A multi-resolution approach for spinal metastasis detection using deep Siamese neural networks

• Published in: Computers in biology and medicine Vol. 84; pp. 137 - 146
• Main Authors: Wang, Juan, Fang, Zhiyuan, Lang, Ning, Yuan, Huishu, Su, Min-Ying, Baldi, Pierre
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.05.2017; Elsevier Limited
• Subjects: Accuracy; Adults; Aged; Algorithms; Artificial neural networks; Automation; Bioinformatics; Bone cancer; Bone marrow; Breast cancer; Cancer; Cardiovascular disease; Classification; Compression; Computed tomography; Computer programs; Computer vision; Data processing; Deep learning; Diagnosis; Feasibility studies; Female; Humans; Image Interpretation, Computer-Assisted - methods; Image processing; Information processing; Internal Medicine; Intervertebral discs; Kidneys; Lesions; Lungs; Machine learning; Magnetic resonance imaging; Magnetic Resonance Imaging - methods; Male; Mammography; Mathematical models; Medical imaging; Membranes; Metastases; Metastasis; Methods; Middle Aged; Multi-resolution analysis; Neural networks; Neural Networks (Computer); NMR; Nuclear magnetic resonance; Other; Pain; Predictions; Probability theory; Quality of life; ROC Curve; Siamese neural network; Solubility; Spatial resolution; Speech; Spinal cancer; Spinal cord; Spinal metastasis; Spinal Neoplasms - diagnostic imaging; Spinal Neoplasms - secondary; Thorax; Tumors; Vertebrae; Visual perception; United States > US; Deep learning; Multi-resolution analysis; Spinal metastasis; Siamese neural network; Magnetic resonance imaging
• DOI: 10.1016/j.compbiomed.2017.03.024
• ISSN: 0010-4825; 1879-0534

Spinal metastasis, a metastatic cancer of the spine, is the most common malignant disease in the spine. In this study, we investigate the feasibility of automated spinal metastasis detection in magnetic resonance imaging (MRI) by using deep learning methods. To accommodate the large variability in metastatic lesion sizes, we develop a Siamese deep neural network approach comprising three identical subnetworks for multi-resolution analysis and detection of spinal metastasis. At each location of interest, three image patches at three different resolutions are extracted and used as the input to the networks. To further reduce the false positives (FPs), we leverage the similarity between neighboring MRI slices, and adopt a weighted averaging strategy to aggregate the results obtained by the Siamese neural networks. The detection performance is evaluated on a set of 26 cases using a free-response receiver operating characteristic (FROC) analysis. The results show that the proposed approach correctly detects all the spinal metastatic lesions while producing only 0.40 FPs per case. At a true positive (TP) rate of 90%, the use of the aggregation reduces the FPs from 0.375 FPs per case to 0.207 FPs per case, a nearly 44.8% reduction. The results indicate that the proposed Siamese neural network method, combined with the aggregation strategy, provide a viable strategy for the automated detection of spinal metastasis in MRI images. •A multi-resolution approach is proposed to detect spinal metastasis in MRI.•The multi-resolution approach is implemented using deep Siamese neural networks.•A slice-based aggregation method is used to minimize the number of false positives.•The proposed approach detects spinal metastasis accurately and effectively.