End-To-End Vertebra Localization and Level Detection in Weakly Labelled 3D Spinal Mr using Cascaded Neural Networks

• Published in: 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019) pp. 1178 - 1182
• Main Authors: van Sonsbeek, Tom, Danaei, Pardiss, Behnami, Delaram, Jafari, Mohammad Hossein, Asgharzadeh, Parisa, Rohling, Robert, Abolmaesumi, Purang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.04.2019
• Subjects: deep learning; detection; Image segmentation; Imaging; Labeling; localization; neural networks; Slabs; Three-dimensional displays; Training; Two dimensional displays; Vertebra identification
• ISSN: 1945-8452
• DOI: 10.1109/ISBI.2019.8759280

Localization and identification of vertebrae in 3D MR volumes is a crucial first step for diagnosis and management of spinal conditions. Automating this process can save radiologists significant time and clicks. In this paper, we propose a novel learning-based approach consisting of two cascaded networks that perform simultaneous identification and localization of vertebrae. The first network performs slice-based level detection of full 3D sagittal volumes using an adaptive loss function that adjusts the weights of its loss terms during training, and outputs estimated center slices of each vertebrae. The sagittal slice is then divided into sub-volumes each containing a single vertebra. These sub-volumes are inputted into the second network for binary classification and localization of the vertebrae. Our method only requires centroid annotation (performed manually), a statistical model then provides an approximation of the volumetric segmentation for ground truth data. With this method, a vertebra identification rate of 82% was achieved.