Computed tomographic colonography (CTC) is a promising emerging technology for imaging of the colon. This concise review discusses the currently available data on CTC technique,test characteristics,acceptance,safety,c...Computed tomographic colonography (CTC) is a promising emerging technology for imaging of the colon. This concise review discusses the currently available data on CTC technique,test characteristics,acceptance,safety,cost-effectiveness,follow-up strategy,and extracolonic findings. In summary,CTC technique is still evolving,and further research is needed to clarify the role of automated colonic insufflation,smooth-muscle relaxants,intravenous and oral contrast,soft-ware rendering,and patient positioning. Currently,full bowel preparation is still required to achieve optimal results. The sensitivity for detecting large polyps (> 1 cm) can be as high as 85%,with specificity of up to 97%. These test characteristics are almost comparable to those of conventional colonoscopy. Patient acceptance of CTC is generally higher than that for colonoscopy,especially in patients who have never undergone either procedure. CTC is generally safe,although uncommon instances of colonic perforation have been documented. In terms of cost-effectiveness,most decision analyses have concluded that CTC would only be cost-effective if it were considerably cheaper than conventional colonoscopy. The proper follow-up strategy for small polyps or incidental extracolonic findings discovered during CTC is still under debate. At present,the exact clinical role of virtual colonoscopy still awaits determination. Even though widespread CTC screening is not available today,in the future there may eventually be a role for this technology. Technological advances in this area will undoubtedly continue,with multi-detector row CT scanners allowing thinner collimation and higher reso-lution images. Stool-tagging techniques are likely to evolve and may eventually allow for low-preparation CTC. Perceptual and fatigue-related reading errors can potentially be minimized with the help of computer-aided detection software. Further research will define the exact role of this promising technology in our diagnostic armamentarium.展开更多
Texture features have played an essential role in the field of medical imaging for computer-aided diagnosis.The gray-level co-occurrence matrix(GLCM)-based texture descriptor has emerged to become one of the most succ...Texture features have played an essential role in the field of medical imaging for computer-aided diagnosis.The gray-level co-occurrence matrix(GLCM)-based texture descriptor has emerged to become one of the most successful feature sets for these applications.This study aims to increase the potential of these features by introducing multi-scale analysis into the construction of GLCM texture descriptor.In this study,we first introduce a new parameter-stride,to explore the definition of GLCM.Then we propose three multi-scaling GLCM models according to its three parameters,(1)learning model by multiple displacements,(2)learning model by multiple strides(LMS),and(3)learning model by multiple angles.These models increase the texture information by introducing more texture patterns and mitigate direction sparsity and dense sampling problems presented in the traditional Haralick model.To further analyze the three parameters,we test the three models by performing classification on a dataset of 63 large polyp masses obtained from computed tomography colonoscopy consisting of 32 adenocarcinomas and 31 benign adenomas.Finally,the proposed methods are compared to several typical GLCM-texture descriptors and one deep learning model.LMS obtains the highest performance and enhances the prediction power to 0.9450 with standard deviation 0.0285 by area under the curve of receiver operating characteristics score which is a significant improvement.展开更多
文摘Computed tomographic colonography (CTC) is a promising emerging technology for imaging of the colon. This concise review discusses the currently available data on CTC technique,test characteristics,acceptance,safety,cost-effectiveness,follow-up strategy,and extracolonic findings. In summary,CTC technique is still evolving,and further research is needed to clarify the role of automated colonic insufflation,smooth-muscle relaxants,intravenous and oral contrast,soft-ware rendering,and patient positioning. Currently,full bowel preparation is still required to achieve optimal results. The sensitivity for detecting large polyps (> 1 cm) can be as high as 85%,with specificity of up to 97%. These test characteristics are almost comparable to those of conventional colonoscopy. Patient acceptance of CTC is generally higher than that for colonoscopy,especially in patients who have never undergone either procedure. CTC is generally safe,although uncommon instances of colonic perforation have been documented. In terms of cost-effectiveness,most decision analyses have concluded that CTC would only be cost-effective if it were considerably cheaper than conventional colonoscopy. The proper follow-up strategy for small polyps or incidental extracolonic findings discovered during CTC is still under debate. At present,the exact clinical role of virtual colonoscopy still awaits determination. Even though widespread CTC screening is not available today,in the future there may eventually be a role for this technology. Technological advances in this area will undoubtedly continue,with multi-detector row CT scanners allowing thinner collimation and higher reso-lution images. Stool-tagging techniques are likely to evolve and may eventually allow for low-preparation CTC. Perceptual and fatigue-related reading errors can potentially be minimized with the help of computer-aided detection software. Further research will define the exact role of this promising technology in our diagnostic armamentarium.
基金This work was supported by the NIH/NCI,No.CA206171.
文摘Texture features have played an essential role in the field of medical imaging for computer-aided diagnosis.The gray-level co-occurrence matrix(GLCM)-based texture descriptor has emerged to become one of the most successful feature sets for these applications.This study aims to increase the potential of these features by introducing multi-scale analysis into the construction of GLCM texture descriptor.In this study,we first introduce a new parameter-stride,to explore the definition of GLCM.Then we propose three multi-scaling GLCM models according to its three parameters,(1)learning model by multiple displacements,(2)learning model by multiple strides(LMS),and(3)learning model by multiple angles.These models increase the texture information by introducing more texture patterns and mitigate direction sparsity and dense sampling problems presented in the traditional Haralick model.To further analyze the three parameters,we test the three models by performing classification on a dataset of 63 large polyp masses obtained from computed tomography colonoscopy consisting of 32 adenocarcinomas and 31 benign adenomas.Finally,the proposed methods are compared to several typical GLCM-texture descriptors and one deep learning model.LMS obtains the highest performance and enhances the prediction power to 0.9450 with standard deviation 0.0285 by area under the curve of receiver operating characteristics score which is a significant improvement.