A computer-aided diagnosis system using white-light endoscopy for the prediction of conventional adenoma with high grade dysplasia.

OBJECTIVES: We developed a computer-aided diagnosis system called ECRCCAD using standard white-light endoscopy (WLE) for predicting conventional adenomas with high-grade dysplasia (HGD) to optimise the patients' management decisions during colonoscopy. METHODS: Pretraining model was used to fine-tune the model parameters by transfer learning. 2,397 images of HGD and 2,487 low-grade dysplasia (LGD) images were randomly assigned (8:1:1) to the training, optimising, and internal validation dataset. The prospective validation dataset is the frames accessed from colonoscope videoes. One independent rural hospital provided an external validation dataset. Histopathological diagnosis was used as the standard criterion. The capability of the ECRCCAD to distinguish HGD was assessed and compared with two expert endoscopists. RESULTS: The accuracy, sensitivity and specificity for diagnosis of HGD in the internal validation set were 90.5%, 93.2%, 87.9%, respectively. While 88.2%, 85.4%, 89.8%, respectively, for the external validation set. For the prospective validation set, ECRCCAD achieved an AUC of 93.5% in diagnosing HGD. The performance of ECRCCAD in diagnosing HGD was better than that of the expert endoscopist in the external validation set (88.2% vs. 71.5%, P < 0.0001). CONCLUSION: ECRCCAD had good diagnostic capability for HGD and enabled a more convenient and accurate diagnosis using WLE.

Accurate measurement of colorectal polyps using computer-aided analysis.

INTRODUCTION: As we know, the majority of colorectal cancers are thought to evolve from colorectal adenomas. In this study, we explored the use of Computer-aided diagnosis (CAD) in the detection of colorectal polyps and the estimation of their sizes, which is important for the diagnosis and management of colorectal cancer. MATERIALS AND METHODS: As the distance between colonoscopy and lesion increases, magnification tends to decrease. Therefore, the size of colorectal polyps can be calculated by taking into account the captured image and the shooting distance. In this study, the fitting curve of the magnification of electronic colonoscopy was obtained by simulating intestinal tract and polyps in vitro. Then, the distance was artificially controlled in the endoscopic operation, and the image was taken at a preset distance. The CAD system was then trained on the overall shape of colorectal polyps. Image segmentation was employed to accurately identify colorectal polyps. Finally, on the basis of the magnification factor, the real value of polyps was predicted from the shooting distance and the segmentation image size. RESULTS: The CAD system can automatically calculate the range of colorectal polyps and calculate the true size of the colorectal polyps according to the magnification of the corresponding distance. CONCLUSIONS: In this study, we developed a method of accurately estimating the size of colorectal polyps. This approach is compatible with many devices, which would expand its range of applications. This method has the potential for application in other areas of clinical diagnosis.

[Diagnosis of stress urinary incontinence using poster urethrovesical angel measured by ultrasonography].

OBJECTIVE: To evaluate the clinical value of ultrasonography detecting the poster urethrovesical angle (PUVA) in diagnosis of stress urinary incontinence (SUI). METHODS: From Jan. 2006 to Dec. 2008, the PUVA in resting phase (PUVA-r) and stress phase (PUVA-s) between 84 patients with SUI and 100 healthy women were measured by color Doppler ultrasound, including 57 cases in SUI degree I, 22 cases in SUI degree II, 5 cases in SUI degree III. RESULTS: (1) The PUVA-r and PUVA-s were (130+/-29) degrees and (158+/-36) degrees in SUI patients, which were significantly higher (113+/-19) degrees and (115+/-23) degrees in control group (P<0.01). (2) Correlation analysis showed that PUVA-s was positively correlated with the severity of SUI (P<0.01). (3) The area under the receiver operating characteristic curves (ROC) of PUVA-s was 0.82 (95% CI: 0.76-0.89) between SUI patients and healthy women. When PUVA-s>or=140 degrees was chosed as the cut-off value, the specificity, accuracy rate and positive predictive value in diagnosis of SUI were 84%, 79%, 81%, which were significantly higher than those when cut-off value of PUVA-s>or=120 degrees (54%, 68%, 62%; P<0.05). CONCLUSION: The PUVA-s in SUI patients are significantly increased and is positively correlated with the severity of SUI, which are indicated that PUVA-s>or=140 degrees should be used as cut-off value in diagnosis of SUI by ultrasonography.