Artificial intelligence for the assessment of bowel preparation

560 Risk of Early Post-Procedure Bleeding With Percutaneous Endoscopic Gastrostomy (PEG) Tube Placement in Patients Using Clopidogrel

Colonoscopy is a diagnostic and therapeutic procedure that reduces colorectal cancer incidence and mortality but requires adequate bowel cleansing for high-quality examination. Past studies have suggested cirrhosis as a risk factor for worse bowel preparation. We carried out a match-controlled retrospective study evaluating patients with and without cirrhosis who underwent outpatient screening colonoscopies to assess the effect of cirrhosis and portal hypertension complications on preparation quality and endoscopic measures. We also did a subgroup analysis excluding patients with obesity. We examined 1464 patients with cirrhosis and matched controls. Cirrhotic patients had lower mean Boston Bowel Preparation Scale (BBPS) scores and slower cecal intubation times. We found a single point increase in the Model for End-stage Liver Disease (MELD) score, as well as ascites, hepatic encephalopathy, and variceal hemorrhage were all associated with a longer cecal intubation time. Subgroup analysis excluding patients with obesity again found a significantly lower BBPS score and longer cecal intubation time while also finding a 24% drop in polyp detection. Patients with cirrhosis have worse BBPS scores and longer cecal intubation times. Nonobese cirrhotic patients additionally have a lower polyp detection rate. Portal hypertension complications were associated with worsened preparation quality and longer cecal intubation times. Each incremental increase in MELD score lengthened cecal intubation time. These findings support a more aggressive bowel preparation strategy for patients with cirrhosis, especially patients with severe disease or portal hypertension complications.

Adequate bowel preparation is crucial for effective screening colonoscopy. However, it is unclear whether higher bowel preparation scores correspond to beneficial effects on the adenoma and polyp detection rate (ADR and PDR) in the adequate bowel preparation group. This study aimed to evaluate the effects of bowel preparation, according to the Boston Bowel Preparation Scale (BBPS), and colonoscopy withdrawal time (CWT) on ADR and PDR in the adequate bowel preparation group. Healthy examinees between 50 and 75years old who underwent colonoscopy between September 2015 and August 2016 were included. BBPS scores, CWT, ADR, and PDR were reviewed retrospectively. Predictors of ADR and PDR were analyzed with a generalized linear mixed model. A total of 5073 cases with adequate bowel preparation (BBPS ≥ 6) were analyzed. Examinees with good (BBPS = 6, 7) and excellent (BBPS = 8, 9) bowel preparation were 1898 (37.4%) and 3175 (62.6%), respectively. Both ADR and PDR were higher in the good bowel preparation group than in the excellent bowel preparation group (ADR 47.3% vs. 45.0%, P = 0.035; PDR 73.7% vs. 69.5%, P = 0.004, respectively). In the multivariate analysis, CWT, rather than BBPS, was significantly associated with both ADR (OR 1.04; 95% CI 1.02-1.06; P < 0.001) and PDR (OR 1.05; 95% CI 1.02-1.07; P = 0.002). Both ADR and PDR were lower when bowel preparation was excellent rather than good. However, CWT, not BBPS, was significantly associated with ADR and PDR in the adequate bowel preparation group. Therefore, meticulous inspection is important for high-quality colonoscopy regardless of the BBPS score in examinees with adequate bowel preparation.

Smartphone Application to Reinforce Education Increases High-Quality Preparation for Colorectal Cancer Screening Colonoscopies in a Randomized Trial

Background and AimBowel preparation is a crucial factor affecting the diagnostic accuracy of colonoscopy, and few randomized control trials evaluated enhancement in bowel preparation. In this study, we aimed to evaluate the effectiveness of walking exercises on bowel preparation before a colonoscopy procedure.MethodsThe present study is a single-blind randomized controlled trial involving 262 patients scheduled for colonoscopy procedures. These patients were randomly assigned to two groups: an intervention group (n = 131) and a control group (n = 131). In the intervention group, participants followed a predetermined plan that included the consumption of specific liquids and foods, bisacodyl pills, polyethylene glycol powder, and a regimen of walking exercises in preparation for their colonoscopy. Conversely, individuals in the control group followed the same regimen but were not instructed to engage in walking exercises. On the day of the colonoscopy, both groups were assessed for their level of physical activity using a foot counter. Additionally, an experienced gastroenterologist evaluated and compared the bowel preparation between the two groups using the Boston Bowel Preparation Scale (BBPS).ResultsThe number of footsteps recorded in the two groups exhibited a significant difference (P < 0.001). Although there was no statistically significant difference between the intervention and control groups in terms of mean BBPS scores (6.26 ± 1.9 vs. 6.29 ± 1.9, P = 0.416), individuals who took more than 6900 steps had significantly higher BBPS scores compared to those with fewer than 6900 footsteps (6.62 ± 1.8 vs. 5.92 ± 1.9, P = 0.003).In the univariate analysis, BBPS was found to be significantly associated with individuals under the age of 50 (OR: 2.45, 95% CI: 1.30–4.61, P = 0.006) and smoking status (OR: 0.41, 95% CI: 0.17–0.94, P = 0.043). In the multivariate analysis, the relationship between BBPS and age below 50 and smoking remained significant (OR: 2.50, 95% CI: 1.30–4.70, P = 0.005, and OR: 0.38, 95% CI: 0.16–0.93, P = 0.034, respectively).ConclusionA higher number of footsteps taken especially more than 6900 can significantly enhance bowel preparation; however, walking exercise as an intervention before colonoscopy is not significantly associated with BBPS. Also, older people and smokers seem to have fewer benefits from walking exercises for bowel preparation.Trial registrationISRCTN32724024 (Registration date:22/08/2018).

The performance of existing image-based training models in evaluating bowel preparation on colonoscopy videos was relatively low, and only a few models used external data to prove their generalization. Therefore, this study attempted to develop a more precise and stable AI system for assessing bowel preparation of colonoscopy video. We proposed a system named ViENDO to assess the bowel preparation quality, including two CNNs. First, Information-Net was used to identify and filter out colonoscopy video frames unsuitable for Boston bowel preparation scale (BBPS) scoring. Second, BBPS-Net was trained and tested with 5,566 suitable short video clips through three-dimensional (3D) convolutional neural network (CNN) technology to detect BBPS-based insufficient bowel preparation. Then, ViENDO was applied to complete withdrawal colonoscopy videos from multiple centers to predict BBPS segment scores in clinical settings. We also conducted a human-machine contest to compare its performance with endoscopists. In video clips, BBPS-Net for determining inadequate bowel preparation generated an area under the curve of up to 0.98 and accuracy of 95.2%. When applied to full-length withdrawal colonoscopy videos, ViENDO assessed bowel cleanliness with an accuracy of 93.8% in the internal test set and 91.7% in the external dataset. The human-machine contest demonstrated that the accuracy of ViENDO was slightly superior compared to most endoscopists, though no statistical significance was found. The 3D-CNN-based AI model showed good performance in evaluating full-length bowel preparation on colonoscopy video. It has the potential as a substitute for endoscopists to provide BBPS-based assessments during daily clinical practice.

Complete colonoscopy is a difficult task in patients with gastric surgeries

BackgroundColorectal cancer (CRC) remains a global health concern, underscoring the importance of effective bowel preparation for diagnostic procedures such as colonoscopy. This randomized controlled trial was designed to compare the efficacy and tolerability of orally administered mannitol and polyethylene glycol (PEG) solutions for repeat colonoscopy in patients with inadequate bowel preparation.MethodsThis prospective, open-label, noninferiority, blinded end point, randomized controlled clinical trial enrolled patients from two centers. Participants aged 18–75 years with Boston Bowel Preparation Scale (BBPS) scores ≤ 1 in any colon segment were included. The study employed a 1-L mannitol or PEG solution, with primary outcomes assessed using BBPS scores. The secondary outcomes included adverse events, taste preferences, and patient satisfaction. A simple randomization strategy was employed, and both intention-to-treat (ITT) and per-protocol (PP) analyses were conducted.ResultsA total of 134 patients were included in the study, and the trial demonstrated the noninferiority of mannitol compared to PEG in achieving adequate bowel preparation (difference 3.0%, 95% confidence interval − 5.0 to 11.0%). Mannitol exhibited favorable trends in BBPS scores, highlighting its effectiveness. The safety profiles of the two solutions were comparable, with a low incidence of adverse events (17.9% vs. 19.4%, P = 0.825). Notably, mannitol outperformed PEG in terms of patient satisfaction and a greater inclination for reuse.ConclusionThe findings suggest that orally administered mannitol is not inferior to orally administered PEG in patients with inadequate bowel preparation.Registration numberClinicalTrials.gov, NCT05912114. Registered on 12/6/2023.

The quality of bowel preparation is one of the key determinants of a successful colonoscopy. Bowel preparation regimens have evolved greatly over the past few decades, with attempts to improve the efficiency and tolerability; still an ideal agent or regimen continues to be oblivious. To compare the efficacy, safety, and tolerance of three bowel preparation regimens for colonoscopy: split dose of oral sulfate solution (OSS), split dose of polyethylene glycol (PEG), and same-day single dose PEG. This study was a randomized, single-blind control design with three study groups. Group A received a split dose of OSS, group B received a split dose of PEG, and Group C received a single dose of PEG for bowel preparation. The quality of preparation was assessed using the Boston Bowel Preparation Scale (BBPS), and the adverse effects and tolerance were noted. The data were compared statistically for any significant difference between the regimens. Mean total BBPS scores were 8.08, 7.52, and 7.92 for groups A, B, and C, respectively (P=0.076). Segmental BBPS scores were statistically similar for the right and transverse colon but differed for the left colon (A: B: C=2.79: 2.54: 2.75; P<0.01). Gastrointestinal side effects and electrolyte disturbances were similar across the three groups. Split-dose preparations were associated with more significant sleep disturbances than single-dose PEG (P<0.001). Patients who received OSS reported more taste intolerance (P<0.01), while those who received single PEG reported more volume intolerance (P<0.001). Split-dose OSS, split-dose PEG, and single-dose PEG regimens provide adequate and comparable bowel preparation for colonoscopy with good patient tolerance and no significant adverse effects. Overnight PEG and OSS preparations were associated with more substantial sleep disturbances. OSS is associated with more taste intolerance, while single PEG is associated with more volume intolerance.

BackgroundBowel cleanliness is a critical factor affecting the detection of adenomatous polyps and early tumors. The Boston Bowel Preparation Scale (BBPS), a widely used evaluation tool, has limitations, including interobserver variability and insufficient standardized training. This study aims to develop an artificial intelligence-driven automatic BBPS scoring and teaching system.MethodsColonoscopy image and video data were collected from three centers between June 2019 and August 2024, categorized into different BBPS scores (0, 1, 2, 3), ileocecal part, and instrument operation frames. Transfer learning and fine-tuning were performed on four pre-trained YOLOv11 models. Performance metrics included accuracy, precision, sensitivity, and AUC. Grad-CAM was used to provide visual explanations of the best-performing model, which was further developed into a system capable of real-time and cumulative BBPS assessment for every video frame.ResultsAmong the four models, YOLOv11m performed the best, achieving an accuracy of 99.86%, precision of 99.74%, sensitivity of 99.74%, and an F1 score of 99.75% on the validation set. On the test set, the model attained a weighted average precision of 95.37%, specificity of 98.25%, and an AUC of 0.996. Based on this model, the AutoBBPS system was developed, which automatically initiates real-time cumulative BBPS scoring once the cecum is reached. In image-level human-machine comparison experiments, the system outperformed junior endoscopists in recognition accuracy and was comparable to senior endoscopists. Video-level human-machine comparison experiments further evaluated the accuracy of the AutoBBPS system against endoscopists under varying confidence thresholds.ConclusionThe AutoBBPS system, developed using YOLOv11, provides real-time and cumulative BBPS scoring for every video frame, effectively assisting endoscopists in improving scoring efficiency and accuracy. Additionally, the intelligent BBPS teaching assistant is particularly beneficial for junior endoscopists, promoting standardized training and enhancing overall scoring quality.

Aims Radio-opaque markers have been widely used in the study of colon motility in patients with chronic functional constipation (FC). Here, we evaluate the relationship between the colon transit time (CTT) and the Boston Bowel Preparation Scale (BBPS) to determine whether CTT is a sufficient predictor of bowel preparation in patients with chronic functional constipation. Methods A total of fifty-six patients with constipation and fifty-two healthy controls (HC) were enrolled in this study. All subjects underwent the colonic transit study using radio-opaque markers and were given a follow-up colonoscopy examination on day 3 to 7 to determine BBPS. The correlation between total and segmental CTT and BBPS was evaluated, and risk factors for predicting inadequate bowel preparation were determined. Results In our study, we found some distinct outcomes compared with previous studies. The mean total CTT (TCTT) was determined to be 43.37 ± 18.82 h in the FC group and 23.08 ± 10.18 h in the HC group. This difference was found to be significant for both the total and segmental CTTs between the two groups (P < 0.05). Further, TCTT was negatively correlated with BBPS both in the FC (r = −0.899, 95% CI -0.748 to -0.925, P < 0.001) and the HC (r = −0.978, 95% CI -0.854 to -1.003, P = 0.004) groups, as was segmental CTTs and segmental BBPS (P < 0.05). In the case of patients with slow transit constipation, multivariate logistic regression analysis indicated that prolonged TCTT (OR 0.722, 95% CI 0.589-0.885, P = 0.002) was independently associated with poor bowel preparation. The total and right to left CTTs were found to predict inadequate bowel preparation and exhibited the best sensitivity and specificity at 48.0 h, 15.5 h, 17.5 h, and 19.0 h, based on ROC curve analysis. Conclusions The CTT test represents a valuable method for predicting the level of bowel preparation prior to a colonoscopy examination. That is, both total and segmental CTTs can be considered an objective predictor of bowel preparation prior to colonoscopy. The present study demonstrates some distinct results relative to previous studies, including STC subtype proportion in FC, the proportion of inadequate bowel preparation in the STC subtype, and the cut-off value of TCTT for predicting inadequate bowel preparation.

Inadequate inpatient bowel preparation can lead to repeat procedures, prolonged hospital stays, and increased financial burden. In this quality improvement project, the authors developed an electronic medical record (EMR)-based order set with precise instructions and anticipatory guidance for inpatient bowel preparation before colonoscopy. The current study is a nonrandomized intervention study. The authors compared 2 groups: an intervention group using a newly developed, consensus-based, standardized EMR bowel preparation order set and a control group using previously existing EMR bowel preparation orders. Bowel preparation outcomes were followed over the course of 16 months. The aim was to improve inpatient colonoscopy bowel preparation, as evaluated by the Boston Bowel Preparation Scale, procedure delays, and length of hospital stay. We additionally evaluated the groups' demographics and patient-level factors. A total of 459 inpatient colonoscopies were evaluated over a 16-month period. The intervention group consisted of 227 inpatient colonoscopies, while the control group consisted of 232. The intervention group showed superior Boston Bowel Preparation Scale score and decreased length of hospital stay. The number of adequate bowel preparations increased in the intervention group when compared to the control group from 77% to 86%. The creation of an EMR-based order set is a low-cost and sustainable action that can be easily implemented throughout a hospital system.

Background and study aims Diabetes mellitus (DM) is an independent risk factor for poor bowel preparation prior to colonoscopy. Bisacodyl is a stimulant laxative that may mitigate colonic dysmotility associated with diabetes. We hypothesized that adding bisacodyl to split-dose bowel preparation (SDBP) would improve the quality of bowel preparation among patients with diabetes. Patients and methods Adult outpatients aged 18 to 80 years undergoing colonoscopy were recruited. One hundred and eighty-six patients with diabetes were randomly assigned to 1 of 3 treatment arms: 1) conventional 4 L of polyethylene glycol electrolyte lavage solution (PEG-ELS; conventional bowel preparation [CBP]); 2) split-dose of 4 L PEG-ELS (split-dose bowel preparation [SDBP]); or 3) split-dose of 4 L PEG-ELS preceded by 10 mg of oral bisacodyl 10 (SDBP-B). The primary outcome measure was bowel cleansing as indicated by Boston Bowel Preparation Scale (BBPS) score. Endoscopists were blinded to the preparation used. Secondary outcome measures were safety and patient tolerability. Results Of the 212 patients randomized, only 186 received assigned bowel preparation. There were no differences among the three study groups with regard to age, indication, duration of DM, insulin use, narcotic use, or presence of end-organ diabetic complications. There was a trend toward better bowel preparation quality among those receiving SDBP and SDBP-B compared to those receiving CBP, but the trend was not statistically significant ≥ 6 BBPS; 67 % vs. 83 % vs. 75 %, P = 0.1). In terms of safety and tolerability, there were no differences among the three groups. Conclusion Adding bisacodyl to SDBP does not improve the quality of bowel preparation in patients with DM. Further efforts are needed to optimize colonoscopy bowel preparation in this population.

IntroductionColonoscopy is the gold standard for investigation of the large bowel. Adequate bowel preparation is vital to an effective procedure. A well-informed, motivated patient, who understands the process to prepare...

Background and Aim: Post-marketing studies comparing low-volume polyethylene glycol (PEG)-based regimens are limited. The aim of this randomized controlled non-inferiority study was to compare the efficacy and tolerability of two commercially available low-volume PEG-based preparations: PLENVU (Norgine) macrogol 3350, sodium ascorbate, sodium sulfate, ascorbic acid, sodium chloride and potassium chloride and Prepkit-C (Fresenius Kabi) macrogol 3350, sodium chloride, potassium chloride and sodium sulfate, sodium picosulfate, magnesium oxide, citric acid and aspartame in an unselected adult colonoscopy outpatient population. Methods: Patients undergoing outpatient colonoscopy were randomized to receive either split-dose Prepkit-C or PLENVU. Patient compliance, satisfaction, and tolerability were captured using a standardized questionnaire on the day of the procedure. A blinded endoscopist assessed the quality of bowel preparation using the Boston Bowel Preparation Scale (BBPS) and Aronchick Scale during the colonoscopy. The procedure times, rates of cecal intubation, and polyp detection rates were also recorded. Results: The study was terminated at 19 months, as recruitment was halted due to the coronavirus disease 2019 pandemic, and an unplanned interim analysis was performed. A total of 164 patients were randomly assigned to either Prepkit-C (n = 82) or PLENVU (n = 82). In the PLENVU group, 89% of patients achieved satisfactory preparation (total BPPS score ≥6 and each subscore ≥2), compared with 85.4% in the Prepkit-C group, resulting in an estimated difference of 3.7% (95% CI, -6.6% to 13.9%;P = 0.48). The median BBPS score with PLENVU was significantly better in the right, middle, and left colonic subsegments with PLENVU (BBPS score, 3;IQR, 2-3) than with Prepkit-C (BBPS score, 2;IQR, 2-3) (P < 0.001). A significantly higher proportion of patients in the PLENVU group had a BBPS score of ≥7, compared with patients in the Prepkit-C group (80% vs 57%;P = 0.002). A significantly greater proportion of patients had good-excellent quality of bowel preparation in the PLENVU group compared with the Prepkit-C group (84% vs 57%;P < 0.001) (Fig. 1). More patients in the PLENVU group reported moderate to severe nausea compared with those in the Prepkit-C group (P = 0.028);however, there were no other differences in patient satisfaction or tolerability between the two groups. Conclusion: The quality of bowel preparation achieved with PLENVU is at least non-inferior to that with Prepkit-C in a real-world unselected group of colonoscopy outpatients, with equivalent tolerability outcomes. This is despite a lower volume of preparation (by 500 mL) that is required to be ingested with PLENVU. Further studies are required in patients at risk of suboptimal bowel preparation.