Semantic segmentation of pancreatic medical images by using convolutional neural network. (March 2022)

• Record Type: Journal Article
• Authors: Huang, Mei-Ling; Wu, Yi-Zhen
• Is Part Of: Biomedical signal processing and control. Volume 73(2022)
• Publication Date: 2022-03
• Subjects: MBU-Net MobileNet-U-Net -- NIH National Institutes of Health -- CT Computed tomography -- MRI Magnetic resonance imaging -- ML Machine learning -- DL Deep learning -- CNNs Convolutional neural networks -- Dice Dice coefficient -- Loss weighted cross-entropy loss -- Jaccard Jaccard similarity coefficient -- AUC Area under the ROC Curve; CT image -- Convolutional neural network -- Semantic segmentation -- Pancreas; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2021.103458
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: The first study combines U-Net and MobileNet-V2 (MBU-Net) with data augmentation for the pancreatic image semantic segmentation. The performances of the proposed U-Net and MBU-Net models before and after data augmentation were compared. The performance evaluation considers Dice, Jaccard, AUC, Precision, Specificity, Recall and model parameters. The proposed MBU-Net exhibited higher Dice, Jaccard and Recall than proposed U-Net after data augmentation. The proposed MBU-Net model requires less number of parameters (6.30 M) during the training phase as compared to state-of-the-art models. Abstract: Pancreatic cancer is the most difficult-to-detect cancer with the highest fatality rate. Pancreas analysis through abdominal computed tomography (CT) is challenging because the pancreas possesses a complex background and blurred boundaries with other organs. Traditionally, the notation of pancreatic area requires manual semantic segmentation by professional radiologists, but this is time-consuming. The developments of computer-assisted diagnosis of deep convolutional neural networks on pancreas segmentation are successful but require heavy computational complexity. This research is dedicated to the semantic segmentation of pancreatic CT images using convolutional neural networks to achieve favorable performance on pancreas segmentation with low computational parameters. We propose MobileNet-U-Net (MBU-Net) by combining U-Net model with light-weight MobileNet-V2 network. The proposed MBU-Net is assessed on the NIH pancreas-CT dataset. The averages for dice coefficient, Jaccard similarity coefficient, AUC, precision, specificity, and recall are 82.87%, 70.97%, 90.54%, 89.29%, 99.95%, and 77.37%, respectively. Results have demonstrated that the proposed MBU-Net can effectively reduce the computational parameters and achieve comparable performance when compared to the state-of-the-art methods.