An Automated Self-Learning Quantification System to Identify Visible Areas in Capsule Endoscopy Images.

Visibility in capsule endoscopic images is presently evaluated through intermittent analysis of frames selected by a physician. It is thus subjective and not quantitative. A method to automatically quantify the visibility on capsule endoscopic images has not been reported. Generally, when designing automated image recognition programs, physicians must provide a training image; this process is called supervised learning. We aimed to develop a novel automated self-learning quantification system to identify visible areas on capsule endoscopic images. The technique was developed using 200 capsule endoscopic images retrospectively selected from each of three patients. The rate of detection of visible areas on capsule endoscopic images between a supervised learning program, using training images labeled by a physician, and our novel automated self-learning program, using unlabeled training images without intervention by a physician, was compared. The rate of detection of visible areas was equivalent for the supervised learning program and for our automatic self-learning program. The visible areas automatically identified by self-learning program correlated to the areas identified by an experienced physician. We developed a novel self-learning automated program to identify visible areas in capsule endoscopic images.

Evaluation of the Diagnostic Ability of Optical Enhancement System in Early Gastric Cancer Demarcation.

This study aimed to evaluate the utility of optical enhancement (OE) in early gastric cancer demarcation. Twenty lesions of early gastric cancer were examined by PENTAX endoscopy system with OE-1 and OE-2 functions. The areas of tumor demarcation identified by 12 evaluators (6 novice and 6 experienced) were compared to the corresponding correct areas determined by postoperative histopathology findings. The misdiagnosed scores that were the sums of false-positive and false-negative areas were compared. Color of one hundred pixels from the inside of the cancerous area and the outside of the cancerous area was expressed as three-dimensional RGB component vectors. The mean vectors and covariance matrixes were calculated and the Mahalanobis distance, indicative of color differences between two areas, was tested. Comparisons of the misdiagnosed score revealed that OE-1 was preferred over WL-1 for gastric cancer demarcation for all 12 evaluators (p = 0.008) and in novice evaluators (p = 0.026). OE-2 was not significantly different from WL-2 in all cases. OE-1 images gave significantly larger Mahalanobis distances, indicative of color differences, than WL-1 images (p = 0.002). It was demonstrated that the OE Mode 1 has a significant advantage over the white light mode in demarcation of early gastric cancer.

Efficacy of i-Scan Imaging for the Detection and Diagnosis of Early Gastric Carcinomas.

We determined comparative efficacy of i-Scan for detection and diagnosis of gastric cancer. Ten patients diagnosed with early gastric cancer based on histopathological findings were analyzed. White light and i-Scan moving images recorded from these patients in twin mode were separated into white light and i-Scan. Twelve endoscopists (three different skill levels) blinded to patient information evaluated the images. Correlation between demarcation accuracy and lesion brightness on still images was investigated. No significant differences were found in diagnostic accuracy between white light and i-Scan moving images for tumor detection rate (91.7% versus 90.8%, P = 0.777). Diagnostic accuracy of tumor size was comparable between novice and experienced endoscopists for i-Scan moving images (65.7% versus 71.1%, P = 0.528), whereas it was significantly lower for white light moving images (41.2% versus 79.5%, P = 0.019). Tumor demarcation accuracy was significantly better with white light than i-Scan still images (71.0% versus 65.8%, P = 0.033). Correlations between demarcation accuracy and brightness reached highs of 0.75 for white light and 0.89 for i-Scan imaging. Efficacy of i-Scan over that of white light imaging for detecting and diagnosing gastric cancer was not shown; however, the diagnostic capability of i-Scan can be improved if imaging conditions are optimized.

Isolation and identification of EG-VEGF/prokineticins as cognate ligands for two orphan G-protein-coupled receptors.

Endocrine gland-derived vascular endothelial growth factor (EG-VEGF, identical to prokineticin 1) is a novel peptide recently identified as a selective mitogen for endocrine gland endothelial cells. The present study demonstrates that EG-VEGF/prokineticin 1 and a peptide closely related to EG-VEGF, prokineticin 2, are cognate ligands of two orphan G-protein-coupled receptors designated ZAQ (=EG-VEGF/PK-R1) and I5E (=EG-VEGF/PK-R2). EG-VEGF/prokineticin 1 and prokineticin 2 induced a transient increase in intracellular calcium ion concentration ([Ca(2+)](i)) with nanomolar potency in Chinese hamster ovary (CHO) cells expressing EG-VEGF/PK-R1 and -R2 and bind to these cells with high affinity and with different receptor selectivity. EG-VEGF/prokineticins provoke rapid phosphorylation of p44/42 MAP kinase and DNA synthesis in the bovine adrenal capillary endothelial cells (BACE). The mRNAs of both EG-VEGF/PK-R1 and -R2 were expressed in BACE. The identification of the receptors for EG-VEGF/prokineticins may provide a novel molecular basis for the regulation of angiogenesis in endocrine glands.