The segmentation of bones in pelvic CT images based on extraction of key frames

Abstract

BackgroundBone segmentation is important in computed tomography (CT) imaging of the pelvis, which assists physicians in the early diagnosis of pelvic injury, in planning operations, and in evaluating the effects of surgical treatment. This study developed a new algorithm for the accurate, fast, and efficient segmentation of the pelvis.MethodsThe proposed method consists of two main parts: the extraction of key frames and the segmentation of pelvic CT images. Key frames were extracted based on pixel difference, mutual information and normalized correlation coefficient. In the pelvis segmentation phase, skeleton extraction from CT images and a marker-based watershed algorithm were combined to segment the pelvis. To meet the requirements of clinical application, physician’s judgment is needed. Therefore the proposed methodology is semi-automated.ResultsIn this paper, 5 sets of CT data were used to test the overlapping area, and 15 CT images were used to determine the average deviation distance. The average overlapping area of the 5 sets was greater than 94%, and the minimum average deviation distance was approximately 0.58 pixels. In addition, the key frame extraction efficiency and the running time of the proposed method were evaluated on 20 sets of CT data. For each set, approximately 13% of the images were selected as key frames, and the average processing time was approximately 2 min (the time for manual marking was not included).ConclusionsThe proposed method is able to achieve accurate, fast, and efficient segmentation of pelvic CT image sequences. Segmentation results not only provide an important reference for early diagnosis and decisions regarding surgical procedures, they also offer more accurate data for medical image registration, recognition and 3D reconstruction.

A Novel Hybrid Active Contour Model for Intracranial Tuberculosis MRI Segmentation Applications

“…The last two terms in (9) are length regularizing term and distance regularizing term which are equivalent to the last two terms in (6). The local fitting term LIF F can attract the contour to stop at the true image edges, which is the same as (7).…”

“…The last two terms in (14) are the length regularizing term and the distance regularizing term, where  and  are both nonnegative constants which are equivalent to the last two terms in (6). Additionally, the energy function gif  which is the same as the first two terms in (1), referred to as the global fitting energy term, is constructed as:…”

“…However, they usually require cultures and immunological tests of tissue and biofluids, which are time-consuming and may delay definitive management [6].…”

Quantification of uptake in pelvis F‐18 FLT PET‐CT images using a 3D localization and segmentation CNN

“…In the literature, several approaches to pelvis segmentation in CT scans have been reported. The methods range from traditional approaches, including watershed, 3 multiatlas and statistical shape 4,5 methods, to state‐of‐the‐art deep learning approaches 6,7 that utilize fully convolutional networks 8 (FCN) or its popular variant the U‐Net 9 . However, none of these methods were designed for or evaluated on image data with the variability in volume size and image quality as depicted in Figures 1 and 2.…”

Deep Learning to Segment Pelvic Bones: Large-scale CT Datasets and Baseline Models

“…Existing methods for pelvic bone segmentation from CT mostly use simple thresholding [1], region growing [33], and handcrafted models, which include deformable models [17,32], statistical shape models [30,19], watershed [35] and others [26,12,22,8,23,4]. These methods focus on local gray information and have limited accuracy due to the weak density differences between cortical and trabecular bones.…”