad). Increase in odds ratio and decrease in SD were observed in both axis-triad and window-triad models compared with individual models, whereas improvements of AUC and AIC were seen in window-triad models. An overall model averaging the probabilities of all individual models showed the best accuracy of 82.0%.
CONCLUSIONS: Data ensemble using different window settings and axes was effective to improve 2D-CNN performance parameters for the classification of brain FDG-PET scans. If prospectively validated with a larger cohort of patients, similar models could provide decision support in a clinical setting.

Entities:  

Keywords:  2D-CNN; Brain; Cancer; Ensemble; FDG-PET; S: Deep learning

Year:  2020        PMID: 31659587      PMCID: PMC7165199          DOI: 10.1007/s10278-019-00289-x

Source DB:  PubMed          Journal:  J Digit Imaging        ISSN: 0897-1889            Impact factor:   4.056


  21 in total

Review 1.  A survey on deep learning in medical image analysis.

Authors:  Geert Litjens; Thijs Kooi; Babak Ehteshami Bejnordi; Arnaud Arindra Adiyoso Setio; Francesco Ciompi; Mohsen Ghafoorian; Jeroen A W M van der Laak; Bram van Ginneken; Clara I Sánchez
Journal:  Med Image Anal       Date:  2017-07-26       Impact factor: 8.545

2.  Deep Learning at Chest Radiography: Automated Classification of Pulmonary Tuberculosis by Using Convolutional Neural Networks.

Authors:  Paras Lakhani; Baskaran Sundaram
Journal:  Radiology       Date:  2017-04-24       Impact factor: 11.105

Review 3.  Deep convolutional neural networks for brain image analysis on magnetic resonance imaging: a review.

Authors:  Jose Bernal; Kaisar Kushibar; Daniel S Asfaw; Sergi Valverde; Arnau Oliver; Robert Martí; Xavier Lladó
Journal:  Artif Intell Med       Date:  2018-09-06       Impact factor: 5.326

Review 4.  Positron emission tomography imaging of brain tumors.

Authors:  Terence Z Wong; Gert J van der Westhuizen; R Edward Coleman
Journal:  Neuroimaging Clin N Am       Date:  2002-11       Impact factor: 2.264

5.  Appropriate Use Criteria for 18F-FDG PET/CT in Restaging and Treatment Response Assessment of Malignant Disease.

Authors:  Hossein Jadvar; Patrick M Colletti; Roberto Delgado-Bolton; Giuseppe Esposito; Bernd J Krause; Andrei H Iagaru; Helen Nadel; David I Quinn; Eric Rohren; Rathan M Subramaniam; Katherine Zukotynski; Julie Kauffman; Sukhjeet Ahuja; Landis Griffeth
Journal:  J Nucl Med       Date:  2017-10-12       Impact factor: 10.057

6.  Rotational data augmentation for electroencephalographic data.

Authors:  Mario Michael Krell
Journal:  Conf Proc IEEE Eng Med Biol Soc       Date:  2017-07

Review 7.  Interpretive Error in Radiology.

Authors:  Stephen Waite; Jinel Scott; Brian Gale; Travis Fuchs; Srinivas Kolla; Deborah Reede
Journal:  AJR Am J Roentgenol       Date:  2016-12-27       Impact factor: 3.959

8.  Deep learning for staging liver fibrosis on CT: a pilot study.

Authors:  Koichiro Yasaka; Hiroyuki Akai; Akira Kunimatsu; Osamu Abe; Shigeru Kiryu
Journal:  Eur Radiol       Date:  2018-05-14       Impact factor: 5.315

9.  Deep convolutional neural network for segmentation of knee joint anatomy.

Authors:  Zhaoye Zhou; Gengyan Zhao; Richard Kijowski; Fang Liu
Journal:  Magn Reson Med       Date:  2018-05-17       Impact factor: 4.668

10.  Ankle Fracture Detection Utilizing a Convolutional Neural Network Ensemble Implemented with a Small Sample, De Novo Training, and Multiview Incorporation.

Authors:  Gene Kitamura; Chul Y Chung; Barry E Moore
Journal:  J Digit Imaging       Date:  2019-08       Impact factor: 4.056

View more
  3 in total

Review 1.  Application of artificial intelligence in brain molecular imaging.

Authors:  Satoshi Minoshima; Donna Cross
Journal:  Ann Nucl Med       Date:  2022-01-14       Impact factor: 2.668

Review 2.  Artificial intelligence for molecular neuroimaging.

Authors:  Amanda J Boyle; Vincent C Gaudet; Sandra E Black; Neil Vasdev; Pedro Rosa-Neto; Katherine A Zukotynski
Journal:  Ann Transl Med       Date:  2021-05

3.  Deep Learning Using Multiple Degrees of Maximum-Intensity Projection for PET/CT Image Classification in Breast Cancer.

Authors:  Kanae Takahashi; Tomoyuki Fujioka; Jun Oyama; Mio Mori; Emi Yamaga; Yuka Yashima; Tomoki Imokawa; Atsushi Hayashi; Yu Kujiraoka; Junichi Tsuchiya; Goshi Oda; Tsuyoshi Nakagawa; Ukihide Tateishi
Journal:  Tomography       Date:  2022-01-05
  3 in total

北京卡尤迪生物科技股份有限公司 © 2022-2023.