An UNet-Based Brain Tumor Segmentation Framework via Optimal Mass Transportation Pre-processing

Jia Wei Liao, Tsung Ming Huang*, Tiexiang Li, Wen Wei Lin, Han Wang, Shing Tung Yau

*此作品的通信作者

研究成果: 書貢獻/報告類型會議論文篇章

1 引文 斯高帕斯(Scopus)

摘要

This article aims to build a framework for MRI images of brain tumor segmentation using the deep learning method. For this purpose, we develop a novel 2-Phase UNet-based OMT framework to increase the ratio of brain tumors using optimal mass transportation (OMT). Moreover, due to the scarcity of training data, we change the density function by different parameters to increase the data diversity. For the post-processing, we propose an adaptive ensemble procedure by solving the eigenvectors of the Dice similarity matrix and choosing the result with the highest aggregation probability as the predicted label. The Dice scores of the whole tumor (WT), tumor core (TC), and enhanced tumor (ET) regions for online validation computed by SegResUNet were 0.9214, 0.8823, and 0.8411, respectively. Compared with random crop pre-processing, OMT is far superior.

原文英語
主出版物標題Brainlesion
主出版物子標題Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries - 8th International Workshop, BrainLes 2022, Held in Conjunction with MICCAI 2022, Revised Selected Papers
編輯Spyridon Bakas, Ujjwal Baid, Bhakti Baheti, Alessandro Crimi, Sylwia Malec, Monika Pytlarz, Maximilian Zenk, Reuben Dorent
發行者Springer Science and Business Media Deutschland GmbH
頁面216-228
頁數13
ISBN(列印)9783031338410
DOIs
出版狀態已發佈 - 2023
事件Proceedings of the 8th International MICCAI Brainlesion Workshop, BrainLes 2022 - Singapore, 新加坡
持續時間: 2022 9月 182022 9月 22

出版系列

名字Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
13769 LNCS
ISSN(列印)0302-9743
ISSN(電子)1611-3349

會議

會議Proceedings of the 8th International MICCAI Brainlesion Workshop, BrainLes 2022
國家/地區新加坡
城市Singapore
期間2022/09/182022/09/22

ASJC Scopus subject areas

  • 理論電腦科學
  • 一般電腦科學

指紋

深入研究「An UNet-Based Brain Tumor Segmentation Framework via Optimal Mass Transportation Pre-processing」主題。共同形成了獨特的指紋。

引用此