Deep neural network for automated simultaneous intervertebral disc (IVDs) identification and segmentation of multi-modal MR images. (June 2021)

• Record Type: Journal Article
• Authors: Das, Pabitra; Pal, Chandrajit; Acharyya, Amit; Chakrabarti, Amlan; Basu, Saumyajit
• Is Part Of: Computer methods and programs in biomedicine. Volume 205(2021)
• Publication Date: 2021-06
• Subjects: Deep learning -- Convolutional neural networks -- Intervertebral disc -- Segmentation -- Identification -- Region-to-image matching (RIM); Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2021.106074
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: In this paper, we introduced for the first time a novel deep neural network architecture coined as 'RIMNet', a Region-to-Image Matching Network model, performing automated simultaneous IVD identification and segmentation for multi-modality MRI images in only 12 seconds. Compared our results with the state-of-the-art methods on MICCAI 2018 IVD challenges, consisting of 16 sets of multi-modality MRI images. The Region-to-image Matching strategy has been applied during the training phase to integrate the multi-modality information more efficiently, improving the network learning capability. Proposed model has attained 94% identification accuracy, dice coefficient value 91.7 ± 1 % in segmentation and MDOC 90.2 ± 1 %. It also achieved 0.87 ± 0.02 for Jaccard Coefficient, 0.54 ± 0.04 for ASD and 0.62 ± 0.02 mm Hausdorff Distance. Abstract: Background and objective: Lower back pain in humans has become a major risk. Classical approaches follow a non-invasive imaging technique for the assessment of spinal intervertebral disc (IVDs) abnormalities, where identification and segmentation of discs are done separately, making it a time-consuming phenomenon. This necessitates designing a robust automated and simultaneous IVDs identification and segmentation of multi-modality MRI images. Methods: We introduced a novel deep neural network architecture coined as 'RIMNet', a Region-to-Image Matching Network model, capable of performing an automated and simultaneous IVDs identification and segmentation of MRI images. The multi-modal input data is being fed to the network with a dropout strategy, by randomly disabling modalities in mini-batches. The performance accuracy as a function of the testing dataset was determined. The execution of the deep neural network model was evaluated by computing the IVDs Identification Accuracy, Dice coefficient, MDOC, Average Symmetric Surface Distance, Jaccard Coefficient, Hausdorff Distance and F1 Score. Results: Proposed model has attained 94% identification accuracy, dice coefficient value of 91.7 ± 1 % in segmentation and MDOC 90.2 ± 1 % . Our model also achieved 0.87 ± 0.02 for Jaccard Coefficient, 0.54 ± 0.04 for ASD and 0.62 ± 0.02 mm Hausdorff Distance. The results have been validated and compared with other methodologies on dataset of MICCAI IVD 2018 challenge. Conclusions: Our proposed deep-learning methodology is capable of performing simultaneous identification and segmentation on IVDs MRI images of the human spine with high accuracy.