Search

Background and study aims Colorectal cancers (CRC) with deep submucosal invasion (T1b) could be metastatic lesions. However, endoscopic images of T1b CRC resemble those of mucosal CRCs (Tis) or with superficial invasion (T1a). The aim of this study was to develop an automatic computer-aided diagnosis (CAD) system to identify T1b CRC based on plain endoscopic images. Patients and methods In two hospitals, 1839 non-magnified plain endoscopic images from 313 CRCs (Tis 134, T1a 46, T1b 56, beyond T1b 37) with sessile morphology were extracted for training. A CAD system was trained with the data augmented by rotation, saturation, resizing and exposure adjustment. Diagnostic performance was assessed using another dataset including 44 CRCs (Tis 23, T1b 21) from a third hospital. CAD generated a probability level for T1b diagnosis for each image, and > 95 % of probability level was defined as T1b. Lesions with at least one image with a probability level > 0.95 were regarded as T1b. Primary outcome is specificity. Six physicians separately read the same testing dataset. Results Specificity was 87 % (95 % confidence interval: 66–97) for CAD, 100 % (85–100) for Expert 1, 96 % (78–100) for Expert 2, 61 % (39–80) for both gastroenterology trainees, 48 % (27–69) for Novice 1 and 22 % (7–44) for Novice 2. Significant differences were observed between CAD and both novices (P = 0.013, P = 0.0003). Other diagnostic values of CAD were slightly lower than of the two experts. Conclusions Specificity of CAD was superior to novices and possibly to gastroenterology trainees but slightly inferior to experts.

Background and study aims Delayed bleeding is believed to occur less frequently after cold snare polypectomy (CSP), but this has not been validated in clinical trials. This study aimed to compare rates of delayed bleeding after CSP and hot snare polypectomy (HSP). Patients and methods We conducted a multicenter, randomized controlled trial. Participants scheduled to undergo endoscopic resection of colorectal polyps ≤ 10 mm were enrolled and randomly assigned to CSP or HSP. Prophylactic clipping was performed at the endoscopists’ discretion. The primary outcome was delayed bleeding rate. Secondary outcomes included immediate bleeding rate and clipping rate. Sample size calculation showed that 451 patients were required in each arm. Results At the end of the study period decided in advance, 308 participants were recruited and an interim analysis was performed. A total of 273 patients (mean age 62.2 ± 8.8 years; 188 males) were analyzed, with 139 patients allocated to CSP and 134 to HSP. In total, 367 polyps were resected with CSP and 360 polyps with HSP. There were no significant differences in patient demographics or polyp characteristics. In per-patient-based analysis, delayed bleeding rates were 0.7 % after CSP and 0.7 % after HSP. Per-polyp analysis showed similar results (CSP: 0.3 % vs. HSP: 0.6 %). The immediate bleeding rate was significantly higher with CSP vs. HSP (54 % vs.14 %, P

Background and study aims Mycobacterium tuberculosis is often detected in the feces of patients with active pulmonary tuberculosis. However, no study has examined the small intestine using small bowel capsule endoscopy (SBCE). This study aimed to investigate intestinal abnormalities in patients with active pulmonary tuberculosis. Patients and methods SBCE was performed in sputum/feces smear-positive pulmonary tuberculosis patients. From December 2013 to November 2016, 15 patients were enrolled from a dedicated tuberculous ward. The primary outcome was intestinal abnormalities identified by SBCE. Results Fourteen patients (median age 81 years, range 29 – 91; 3 female) underwent SBCE to the terminal ileum. The video was not recorded for one patient who was excluded from the analysis. Intestinal lesions, including 5 annular ulcers in 4 patients, were observed in 64 % (9/14). In subgroup analysis, prevalence tended to be higher in patients undergoing SBCE within 1 month of anti-tuberculous therapy (P = 0.051). Distribution of small intestinal lesions tended to be distal. Four of five annular ulcers were located close to the ileocecal valve. Conclusion Prevalence of intestinal lesions in patients with pulmonary tuberculosis is higher than expected before (UMIN 000017292)

Background and study aims Topical peppermint oil prevents intestinal spasm, but can cause rebound spasm. Lidocaine hydrochloride, a local anesthetic, may work as an antispasmodic by blocking Na + channels. The aim of this study was to investigate the effect of topical lidocaine on the inhibition of colonic spasm during colonoscopy, compared with peppermint oil. Patients and methods A randomized, controlled double-blind trial was conducted in an academic endoscopy unit. Patients requiring endoscopic resection were randomly allocated to colonoscopy with topical administration of lidocaine (n = 30) or peppermint oil (n = 30). Similar vials containing different solutions were randomly numbered. Allocation was made based on the vial number. The solution used and the vial number were not revealed during the study. Two endoscopists performed all procedures using midazolam, without anticholinergic agents. When a pre-selected lesion was identified, the solution in the assigned vial was dispersed and the bowel observed for 5 minutes. The primary endpoint was the duration of spasm inhibition, and a secondary endpoint was the occurrence of rebound spasm stronger than before dispersion. Results There were no significant differences in patient demographics. Spasm was inhibited in almost all patients in both groups, with a similar median duration (lidocaine 227 sec vs. peppermint 212.5 sec, P = 0.508). In contrast, rebound spasm occurred less frequently in the lidocaine group (lidocaine 7 % vs. peppermint 47 %, P = 0.001). There were no adverse events or symptoms associated with administration of the solutions. Conclusions The inhibitory effect of lidocaine is not superior to peppermint oil. However, lidocaine significantly decreases the frequency of rebound spasms.

Background and study aims: Endoscopic submucosal dissection (ESD) has been developed to facilitate en bloc resection of large lesions. However, it is laborious to retrieve the large colorectal specimens. We propose a novel retrieval technique using a Valsalva maneuver, known as Tumor Extraction by Defecation (TED). Case series: A total of nine lesions (median size 88 mm, maximum 225 mm; proximal colon three, rectum six) that could not be easily retrieved using net forceps were subsequently removed by TED. The rectum was filled with water through the colonoscope. The patient then strained to evacuate the specimen, facilitated by an almost straight anorectal angle. All specimens were retrieved without fragmentation, within minutes. Histology was assessed appropriately, including an adenoma in two and mucosal cancer in seven. All cut margins were verified to be negative. No adverse events occurred. Conclusions: TED is a promising technique for retrieving large colorectal specimens after ESD.

Background and study aims: Bleeding after colonoscopic resection of pedunculated polyps cannot be easily predicted. The aims of this study were to evaluate the blood supply in pedunculated polyps and to clarify the optimal position on the polyp stalk for snare placement to prevent post-polypectomy hemorrhage. ]Patients and methods: In one institution, 11 pedunculated polyps from 11 patients were studied prospectively. All polyps were resected at the base of the stalk using a snare wire with electrocautery. Histologic axial sections from the apex and base of the stalk were examined with hematoxylin eosin and elastica stains. Elastica stains were used to identify blood vessels. The cross-sectional area of the stalk, total vessel area, maximum diameter of artery/arteriole lumen, number of thick (≥ 0.1 mm) vessels, and number of arteries/arterioles were measured in each section with image processing software. Wilcoxon signed-ranks test was used for comparison. Results: The median polyp diameter was 16 mm (range 7 to 24 mm) and median length of the stalk was 11 mm (range 7 to 23 mm). Two invasive cancers (T1) were included. The maximum diameter of the arterial/arteriolar lumen was greater at the base (P = 0.0044), whereas the ratio of the vessel area to the cross-section area was greater at the apex (P = 0.016). The number of thick vessels and arteries/arterioles were equivalent between apex and base. Conclusions: Morphometric study of the blood supply of pedunculated polyps confirmed that the optimal site for the excision of pedunculated polyps is in the middle of the stalk.

Background and study aims: Diagnostic performance of a computer-aided diagnosis (CAD) system for deep submucosally invasive (T1b) colorectal cancer was excellent, but the “regions of interest” (ROI) within images are not obvious. Class activation mapping (CAM) enables identification of the ROI that CAD utilizes for diagnosis. The purpose of this study was a quantitative investigation of the difference between CAD and endoscopists. Patients and methods: Endoscopic images collected for validation of a previous study were used, including histologically proven T1b colorectal cancers (n = 82; morphology: flat 36, polypoid 46; median maximum diameter 20 mm, interquartile range 15-25 mm; histological subtype: papillary 5, well 51, moderate 24, poor 2; location: proximal colon 26, distal colon 27, rectum 29). Application of CAM was limited to one white light endoscopic image (per lesion) to demonstrate findings of T1b cancers. The CAM images were generated from the weights of the previously fine-tuned ResNet50. Two expert endoscopists depicted the ROI in identical images. Concordance of the ROI was rated by intersection over union (IoU) analysis. Results: Pixel counts of ROIs were significantly lower using CAD (188.9K[x103] ± 109.1K) than by endoscopists (354.5K ± 223.6K; P < 0.0001) and mean ± standard deviation of the IoU was 0.203 ± 0.170. IoU was significantly higher in correctly identified lesions (n = 54, 0.234 ± 0.172) than incorrect ones (n=28, 0.144 ± 0.153, P = 0.0215). Concusions: IoU was larger in correctly diagnosed T1b colorectal cancers. Optimal annotation of the ROI may be the key to improving diagnostic sensitivity of CAD for T1b colorectal cancers.