Bone SPECT/CT in Oncology

AbstractThe aim of this study was to evaluate the quantitative values of short-time scan (STS) of metastatic lesions compared with a standard scan (SS) when acquired by whole-body bone SPECT/CT with cadmium–zinc–telluride (CZT) detectors. We retrospectively reviewed 13 patients with bone metastases from prostate cancer, who underwent SPECT/CT performed on whole-body CZT gamma cameras. STSs were obtained using 75, 50, 25, 10, and 5% of the list-mode data for SS, respectively. Regions of interest (ROIs) were set on the increased uptake areas diagnosed as metastases. Intraclass correlation coefficients (ICCs) of standardized uptake values (SUVs) for the ROIs were calculated between the SS and each STS, and ICC ≥ 0.8 was set as a perfect correlation. Moreover, the repeatability coefficient (RC) was calculated, and RC ≤ 20% was defined as acceptable. A total of 152 metastatic lesions were included in the analysis. The ICCs between the SS vs. 75%-STS, 50%-STS, 25%-STS, 10%-STS, and 5%-STS were 0.999, 0.997, 0.994, 0.983, and 0.955, respectively. The RCs of the SS vs. 75%-STS, 50%-STS, 25%-STS, 10%-STS, and 5%-STS were 7.9, 12.4, 19.8, 30.8, and 41.3%, respectively. When evaluating the quality of CZT bone SPECT/CT acquired by a standard protocol, 25%-STS may provide adequate quantitative values.

Download Full-text

Deep learning based automatic segmentation of metastasis hotspots in thorax bone SPECT images

PLoS ONE ◽

10.1371/journal.pone.0243253 ◽

2020 ◽

Vol 15 (12) ◽

pp. e0243253

Author(s):

Qiang Lin ◽

Mingyang Luo ◽

Ruiting Gao ◽

Tongtong Li ◽

Zhengxing Man ◽

...

Keyword(s):

Deep Learning ◽

Fine Tuning ◽

Disease Stage ◽

Whole Body ◽

Learning Technology ◽

Automatic Assessment ◽

Original Dataset ◽

Bone Spect ◽

Segmentation Models ◽

Spect Images

SPECT imaging has been identified as an effective medical modality for diagnosis, treatment, evaluation and prevention of a range of serious diseases and medical conditions. Bone SPECT scan has the potential to provide more accurate assessment of disease stage and severity. Segmenting hotspot in bone SPECT images plays a crucial role to calculate metrics like tumor uptake and metabolic tumor burden. Deep learning techniques especially the convolutional neural networks have been widely exploited for reliable segmentation of hotspots or lesions, organs and tissues in the traditional structural medical images (i.e., CT and MRI) due to their ability of automatically learning the features from images in an optimal way. In order to segment hotspots in bone SPECT images for automatic assessment of metastasis, in this work, we develop several deep learning based segmentation models. Specifically, each original whole-body bone SPECT image is processed to extract the thorax area, followed by image mirror, translation and rotation operations, which augments the original dataset. We then build segmentation models based on two commonly-used famous deep networks including U-Net and Mask R-CNN by fine-tuning their structures. Experimental evaluation conducted on a group of real-world bone SEPCT images reveals that the built segmentation models are workable on identifying and segmenting hotspots of metastasis in bone SEPCT images, achieving a value of 0.9920, 0.7721, 0.6788 and 0.6103 for PA (accuracy), CPA (precision), Rec (recall) and IoU, respectively. Finally, we conclude that the deep learning technology have the huge potential to identify and segment hotspots in bone SPECT images.

Download Full-text