Search

We investigated the risk of colorectal cancer (CRC) in patients with Crohn’s disease (CD) using the claims data of the Korean National Health Insurance during 2006–2015. The data of 13,739 and 40,495 individuals with and without CD, respectively, were analyzed. Hazard ratios (HRs) were calculated using multivariate Cox proportional hazard regression tests. CRC developed in 25 patients (0.18%) and 42 patients (0.1%) of the CD and non-CD groups, respectively. The HR of CRC in the CD group was 2.07 (95% confidence interval (CI), 1.25–3.41). The HRs of CRC among men and women were 2.02 (95% CI 1.06–3.87) and 2.10 (95% CI, 0.96–4.62), respectively. The HRs of CRC in the age groups 0–19, 20–39, 40–59, and ≥60 years were 0.07, 4.86, 2.32, and 0.66, respectively. The HR of patients with late-onset CD (≥40 years) was significantly higher than that of those with early-onset CD (&lt, 40 years). CD patients were highly likely to develop CRC. Early-onset CD patients were significantly associated with an increased risk of CRC than matched individuals without CD. However, among CD patients, late-onset CD was significantly associated with an increased risk of CRC.

Computer-assisted analysis is expected to improve the reliability of videofluoroscopic swallowing studies (VFSSs), but its usefulness is limited. Previously, we proposed a deep learning model that can detect laryngeal penetration or aspiration fully automatically in VFSS video images, but the evidence for its reliability was insufficient. This study aims to compare the intra- and inter-rater reliability of the computer model and human raters. The test dataset consisted of 173 video files from which the existence of laryngeal penetration or aspiration was judged by the computer and three physicians in two sessions separated by a one-month interval. Intra- and inter-rater reliability were calculated using Cohen’s kappa coefficient, the positive reliability ratio (PRR) and the negative reliability ratio (NRR). Intrarater reliability was almost perfect for the computer and two experienced physicians. Interrater reliability was moderate to substantial between the model and each human rater and between the human raters. The average PRR and NRR between the model and the human raters were similar to those between the human raters. The results demonstrate that the deep learning model can detect laryngeal penetration or aspiration from VFSS video as reliably as human examiners.

Accurate measurement of polyp size during colonoscopy is crucial. The usefulness of cap-assisted colonoscopy and external grid application on monitor (gCAP) was evaluated for polyp size measurement in this 3-year, single-center, single-blind, randomized trial. Using the endoscopic forceps width as reference, the discrepancy percent (DP), error rate (ER), and measurement time were compared between gCAP and visual estimation (VE) after randomization. ER was calculated within a 20% and 33% limit. From the 111 patients, 280 polyps were measured. The mean polyp sizes were 4.0 ± 1.7 mm and 4.2 ± 1.8 mm with gCAP and VE, respectively (p = 0.368). Compared with that by the forceps method, DP was significantly lower in the gCAP group than in the VE group. Moreover, ER was significantly lower in the gCAP group within its preset limit. The measurement time was 4 s longer in the gCAP group than in the VE group (8.2 ± 4.8 s vs. 4.2 ± 1.5 s, p &lt, 0.001). However, the forceps method lasted 28 s longer than the others. On subgroup analysis by size, gCAP was more accurate for polyp size ≥5 mm. The gCAP method was more accurate for polyp size measurement than VE, especially for polyps ≥5 mm, and was more convenient than the forceps method.