Integrating multiple MRI sequences for pelvic organs segmentation via the attention mechanism
Autor: | Lijuan Lai, Sijuan Huang, Junyun Li, Mengxue He, Tianyu Zeng, Zesen Cheng, Wanjia Zheng, Xiaoyan Huang, Xin Yang |
---|---|
Rok vydání: | 2021 |
Předmět: |
Male
Contouring Pelvic organ Jaccard index Computer science business.industry Pattern recognition Image processing General Medicine Pelvic cavity Magnetic Resonance Imaging Pelvis medicine.anatomical_structure Hausdorff distance Similarity (network science) medicine Image Processing Computer-Assisted Humans Segmentation Artificial intelligence Neural Networks Computer business Retrospective Studies |
Zdroj: | Medical physicsREFERENCES. 48(12) |
ISSN: | 2473-4209 |
Popis: | Purpose To create a network which fully utilizes multi-sequence MRI and compares favorably with manual human contouring. Methods We retrospectively collected 89 MRI studies of the pelvic cavity from patients with prostate cancer, and cervical cancer. The dataset contained 89 samples from 87 patients with a total of 84 valid samples. MRI was performed with T1-weighted (T1), T2-weighted (T2) and Enhanced Dixon T1-weighted (T1DIXONC) sequences. There were two cohorts. The training cohort contained 55 samples and the testing cohort 29 samples. The MRI images in the training cohort contained contouring data from radiotherapist α. The MRI images in the testing cohort contained contouring data from radiotherapist α and contouring data from another radiotherapist: radiotherapist β. The training cohort was used to optimize the CNNs, which included the attention mechanism through the proposed Activation Module and the Blended Module into multiple MRI sequences, to perform auto delineation. The testing cohort was used to assess the networks' auto delineation performance. The contoured organs at risk (OAR) were the anal canal, bladder, rectum, femoral head (L), femoral head (R). Results We compared our Proposed Network with UNet and FuseUNet using our dataset. When T1 was the main sequence, we input three sequences to segment five organs and evaluated the results using four metrics: the DSC (Dice Similarity Coefficient), the JSC (Jaccard Similarity Coefficient), the ASD (Average Mean Distance), and the 95% HD (Robust Hausdorff Distance). The Proposed Network achieved improved results compared with the baselines among all metrics. The DSC were 0.834±0.029, 0.818±0.037 and 0.808±0.050 for our Proposed Network, FuseUNet, and UNet, respectively. The 95% HD were 7.256mm±2.748mm, 8.404mm±3.297mm and 8.951mm±4.798mm for our Proposed Network, FuseUNet and, UNet, respectively. Our Proposed Network also had superior performance on the JSC and ASD coefficients. Conclusion Our proposed Activation Module and Blended Module significantly improved the performance of FuseUNet for multi-sequence MRI segmentation. Our Proposed Network integrated multiple MRI sequences efficiently and auto-segmented OAR rapidly and accurately. We also discovered that three-sequence fusion (T1-T1DIXONC-T2) was superior to two-sequence fusion (T1-T2, T1-T1DIXONC, respectively). We infer that the more MRI sequences fused the better the automatic segmentation results. This article is protected by copyright. All rights reserved. |
Databáze: | OpenAIRE |
Externí odkaz: |