Digestive endoscopy : official journal of the Japan Gastroenterological Endoscopy Society最新文献

Background: Rapid on-site evaluation enhances the diagnostic yield of endoscopic ultrasound-guided fine needle aspiration (EUS-FNA), but the cytopathologists are limited. This study aimed to assess the diagnostic capability of an artificial intelligence-assisted ROSE (AI-ROSE) for EUS-FNA.

Methods: The study included 137 patients who underwent EUS-FNA of a pancreatic mass between April 2019 and August 2021. Participants were divided into training (n = 96), validation (n = 15), and test cohorts (n = 26). From the training/validation cohort, 5157/615 digital images of cell clusters were extracted and divided into 288 × 288-pixel patches. These cell clusters were annotated and labeled into five classes. A semantic segmentation architecture was developed. From the test cohort, 120 cell clusters were extracted to compare diagnostic performance between AI-ROSE and 21 endosonographers and 5 cytotechnologists with varying experience levels in ROSE.

Results: In total, 1,097,840 training, 31,817 validation, and 1920 test regions were extracted. In the test cohort, AI-ROSE accuracy for three-category classification (class 1/2, class 3, and class 4/5) was 89.8%. For two-category classification (class 1/2/3 and class 4/5), sensitivity, specificity, and accuracy were 89.3%, 98.1%, and 95.1%, respectively. In the comparison cohort, AI-ROSE accuracy for two-category classification was 93.3%, significantly higher than all endosonographers (68.3%; range, 45.8%-86.7%) and cytotechnologists (76.3%; range, 72.5%-78.3%). The AI-ROSE evaluation time for 120 cell clusters was 6.04 s, much shorter than that of all endosonographers (1800; 480-6000 s) and cytotechnologists (2160; 1020-3600 s).

Conclusions: The AI-ROSE model shows remarkable speed and accuracy in diagnosing pancreatic cell clusters, enabling rapid decision-making during EUS-FNA. UMIN-CTR; No. 000042212.

Objectives: Intestinal metaplasia (IM) is a premalignant lesion within the columnar-lined esophagus (CLE). The detection of IM stratifies the risk of cancer progression. We aim to investigate the random biopsy yield of IM in Asian patients with CLE.

Methods: A total of 117,526 consecutive esophagogastroduodenoscopies (EGDs) in the outpatient setting were retrospectively screened. The histopathological findings independently reviewed by two pathologists were correlated with the clinical and endoscopic features and the number of biopsies.

Results: Among 5963 EGDs, 3135 patients (5.1%) were diagnosed with CLE. A total of 4675 EGDs involving 8994 biopsies (median = 1, range 1-9) exhibited a 35.1% yield of IM per endoscopy. The yield of IM was higher in cases of long-segment CLE (74.8%) compared with short-segment (32.9%). IM detection was significantly associated with male gender (odds ratio, 95% confidence interval = 1.58, 1.37-1.83), chronological age (1.02, 1.02-1.03), longer segment CLE (1.75, 1.58-1.94), and number of biopsies per endoscopy (1.40, 1.32-1.49). One biopsy during each EGD resulted in a 26% IM yield, which was significantly lower than the yields after three biopsies (2.85, 2.25-3.62) and after following the Seattle protocol (2.86, 2.34-3.50). In 1772 patients without IM on index endoscopy, subsequent endoscopies increased the IM detection rate by 6%, with 4% attributed to the second endoscopy and 2% to the third through eighth endoscopy.

Conclusions: This study provides insights into the number of biopsies needed to characterize IM in Barrett's esophagus. Further research is needed to optimize biopsy strategies, particularly in Asian countries.

"Green Endoscopy" is now a worldwide movement that has captured the energy and imagination of the endoscopy community. It describes the effort to lessen the environmental impact of clinical practice by altering equipment, clinical pathways, and purchasing. While many suggestions or recommendations have been put forward, there is still much to be done in order to understand and regulate these attempts, to avoid unintended consequences and further environmental harm. This review highlights current thinking as well as fundamentals of environmental science and analysis, pertaining to our rapidly growing field.

Objectives: White opaque substance (WOS) is an endoscopic finding related to histological grade among superficial nonampullary duodenal epithelial tumors (SNADETs). However, the significance of WOS distribution and chronological changes is unknown.

Methods: Images of two different esophagogastroduodenoscopy (EGD) examinations before treatment for SNADETs were evaluated for the chronological change of WOS distribution. WOS distribution was classified into None, Partial, Whole and Marginal according to the degree of WOS. The frequency of chronological change and their relationship to lesion characteristics, including histological grade and epithelial subtypes (intestinal or gastric), were analyzed.

Results: Among 187 lesions, 27 were gastric-types and 160 were intestinal-types, 130 were low-grade adenoma (LGA) and 57 were high-grade adenoma or adenocarcinoma. The median interval between the EGDs was 49 days. Chronological change was more frequently observed in intestinal-types compared with gastric-types (26% vs. 5%, p = 0.01). Gastric-types mostly showed None WOS with no chronological change. Univariate analysis showed that tumor diameter of 10 mm or smaller, intestinal-type, interval between EGD, and LGA were related to chronological change. Multivariate analyses showed that intestinal-type and LGA were significant factors (p < 0.05). The most frequent change was None to Marginal (33%) among intestinal-type lesions with chronological change of WOS distribution.

Conclusion: This study revealed that WOS easily changes among LGA, especially in small intestinal-type lesions. As WOS is a variable finding, diagnosing histologic grade and epithelial subtypes of SNADETs should be based not only on the distribution of WOS, but also on a comprehensive assessment, including white-light endoscopy and image-enhanced endoscopy with magnification.