Dilated-Inception Net: Multi-Scale Feature Aggregation for Cardiac Right Ventricle Segmentation
| Autor: | Jingcong Li, Zhenghui Gu, Zhu Liang Yu, Yuanqing Li, Hui Liu |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Adult
Male Scale (ratio) Computer science Heart Ventricles 0206 medical engineering Feature extraction Biomedical Engineering 02 engineering and technology Deep Learning Image Interpretation Computer-Assisted medicine Humans Segmentation Aged medicine.diagnostic_test business.industry Cardiac Ventricle Magnetic resonance imaging Pattern recognition Image segmentation Middle Aged Magnetic Resonance Imaging 020601 biomedical engineering Cardiac Imaging Techniques medicine.anatomical_structure Ventricle Feature (computer vision) Female Artificial intelligence business Algorithms |
| Zdroj: | IEEE Transactions on Biomedical Engineering. 66:3499-3508 |
| ISSN: | 1558-2531 0018-9294 |
| Popis: | Segmentation of cardiac ventricle from magnetic resonance images is significant for cardiac disease diagnosis, progression assessment, and monitoring cardiac conditions. Manual segmentation is so time consuming, tedious, and subjective that automated segmentation methods are highly desired in practice. However, conventional segmentation methods performed poorly in cardiac ventricle, especially in the right ventricle. Compared with the left ventricle, whose shape is a simple thick-walled circle, the structure of the right ventricle is more complex due to ambiguous boundary, irregular cavity, and variable crescent shape. Hence, effective feature extractors and segmentation models are preferred. In this paper, we propose a dilated-inception net (DIN) to extract and aggregate multi-scale features for right ventricle segmentation. The DIN outperforms many state-of-the-art models on the benchmark database of right ventricle segmentation challenge. In addition, the experimental results indicate that the proposed model has potential to reach expert-level performance in right ventricular epicardium segmentation. More importantly, DIN behaves similarly to clinical expert with high correlation coefficients in four clinical cardiac indices. Therefore, the proposed DIN is promising for automated cardiac right ventricle segmentation in clinical applications. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|