Lumbar Spine Tumor Segmentation and Localization in T2 MRI Images Using AI

TL;DR

Limited annotated data hampers accurate segmentation, localization, and classification of lumbar spine tumors in MRI. The authors propose a CSF-guided data augmentation combined with a hybrid AI pipeline that fuses fuzzy c-means segmentation, Random Forest refinement, a six-layer CNN classifier, and 3D vertebral labeling for localization. The approach achieves about 99% segmentation accuracy, 98% tumor-type classification accuracy, and 99% localization accuracy, with IoU and Dice metrics exceeding prior methods. This work enhances diagnostic precision and supports clinical decision-making and surgical planning in spinal oncology.

Abstract

In medical imaging, segmentation and localization of spinal tumors in three-dimensional (3D) space pose significant computational challenges, primarily stemming from limited data availability. In response, this study introduces a novel data augmentation technique, aimed at automating spine tumor segmentation and localization through AI approaches. Leveraging a fusion of fuzzy c-means clustering and Random Forest algorithms, the proposed method achieves successful spine tumor segmentation based on predefined masks initially delineated by domain experts in medical imaging. Subsequently, a Convolutional Neural Network (CNN) architecture is employed for tumor classification. Moreover, 3D vertebral segmentation and labeling techniques are used to help pinpoint the exact location of the tumors in the lumbar spine. Results indicate a remarkable performance, with 99% accuracy for tumor segmentation, 98% accuracy for tumor classification, and 99% accuracy for tumor localization achieved with the proposed approach. These metrics surpass the efficacy of existing state-of-the-art techniques, as evidenced by superior Dice Score, Class Accuracy, and Intersection over Union (IOU) on class accuracy metrics. This innovative methodology holds promise for enhancing the diagnostic capabilities in detecting and characterizing spinal tumors, thereby facilitating more effective clinical decision-making.

Figures (12)

• Figure 1: Types of Spine Tumor.
• Figure 2: Flowchart for Proposed Data Augmentation Technique.
• Figure 3: Overview of Proposed Tumor Segmentation Method.
• Figure 4: Proposed CNN Architecture for Tumour Classification.
• Figure 5: Illustration of Tumor Localization in Lumbar Spine.