Design and validation of an artificial intelligence system to detect the quality of colon cleansing before colonoscopy.

BACKGROUND AND AIMS: Patients' perception of their bowel cleansing quality may guide rescue cleansing strategies before colonoscopy. The main aim of this study was to train and validate a convolutional neural network (CNN) for classifying rectal effluent during bowel preparation intake as "adequate" or "inadequate" cleansing before colonoscopy. PATIENTS AND METHODS: Patients referred for outpatient colonoscopy were asked to provide images of their rectal effluent during the bowel preparation process. The images were categorized as adequate or inadequate cleansing based on a predefined 4-picture quality scale. A total of 1203 images were collected from 660 patients. The initial dataset (799 images), was split into a training set (80%) and a validation set (20%). The second dataset (404 images) was used to develop a second test of the CNN accuracy. Afterward, CNN prediction was prospectively compared with the Boston Bowel Preparation Scale (BBPS) in 200 additional patients who provided a picture of their last rectal effluent. RESULTS: On the initial dataset, a global accuracy of 97.49%, a sensitivity of 98.17% and a specificity of 96.66% were obtained using the CNN model. On the second dataset, an accuracy of 95%, a sensitivity of 99.60% and a specificity of 87.41% were obtained. The results from the CNN model were significantly associated with those from the BBPS (P<0.001), and 77.78% of the patients with poor bowel preparation were correctly classified. CONCLUSION: The designed CNN is capable of classifying "adequate cleansing" and "inadequate cleansing" images with high accuracy.

Increasing the low residue diet to 3 days does not improve the bowel cleansing in hard to prepare patients: Post hoc analysis of a randomized controlled trial.

BACKGROUND: Recent evidence suggests that the number of low residue diet (LRD) days does not influence the bowel cleansing quality in non-selected patients. However, there are not data in the subgroup of patients with risk factors of inadequate bowel cleansing. OBJECTIVE: The aim of this study was to assess whether a 3-day LRD improved the bowel cleansing quality in patients with risk factors of poor bowel cleansing. PATIENTS AND METHODS: Post hoc analysis of a randomized controlled trial carried out between December 2017 and March 2018 in a tertiary care hospital. Patients with high risk of poor bowel cleansing were selected following a validated score. The patients were randomized to the 1-day LRD or 3-day LRD groups. All patients received a 2-L split-dose of polyethylene glycol plus ascorbic acid. Intention-to-treat (ITT) and per-protocol (PP) analyses were conducted for the main outcome. RESULTS: 135 patients (1-day LRD group=67, 3-day LRD=68) were included. The rate of adequate cleansing quality was not significantly different between the groups in the ITT analysis: 76.1%, 95% CI: [64.6-84.8] vs. 79.4%, 95% CI: [68.2-87.4]; odds ratio (OR) 1.2, 95% CI [0.54-2.73]) or in the PP analysis: 77.3%, 95% CI: [65.7-85.8] vs. 80.3%, 95% CI: [69.0-88.3]; OR 1.2, 95% CI [0.52-2.77]). Compliance with the diet or cleansing solution, satisfaction or difficulties with the LRD and the polyp/adenoma detection rates were not significantly different. CONCLUSION: Our results suggest that 1-day LRD is not inferior to 3-day LRD in patients with risk factors of inadequate bowel cleansing.

Impact of a 1-day versus 3-day low-residue diet on bowel cleansing quality before colonoscopy: a randomized controlled trial.

BACKGROUND: The aim of this study was to assess whether a 3-day low-residue diet (LRD) improved bowel cleansing quality compared with a 1-day LRD regimen. METHODS: Consecutive patients scheduled for outpatient colonoscopy were randomized to the 1-day LRD or 3-day LRD groups. All patients received a 2-L split-dose of polyethylene glycol plus ascorbic acid. The primary outcome was bowel cleansing quality as evaluated using the Boston Bowel Preparation Scale (BBPS) (adequate cleansing ≥ 2 points per segment). Secondary outcomes were adherence to and level of satisfaction with the LRD, difficulty following the dietary recommendations, and willingness to repeat the same LRD in the future. Intention-to-treat (ITT) and per-protocol (PP) analyses were conducted for the primary outcome. A superiority analysis was performed to demonstrate that a 3-day LRD regimen was superior to a 1-day LRD regimen with a margin of 10 %. RESULTS: 390 patients (1-day LRD group = 196, 3-day LRD = 194) were included. The cleansing quality was not significantly different between the groups: ITT analysis 82.7 % (95 % confidence interval [CI] 77.4 to 88.0) vs. 85.6 % (95 %CI 80.7 to 90.5), with odds ratio (OR) 1.2 (95 %CI 0.72 to 2.15); PP analysis 85.0 % (95 %CI 79.9 to 90.1) vs. 88.6 % (95 %CI 84.0 to 93.2), with OR 1.4 (95 %CI 0.88 to 2.52). No differences were found regarding adherence to the diet or cleansing solution, satisfaction or difficulty with the LRD, and the polyp/adenoma detection rates. CONCLUSION: 3-day LRD did not offer advantages over 1-day LRD in preparation for colonoscopy.

Comparison of Two Intensive Bowel Cleansing Regimens in Patients With Previous Poor Bowel Preparation: A Randomized Controlled Study.

OBJECTIVES: Inadequate bowel cleansing is a major burden for endoscopy units. The aim of this study was to compare two intensive bowel cleansing regimens in patients with previous colonoscopy with inadequate bowel preparation. METHODS: Patients with inadequate cleansing at index colonoscopy were randomized to 4-L split-dose polyethylene-glycol (PEG) regimen vs. 2-L split-dose PEG plus ascorbic acid (PEG+Asc) regimen. All individuals underwent a 3-day low-residue diet and received 10 mg of bisacodyl, the day before colonoscopy. Cleansing was considered to be adequate if the Boston Bowel Preparation Scale scored ≥2 at each colonic segment. A non-inferiority analysis was performed to demonstrate that colonic cleansing with 2-L PEG+Asc was not inferior to 4-l PEG, considering a non-inferiority margin of 10%. RESULTS: Adequate bowel cleansing was significantly higher in patients assigned to 4-L PEG regimen (n=127) vs. those randomized to 2-L PEG+Asc regimen (n=129) by intention-to-treat analysis (81.1 vs. 67.4%, odds ratio (OR) 2.07, 95% confidence interval (CI) (1.163-3.689)) and by per-protocol analysis (86.6 vs. 71.7%, OR: 2.55, 95% CI: (1.316-4.922)). The study was terminated for futility after the interim analysis, because the 95% CI of the difference of proportions was 3.13-24.27% in the intention-to-treat analysis and 3.33-26.47% in the per-protocol analysis, confirming the superiority of 4-L PEG preparation. CONCLUSIONS: After 3-day low-residue diet and oral bisacodyl before colonoscopy, colon cleansing with 4-L split-dose PEG was superior to 2-L split-dose PEG+Asc in patients with previous inadequate cleansing. (EUDRACT: 2013-002506-31, NCT02073552).

Risk factors for inadequate bowel preparation: a validated predictive score.

Background and study aim Inadequate bowel cleansing negatively affects the efficiency of colonoscopy in routine clinical practice. The aim of this study was to design and validate a predictive model for inadequate bowel cleanliness. Patients and methods The model was built from 667 consecutive outpatients (development cohort) who were prospectively scheduled for colonoscopy between June and September 2014. The validation cohort included 409 outpatients who underwent colonoscopy between October and December 2014. Cleansing was evaluated using the Boston Bowel Preparation Scale (BBPS). Bowel preparation was administered on the same day as the examination. Results In the development cohort, BBPS was adequate in 541 patients (81.1 %). At multivariate analysis, antidepressants (odds ratio [OR] 4.25, 95 % confidence interval [CI] 1.91 - 9.47), co-morbidity (OR 3.35, 95 %CI 2.16 - 5.18), constipation (OR 2.09, 95 %CI 1.29 - 3.40), and abdominal/pelvic surgery (OR 1.60, 95 %CI 1.03 - 2.47) were independent predictors for inadequate cleansing. The model built with these variables showed an area under the curve of 0.72 in the development cohort and 0.70 in the validation cohort. A cutoff of 1.225 predicted inadequate bowel preparation with a sensitivity, specificity, positive predictive value, and negative predictive value of 60.3 % (95 %CI 51.6 - 68.4), 75.4 % (95 %CI 71.6 - 78.9), 36.4 % (95 %CI 30.1 - 43.1), and 89.1 % (95 %CI 85.9 - 91.6) in the development cohort, and 50.0 % (95 %CI 38.1 - 61.9), 80.0 % (95 %CI 75.3 - 84.2), 35.7 % (95 %CI 26.4 - 45.6), and 87.9 % (95 %CI 83.7 - 91.3) in the validation cohort. Conclusion A simple score may assist the clinician in predicting which patients are at high risk of inadequate bowel cleanliness. This may guide changes in bowel preparation strategy accordingly.

One-day versus two-day cleansing for colon capsule endoscopy: a prospective randomized pilot study.

BACKGROUND: Standard bowel cleansing for colon capsule endoscopy (CCE) requires a liquid diet and bowel laxatives for at least 2 days, which is a major drawback of this procedure and affects tolerance and acceptability. OBJECTIVE: To compare the quality of colon cleanliness achieved with one-day versus two-day bowel preparation in outpatients undergoing CCE. METHODS: Patients were randomly assigned to one of two groups: group I (one-day schedule, n=20) received a fiber-free diet and 3 L of polyethylene glycol (PEG) on day 0; group II (two-day schedule, n=20) received a liquid diet and 3 L of PEG in the evening of day -1, and 1L of PEG in the early morning of day 0. In both groups, the patients received 15 mg bisacodyl on day -1 and one or two additional sodium phosphate (NaP) boosters following capsule ingestion. Each colon segment was assessed for cleanliness using a four-point grading scale (excellent=1, good=2, fair=3, and poor=4). For the final analysis, colon cleanliness was rated as adequate (good or excellent) or inadequate (fair or poor). RESULTS: Overall colon cleanliness was adequate in 94% (CI 91-97) of patients in group I versus 80% (CI 72-88) in group II (P=0.27). No significant differences were observed in the per-segment quality of colon cleansing between the two groups. CCE reached the rectum in 80% (CI 73-87) of patients in group I versus 75% (CI 67-83) in group II (p=0.59). CONCLUSION: The quality of colon cleanliness achieved with one-day bowel preparation is equivalent to that of the standard two-day schedule in patients undergoing CCE.

An Enhanced High-Volume Preparation for Colonoscopy Is Not Better Than a Conventional Low-Volume One in Patients at Risk of Poor Bowel Cleansing: A Randomized Controlled Trial.

Objective: We tested the hypothesis that an enhanced bowel preparation strategy (EBS) improves colonic cleansing in patients at high risk for inadequate bowel cleansing (HRI). Methods: This prospective randomized clinical trial included consecutive HRI patients referred for outpatient colonoscopy between February and October 2019. HRI was considered if patients scored >1.225 according to a previously validated bowel-cleansing predictive score. HRI patients were randomized (1:1) to a low-volume conventional bowel cleansing strategy (CBS) (1-day low residue diet (LRD) plus 2 L of polyethylene glycol (PEG) plus ascorbic acid) or to an EBS (3-day LRD plus 10 mg oral bisacodyl plus 4 L PEG). The Boston Bowel Preparation Scale (BBPS) was used to assess the quality of cleanliness. Intention-to-treat (ITT) and per protocol (PP) analyses were performed. A sample size of 130 patients per group was estimated to reach a 15% difference in favor of EBP. Results: A total of 253 HRI patients were included (mean age 69.8 ± 9.5 years, 51.8% women). No statistically significant differences were found in the BBPS scale between the two groups in the ITT analysis (CBS 76.8% vs. EBS 79.7%, P = 0.58) or PP analysis (CBS 78% vs. EBS 84.3%, P = 0.21), risk difference 2.9% (95% CI-7.26 to 39.16) in the ITT analysis, or risk difference 6.3% (95% CI-3.48 to 16.08) in PP analysis. No differences in preparation tolerance, compliance, adverse effects, or colonoscopy findings were found. Conclusion: EBS is not superior to CBS in hard-to-prepare patients. (EUDRACT: 2017-000787-15, NCT03830489). Clinical Trial Registration: www.ClinicalTrials.gov, identifier NCT03830489.

Design and validation of an artificial intelligence system to detect the quality of colon cleansing before colonoscopy / Diseño y validación de un sistema de inteligencia artificial para detectar la calidad de la limpieza del colon previa a la realización de la colonoscopia

Background and aims Patients’ perception of their bowel cleansing quality may guide rescue cleansing strategies before colonoscopy. The main aim of this study was to train and validate a convolutional neural network (CNN) for classifying rectal effluent during bowel preparation intake as “adequate” or “inadequate” cleansing before colonoscopy.Patients and methodsPatients referred for outpatient colonoscopy were asked to provide images of their rectal effluent during the bowel preparation process. The images were categorized as adequate or inadequate cleansing based on a predefined 4-picture quality scale. A total of 1203 images were collected from 660 patients. The initial dataset (799 images), was split into a training set (80%) and a validation set (20%). The second dataset (404 images) was used to develop a second test of the CNN accuracy. Afterward, CNN prediction was prospectively compared with the Boston Bowel Preparation Scale (BBPS) in 200 additional patients who provided a picture of their last rectal effluent.ResultsOn the initial dataset, a global accuracy of 97.49%, a sensitivity of 98.17% and a specificity of 96.66% were obtained using the CNN model. On the second dataset, an accuracy of 95%, a sensitivity of 99.60% and a specificity of 87.41% were obtained. The results from the CNN model were significantly associated with those from the BBPS (P<0.001), and 77.78% of the patients with poor bowel preparation were correctly classified.ConclusionThe designed CNN is capable of classifying “adequate cleansing” and “inadequate cleansing” images with high accuracy. (AU)

Antecedentes y objetivos La percepción de los pacientes sobre la calidad de su limpieza intestinal puede guiar las estrategias de limpieza de rescate antes de una colonoscopia. El objetivo principal de este estudio fue entrenar y validar una red neuronal convolucional (CNN) para clasificar el efluente rectal durante la preparación intestinal como «adecuado» o «inadecuado».Pacientes y métodosPacientes no seleccionados proporcionaron imágenes del efluente rectal durante el proceso de preparación intestinal. Las imágenes fueron categorizadas como una limpieza adecuada o inadecuada según una escala de calidad de 4 imágenes predefinida. Se recopilaron un total de 1.203 imágenes de 660 pacientes. El conjunto de datos inicial (799 imágenes) se dividió en un conjunto de entrenamiento (80%) y un conjunto de validación (20%). Un segundo conjunto de datos (404 imágenes) se utilizó para evaluar la precisión de la CNN. Posteriormente, la predicción de la CNN se comparó prospectivamente con la escala de preparación colónica de Boston (BBPS) en 200 pacientes que proporcionaron una imagen de su último efluente rectal.ResultadosEn el conjunto de datos inicial, la precisión global fue del 97,49%, la sensibilidad del 98,17% y la especificidad del 96,66%. En el segundo conjunto de datos, se obtuvo una precisión del 95%, una sensibilidad del 99,60% y una especificidad del 87,41%. Los resultados del modelo de CNN se asociaron significativamente con la escala de preparación colónica de Boston (p<0,001), y el 77,78% de los pacientes con una preparación intestinal deficiente fueron clasificados correctamente.ConclusiónLa CNN diseñada es capaz de clasificar imágenes de «limpieza adecuada» y «limpieza inadecuada» con alta precisión. (AU)

One-day versus two-day cleansing for colon capsule endoscopy: a prospective randomized pilot study / Comparación de la limpieza intestinal en un día versus 2 días para la colonoscopia con videocápsula: estudio piloto prospectivo y aleatorizado

BACKGROUND: Standard bowel cleansing for colon capsule endoscopy (CCE) requires a liquid diet and bowel laxatives for at least 2 days, which is a major drawback of this procedure and affects tolerance and acceptability. Objective To compare the quality of colon cleanliness achieved with one-day versus two-day bowel preparation in outpatients undergoing CCE. Methods Patients were randomly assigned to one of two groups: group I (one-day schedule, n = 20) received a fiber-free diet and 3 L of polyethylene glycol (PEG) on day 0; group II (two-day schedule, n = 20) received a liquid diet and 3 L of PEG in the evening of day −1, and 1 L of PEG in the early morning of day 0. In both groups, the patients received 15 mg bisacodyl on day −1 and one or two additional sodium phosphate (NaP) boosters following capsule ingestion. Each colon segment was assessed for cleanliness using a four-point grading scale (excellent = 1, good = 2, fair = 3, and poor = 4). For the For the final analysis, colon cleanliness was rated as adequate (good or excellent) or inadequate (fair or poor). RESULTS: Overall colon cleanliness was adequate in 94% (CI 91-97) of patients in group I versus 80% (CI 72-88) in group II (P = 0.27). No significant differences were observed in the per-segment quality of colon cleansing between the two groups. CCE reached the rectum in 80% (CI 73-87) of patients in group I versus 75% (CI 67-83) in group II (p = 0.59). CONCLUSION: The quality of colon cleanliness achieved with one-day bowel preparation is equivalent to that of the standard two-day schedule in patients undergoing CCE

ANTECEDENTES: El inconveniente principal de la cápsula endoscópica de colon (CEC) es unapreparación intestinal que exige una dieta estricta y la toma de soluciones evacuantes durante al menos los 2 días previos a la prueba, limitando su aceptabilidad. OBJETIVO: Comparar la calidad de la preparación intestinal administrada en un día respecto a la de 2 días para la CEC. MÉTODOS: Se aleatorizan 2 grupos: Grupo I (preparación-1 día, n = 20) dieta sin fibra con 3 Lde polietilenglicol (PEG) el día 0; Grupo II (preparación-2 días, n = 20) dieta líquida y 3 L de PEG la noche previa (día 1) y 1 L de PEG el día 0. En ambos grupos se administra 15 mg debisacodilo el día 1 y fosfato sódico (NaP) como propulsor tras la ingesta de CEC. Cada segmento del colon fue valorado según la escala de limpieza de 4 grados (excelente = 1, buena = 2, regular = 3, y mala = 4). Para el análisis final, la limpieza global se cataloga como adecuada (1 y 2)o inadecuada (3 y 4). RESULTADOS: La limpieza global fue adecuada en un 94% (IC: 91-97) de pacientes del Grupo I frente al 80% (IC: 72-88) en el Grupo II (P = 0,27). No se determinaron diferencias cuando se compara la limpieza por segmentos colónicos. La CEC alcanza el recto en el 80% (IC: 73-87) del Grupo I frente al 75% (IC: 67-83) del Grupo II (p = 0,59).CONCLUSIÓN: La calidad de la limpieza colónica para la CEC obtenida con la preparación de un día único es similar a la preparación estándar de 2 días