Bone Segmentation and Fracture Detection in Ultrasound Using 3D Local Phase Features

Abstract

Abstract. 3D ultrasound (US) is increasingly considered as a viable alternative imaging modality in computer-assisted orthopaedic surgery (CAOS) applications. Automatic bone segmentation from US images, however, remains a challenge due to speckle noise and various other artifacts inherent to US. In this paper, we present intensity invariant three dimensional (3D) local image phase features, obtained using 3D Log-Gabor filter banks, for extracting ridge-like features similar to those that occur at soft tissue/bone interfaces. Our contributions include the novel extension of 2D phase symmetry features to 3D and their use in automatic extraction of bone surfaces and fractured fragments in 3D US. We validate our technique using phantom, in vitro, and in vivo experiments. Qualitative and quantitative results demonstrate remarkably clear segmentations results of bone surfaces with a localization accuracy of better than 0.62mm and mean errors in estimating fracture displacements below 0.65mm, which will likely be of strong clinical utility.

Towards Real-Time 3D US to CT Bone Image Registration Using Phase and Curvature Feature Based GMM Matching

“…The consistent SRE demonstrates the ability of GMM registration to reliably match surfaces; the improved FRE for tests including K indicates that curvature features improve registration when other surface features are too similar. The average run-time of the registration of all tests was less than 2.11 s. It should be noted that the PS surface has been reported to be biased slightly inwards relative to the bone surface (on the order of 0.4 mm) [9], which likely contributes slightly to the FRE.…”

“…Recently, 3D local phase information has been proposed to extract the surfaces of bones in US volumes [9] using a 3D Log-Gabor filter, LG(ω) to extract local phase information in the B-Mode US volume:…”

Confidence Weighted Local Phase Features for Robust Bone Surface Segmentation in Ultrasound

“…Furthermore, in spine imaging, US bone surface extraction was shown to be promising for needle-insertion applications [5]. Bone surface extraction based on the detection of symmetric response features was recently shown to be a powerful approach [6,7]. However, local symmetry features remain prone to false detection of soft-tissue interfaces that often exhibit features similar to those of bone.…”

Robust Bone Detection in Ultrasound Using Combined Strain Imaging and Envelope Signal Power Detection

“…The associated radiation exposure posing risks to both patients and surgical teams gave rise to recent interest in safer non-ionizing real-time intraoperative imaging alternatives such as US [1].…”

“…during fragment positioning in fracture reduction surgeries [1]. Despite recent advancement in US intensity-based automatic bone segmentation, results remain unpredictable due to the high levels of speckle noise, reverberation, and signal drop out [1].…”