Computer-aided detection of sclerotic bone metastases in the spine using watershed algorithm and support vector machines

Abstract

This work presents a computer-aided detection (CAD) system to aid radiologists in finding sclerotic bone metastases in the spine on CT images. The spine is first segmented using thresholding, region growing and a vertebra template. A watershed algorithm and a merging routine segment potential lesion candidates in each twodimensional (2-D) axial CT image. Next, overlapping 2-D detections on sequential CT slices are merged to form 3-D candidate lesions. For each of these, 30 quantitative features based on shape, density, and location are computed. After a feature filter eliminates clearly false candidates, a ground truth on 10 clinical cases segmented manually by an expert, and the features of each CAD candidate are used to train seven support vector machines. The segmentation algorithm detects 164 out of the 212 manually segmented lesions. A ten-fold cross-validation trained on these detections results in 77.4% sensitivity at an average of 9.44 false positives per case.

Improving Computer-Aided Detection UsingConvolutional Neural Networks and Random View Aggregation

“…In this study, we use three existing CADe systems that have previously been described in the literature: a) Detection of sclerotic spine metastases: we use a recent CADe method for detecting sclerotic metastases candidates from CT volumes [4], [33] (see Sec. III-D).…”

Artificial Intelligence in Musculoskeletal Imaging: A Paradigm Shift

“…(55,56) Radiologists may overlook spine compression fractures if they do not routinely review sagittal midline images on body CT. (57) In response, a system was designed for the automated detection and localization of thoracic and lumbar vertebral body compression fractures on CT. (15,16,(59)(60)(61) In one such system, the sensitivity (and false-positive rate per patient) was 81% (2.1), 81% (1.3), and 76% (2.1) for sclerotic, lytic, and mixed lesions of the spine, respectively, using SVM classifiers. 5).…”

“…2 and 3). (15,16,(59)(60)(61) In one such system, the sensitivity (and false-positive rate per patient) was 81% (2.1), 81% (1.3), and 76% (2.1) for sclerotic, lytic, and mixed lesions of the spine, respectively, using SVM classifiers. (27) This system is a first step toward the quantitative analysis of metastatic spine disease for determination of tumor burden, assessment of lesion change over time, and inclusion of bone lesions into treatment response criteria such as RECIST.…”

Detection of Sclerotic Spine Metastases via Random Aggregation of Deep Convolutional Neural Network Classifications

“…We use a state-of-the-art CADe method for detecting sclerotic metastases candidates from CT volumes (Burns et al, 2013, Wiese et al, 2011. The spine is initially segmented by thresholding at certain attenuation levels and performing region growing.…”