The Application of Knowledge Distillation Toward Fine-Grained Segmentation for Three-Vessel View of Fetal Heart Ultrasound Images
Qiwen Cai,Ran Chen,Lu Li,Chao Huang,Haisu Pang,Yuanshi Tian,Min Di,Mingxuan Zhang,Mingming Ma,Dexing Kong,Bowen Zhao
DOI: https://doi.org/10.1155/2022/1765550
IF: 3.12
2022-01-01
Computational Intelligence and Neuroscience
Abstract:Objectives. Measuring anatomical parameters in fetal heart ultrasound images is crucial for the diagnosis of congenital heart disease (CHD), which is highly dependent on the clinical experience of the sonographer. To address this challenge, we propose an automated segmentation method using the channel-wise knowledge distillation technique. Methods. We design a teacher-student architecture to conduct channel-wise knowledge distillation. ROI-based cropped images and full-size images are used for the teacher and student models, respectively. It allows the student model to have both the fine-grained segmentation capability inherited from the teacher model and the ability to handle full-size test images. A total of 1,300 fetal heart ultrasound images of three-vessel view were collected and annotated by experienced doctors for training, validation, and testing. Results. We use three evaluation protocols to quantitatively evaluate the segmentation accuracy: Intersection over Union (IoU), Pixel Accuracy (PA), and Dice coefficient (Dice). We achieved better results than related methods on all evaluation metrics. In comparison with DeepLabv3+, the proposed method gets more accurate segmentation boundaries and has performance gains of 1.8% on mean IoU (66.8% to 68.6%), 2.2% on mean PA (79.2% to 81.4%), and 1.2% on mean Dice (80.1% to 81.3%). Conclusions. Our segmentation method could identify the anatomical structure in three-vessel view of fetal heart ultrasound images. Both quantitative and visual analyses show that the proposed method significantly outperforms the related methods in terms of segmentation results.