Next Article in Journal
Age-Related Biology of Early-Stage Operable Breast Cancer and Its Impact on Clinical Outcome
Next Article in Special Issue
Distinct Cerebrovascular Reactivity Patterns for Brain Radiation Necrosis
Previous Article in Journal
Liver Transplantation for T2 Hepatocellular Carcinoma during the COVID-19 Pandemic: A Novel Model Balancing Individual Benefit against Healthcare Resources
Previous Article in Special Issue
[18F]FET PET Uptake Indicates High Tumor and Low Necrosis Content in Brain Metastasis
Article

Fine-Tuning Approach for Segmentation of Gliomas in Brain Magnetic Resonance Images with a Machine Learning Method to Normalize Image Differences among Facilities

1
Division of Molecular Modification and Cancer Biology, National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan
2
Cancer Translational Research Team, RIKEN Center for Advanced Intelligence Project, 1-4-1 Nihonbashi, Chuo-ku, Tokyo 103-0027, Japan
3
Department of Neurosurgery and Neuro-Oncology, National Cancer Center Hospital, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan
4
Division of Brain Tumor Translational Research, National Cancer Center Research Institute, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan
5
Department of Neurosurgery, Osaka University Graduate School of Medicine, 2-2 Yamadaoka, Suita, Osaka 565-0871, Japan
6
Department of Diagnostic Radiology, National Cancer Center Hospital, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan
7
Artificial Intelligence Research Center, National Institute of Advanced Industrial Science and Technology, 2-3-26, Aomi, Koto-ku, Tokyo 135-0064, Japan
8
Department of Neurosurgery, Graduate School of Medical Sciences, Kumamoto University, 1-1-1 Honjo, Chuo-ku, Kumamoto 860-8556, Japan
9
Department of Neurosurgery, Kyorin University Faculty of Medicine, 6-20-2, Sinkawa, Mitaka, Tokyo 181-8611, Japan
10
Department of Neurosurgery, Dokkyo Medical University, 880 Kitakobayashi, Mibu, Shimotsugagun, Tochigi 321-0293, Japan
11
Department of Neurosurgery, Tokyo Metropolitan Komagome Hospital, 3-18-22 Honkomagome, Bunkyo-ku, Tokyo 113-8677, Japan
12
Department of Neurosurgery, Faculty of Medicine, The University of Tokyo, 7-3-1 Hongo Bunkyo-ku, Tokyo 113-8655, Japan
13
Department of Neurosurgery, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan
14
Department of Neurosurgery, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8510, Japan
15
Department of Neurosurgery, Graduate School of Medicine, Yokohama City University, 3-9 Fukuura, Kanazawa-ku, Yokohama, Kanagawa 236-0004, Japan
16
Department of Neuro-Oncology/Neurosurgery, Saitama Medical University International Medical Center, 397-1 Yamane, Hidaka, Saitama 350-1298, Japan
17
Department of Neurosurgery, Kansai Medical University, 2-5-1 Shinmachi, Hirakata, Osaka 573-1010, Japan
18
Department of Neurosurgery, National Hospital Organization Osaka National Hospital, 2-1-14 Hoenzaka, Chuo-ku, Osaka 540-0006, Japan
19
Department of Neurosurgery, Osaka City University Graduate School of Medicine, 1-4-3 Asahi-machi, Abeno-ku, Osaka 545-8585, Japan
20
Department of Neurological Surgery, Wakayama Medical University School of Medicine Wakayama, 811-1 Kimiidera, Wakayama 641-8509, Japan
21
Department of Neurosurgery, Osaka International Cancer Institute, 3-1-69 Ootemae, Chuo-ku, Osaka 541-8567, Japan
22
Department of Neurosurgery, Kindai University Faculty of Medicine, 377-2 Ohnohigashi, Osaka-Sayama, Osaka 589-8511, Japan
23
Department of Biomedical Research and Innovation, Institute for Clinical Research, National Hospital Organization Osaka National Hospital, 2-1-14 Hoenzaka, Chuo-ku, Osaka 540-0006, Japan
24
Humanome Lab, 2-4-10 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Academic Editor: Jorn Fierstra
Cancers 2021, 13(6), 1415; https://doi.org/10.3390/cancers13061415
Received: 27 December 2020 / Revised: 14 March 2021 / Accepted: 16 March 2021 / Published: 19 March 2021
(This article belongs to the Special Issue Advanced Neuroimaging Approaches for Malignant Brain Tumors)
This study evaluates the performance degradation of machine learning models for segmenting gliomas in brain magnetic resonance images caused by domain shift and proposed possible solutions. Although machine learning models exhibit significant potential for clinical applications, performance degradation in different cohorts is a problem that must be solved. In this study, we identify the impact of the performance degradation of machine learning models to be significant enough to render clinical applications difficult. This demonstrates that it can be improved by fine-tuning methods with a small number of cases from each facility, although the data obtained appeared to be biased. Our method creates a facility-specific machine learning model from a small real-world dataset and public dataset; therefore, our fine-tuning method could be a practical solution in situations where only a small dataset is available.
Machine learning models for automated magnetic resonance image segmentation may be useful in aiding glioma detection. However, the image differences among facilities cause performance degradation and impede detection. This study proposes a method to solve this issue. We used the data from the Multimodal Brain Tumor Image Segmentation Benchmark (BraTS) and the Japanese cohort (JC) datasets. Three models for tumor segmentation are developed. In our methodology, the BraTS and JC models are trained on the BraTS and JC datasets, respectively, whereas the fine-tuning models are developed from the BraTS model and fine-tuned using the JC dataset. Our results show that the Dice coefficient score of the JC model for the test portion of the JC dataset was 0.779 ± 0.137, whereas that of the BraTS model was lower (0.717 ± 0.207). The mean Dice coefficient score of the fine-tuning model was 0.769 ± 0.138. There was a significant difference between the BraTS and JC models (p < 0.0001) and the BraTS and fine-tuning models (p = 0.002); however, no significant difference between the JC and fine-tuning models (p = 0.673). As our fine-tuning method requires fewer than 20 cases, this method is useful even in a facility where the number of glioma cases is small. View Full-Text
Keywords: glioma; machine learning; MR images; fine-tuning; deep learning glioma; machine learning; MR images; fine-tuning; deep learning
Show Figures

Figure 1

MDPI and ACS Style

Takahashi, S.; Takahashi, M.; Kinoshita, M.; Miyake, M.; Kawaguchi, R.; Shinojima, N.; Mukasa, A.; Saito, K.; Nagane, M.; Otani, R.; Higuchi, F.; Tanaka, S.; Hata, N.; Tamura, K.; Tateishi, K.; Nishikawa, R.; Arita, H.; Nonaka, M.; Uda, T.; Fukai, J.; Okita, Y.; Tsuyuguchi, N.; Kanemura, Y.; Kobayashi, K.; Sese, J.; Ichimura, K.; Narita, Y.; Hamamoto, R. Fine-Tuning Approach for Segmentation of Gliomas in Brain Magnetic Resonance Images with a Machine Learning Method to Normalize Image Differences among Facilities. Cancers 2021, 13, 1415. https://doi.org/10.3390/cancers13061415

AMA Style

Takahashi S, Takahashi M, Kinoshita M, Miyake M, Kawaguchi R, Shinojima N, Mukasa A, Saito K, Nagane M, Otani R, Higuchi F, Tanaka S, Hata N, Tamura K, Tateishi K, Nishikawa R, Arita H, Nonaka M, Uda T, Fukai J, Okita Y, Tsuyuguchi N, Kanemura Y, Kobayashi K, Sese J, Ichimura K, Narita Y, Hamamoto R. Fine-Tuning Approach for Segmentation of Gliomas in Brain Magnetic Resonance Images with a Machine Learning Method to Normalize Image Differences among Facilities. Cancers. 2021; 13(6):1415. https://doi.org/10.3390/cancers13061415

Chicago/Turabian Style

Takahashi, Satoshi, Masamichi Takahashi, Manabu Kinoshita, Mototaka Miyake, Risa Kawaguchi, Naoki Shinojima, Akitake Mukasa, Kuniaki Saito, Motoo Nagane, Ryohei Otani, Fumi Higuchi, Shota Tanaka, Nobuhiro Hata, Kaoru Tamura, Kensuke Tateishi, Ryo Nishikawa, Hideyuki Arita, Masahiro Nonaka, Takehiro Uda, Junya Fukai, Yoshiko Okita, Naohiro Tsuyuguchi, Yonehiro Kanemura, Kazuma Kobayashi, Jun Sese, Koichi Ichimura, Yoshitaka Narita, and Ryuji Hamamoto. 2021. "Fine-Tuning Approach for Segmentation of Gliomas in Brain Magnetic Resonance Images with a Machine Learning Method to Normalize Image Differences among Facilities" Cancers 13, no. 6: 1415. https://doi.org/10.3390/cancers13061415

Find Other Styles
Note that from the first issue of 2016, MDPI journals use article numbers instead of page numbers. See further details here.

Article Access Map by Country/Region

1
Back to TopTop