Automatic Lumbar MRI Detection and Identification Based on Deep Learning
Tóm tắt
The aim of this research is to automatically detect lumbar vertebras in MRI images with bounding boxes and their classes, which can assist clinicians with diagnoses based on large amounts of MRI slices. Vertebras are highly semblable in appearance, leading to a challenging automatic recognition. A novel detection algorithm is proposed in this paper based on deep learning. We apply a similarity function to train the convolutional network for lumbar spine detection. Instead of distinguishing vertebras using annotated lumbar images, our method compares similarities between vertebras using a beforehand lumbar image. In the convolutional neural network, a contrast object will not update during frames, which allows a fast speed and saves memory. Due to its distinctive shape, S1 is firstly detected and a rough region around it is extracted for searching for L1–L5. The results are evaluated with accuracy, precision, mean, and standard deviation (STD). Finally, our detection algorithm achieves the accuracy of 98.6% and the precision of 98.9%. Most failed results are involved with wrong S1 locations or missed L5. The study demonstrates that a lumbar detection network supported by deep learning can be trained successfully without annotated MRI images. It can be believed that our detection method will assist clinicians to raise working efficiency.
Tài liệu tham khảo
citation_journal_title=Bmc Musculoskelet Disord; citation_title=Percutaneous endoscopic lumbar discectomy: minimally invasive technique for multiple episodes of lumbar disc herniation; citation_author=KC Choi, JS Kim, CL Dong; citation_volume=18; citation_publication_date=2017; citation_pages=329-324; citation_doi=10.1186/s12891-017-1697-8; citation_id=CR1
citation_journal_title=Acta Neurologica Belgica; citation_title=Treatment of cauda equina syndrome caused by lumbar disc herniation with percutaneous endoscopic lumbar discectomy; citation_author=X Li, Q Dou, S Hu, J Liu, Q Kong, J Zeng, Y Song; citation_volume=116; citation_issue=2; citation_publication_date=2016; citation_pages=185-190; citation_doi=10.1007/s13760-015-0530-0; citation_id=CR2
citation_journal_title=J Digit Imaging; citation_title=Identification of Apical Vertebra for Grading of Idiopathic Scoliosis using Image Processing; citation_author=H Anitha, GK Prabhu; citation_volume=25; citation_issue=1; citation_publication_date=2012; citation_pages=155-161; citation_doi=10.1007/s10278-011-9394-x; citation_id=CR3
citation_journal_title=J Digit Imaging; citation_title=Improving Visibility of Stereo-Radiographic Spine Reconstruction with Geometric Inferences; citation_author=S Kumar, KP Nayak, KS Hareesha; citation_volume=29; citation_issue=2; citation_publication_date=2016; citation_pages=226-234; citation_doi=10.1007/s10278-015-9841-1; citation_id=CR4
citation_journal_title=J Digit Imaging; citation_title=Automatic estimation of orientation and position of spine in digitized X-rays using mathematical morphology; citation_author=VP Kumar, T Thomas; citation_volume=18; citation_issue=3; citation_publication_date=2005; citation_pages=234-241; citation_doi=10.1007/s10278-005-5150-4; citation_id=CR5
citation_journal_title=Journal of Digital Imaging; citation_title=Spine localization in X-ray images using interest point detection; citation_author=M Benjelloun, S Mahmoudi; citation_volume=22; citation_issue=3; citation_publication_date=2009; citation_pages=309-318; citation_doi=10.1007/s10278-007-9099-3; citation_id=CR6
citation_journal_title=Science; citation_title=Reducing the dimensionality of data with neural networks; citation_author=GE Hinton, R Salakhutdinov; citation_volume=313; citation_issue=5786; citation_publication_date=2006; citation_pages=504-507; citation_doi=10.1126/science.1127647; citation_id=CR7
citation_journal_title=Comput Methods Prog Biomed; citation_title=Fully automatic cervical vertebrae segmentation framework for X-ray images; citation_author=SMMR Al Arif, K Knapp, G Slabaugh; citation_volume=157; citation_publication_date=2018; citation_pages=95-111; citation_doi=10.1016/j.cmpb.2018.01.006; citation_id=CR8
citation_journal_title=Comput Med Imaging Graph; citation_title=Vertebrae localization in CT using both local and global symmetry features; citation_author=K Kim, S Lee; citation_volume=58; citation_publication_date=2017; citation_pages=45-55; citation_doi=10.1016/j.compmedimag.2017.02.002; citation_id=CR9
citation_journal_title=IEEE Trans Biomed Eng; citation_title=Simultaneous Localization of Lumbar Vertebrae and Intervertebral Discs With SVM-Based MRF; citation_author=AB Oktay, YS Akgul; citation_volume=60; citation_issue=9; citation_publication_date=2013; citation_pages=2375-2383; citation_doi=10.1109/TBME.2013.2256460; citation_id=CR10
citation_journal_title=Neuro informatics; citation_title=Automated pathogenesis-based diagnosis of lumbar neural foraminal stenosis via deep multiscale multitask learning; citation_author=Z Han, B Wei, S Leung; citation_volume=1; citation_publication_date=2018; citation_pages=1-13; citation_id=CR11
citation_journal_title=Comput Biol Med; citation_title=A multi-resolution approach for spinal metastasis detection using deep Siamese neural networks; citation_author=J Wang, Z Fang, N Lang, H Yuan, MY Su, P Baldi; citation_volume=84; citation_issue=C; citation_publication_date=2017; citation_pages=137-146; citation_doi=10.1016/j.compbiomed.2017.03.024; citation_id=CR12
citation_journal_title=Energy Convers Manag; citation_title=Neural network based method for conversion of solar radiation data; citation_author=AN Celik, T Muneer; citation_volume=67; citation_issue=1; citation_publication_date=2013; citation_pages=117-124; citation_doi=10.1016/j.enconman.2012.11.010; citation_id=CR13
citation_journal_title=J Digit Imaging; citation_title=Detection and labeling of vertebrae in mr images using deep learning with clinical annotations as training data; citation_author=D Forsberg, E Sjöblom, JL Sunshine; citation_volume=30; citation_issue=4; citation_publication_date=2017; citation_pages=1-7; citation_doi=10.1007/s10278-017-9945-x; citation_id=CR14
Gao J, Ling H, Hu W et al.: Transfer Learning Based Visual Tracking with Gaussian Processes Regression. Computer Vision – ECCV 2014. Springer International Publishing, 2014, pp. 188–203
Bertinetto L, Valmadre J, Henriques J F, et al. Fully-Convolutional Siamese Networks for Object Tracking. European Conference on Computer Vision – ECCV2016, 2016:850–865.
Krizhevsky A, Sutskever I, Hinton G E. ImageNet classification with deep convolutional neural networks. International Conference on Neural Information Processing Systems. Curran Associates Inc. 60(2):1097–1105, 2012.
Russakovsky O, Deng J, Su H, Krause J, Satheesh S, Ma S, Huang Z, Karpathy A, Khosla A, Bernstein M, Berg AC, Fei-Fei L: ImageNet Large Scale Visual Recognition Challenge. IJCV, 2015
citation_journal_title=Computerized Medical Imaging and Graphics; citation_title=Multimodal vertebrae recognition using transformed deep convolution network; citation_author=Y Cai, M Landis, DT Laidley, A Kornecki, A Lum, S Li; citation_volume=51; citation_publication_date=2016; citation_pages=11-19; citation_doi=10.1016/j.compmedimag.2016.02.002; citation_id=CR19
Suzani A, Seitel A, Liu Y, et al. Fast Automatic Vertebrae Detection and Localization in Pathological CT Scans - A Deep Learning Approach. Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, 9353:678–686, 2015.