Cited 2 time in
Cited 1 time in
Generative approach for data augmentation for deep learning-based bone surface segmentation from ultrasound images
- Generative approach for data augmentation for deep learning-based bone surface segmentation from ultrasound images
- Zama, Asaduz; Park, Sang Hyun; Bang, Hyunhee; Park, Chul-woo; Park, Ilhyung; Joung, Sanghyun
- DGIST Authors
- Park, Sang Hyun
- Issue Date
- International Journal of Computer Assisted Radiology and Surgery
- Article Type
- Article; Early Access
- Author Keywords
- Cystic bone lesion; Ultrasound bone segmentation; Deep learning; GAN; Pix2Pix; Augmentation; Osteolytic bone tumor surgery
- Purpose: Precise localization of cystic bone lesions is crucial for osteolytic bone tumor surgery. Recently, there is a move toward ultrasound imaging over plain radiographs (X-rays) for intra-operative navigation due to the radiation-free and cost-effectiveness of the modality. In this process, the intra-operative bone model reconstructed from the segmented ultrasound image is registered with the pre-operative bone model. Deep learning approaches have recently shown remarkable success in bone surface segmentation from ultrasound images. However, to train deep learning models effectively with limited dataset size, data augmentation is essential. This study investigates the applicability of the generative approach for data augmentation as well as identifies standard data augmentation approaches for bone surface segmentation from ultrasound images. Methods: The generative approach we used in our work is based on Pix2Pix image-to-image translation network. We have proposed a multiple-snapshot approach, which mitigates the uni-modal deterministic output issue in the Pix2Pix network without using any complex architecture and training process. We also identified standard data augmentation approaches necessary for ultrasound bone surface segmentation through experiments. Results: We have evaluated our networks using 800 ultrasound images from trained regions (humerus bone) and 1200 ultrasound images from untrained regions (tibia and femur bones) using four different augmentation approaches. The results show that the generative augmentation approach has a positive impact on accuracy in both trained (+ 4.88%) and untrained regions (+ 25.84%) compared to using only standard augmentations. Moreover, compared to standard augmentation approaches, the addition of the generative augmentation approach also showed a similar trend in both trained (+ 8.74%) and untrained (+ 11.55%) regions. Conclusion: Generative approaches are very beneficial for data augmentation, where limited dataset size is prevalent, such as ultrasound bone segmentation. The proposed multiple-snapshot Pix2Pix approach has the potential to generate multimodal images, which enlarges the dataset considerably. © 2020, CARS.
- Springer Verlag
- Related Researcher
Park, Sang Hyun
Medical Image & Signal Processing Lab
컴퓨터비전, 인공지능, 의료영상처리
There are no files associated with this item.
- Department of Robotics EngineeringMedical Image & Signal Processing Lab1. Journal Articles
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.