Validation of the computed assessment of cleansing score with the Mirocam® system

From Data to Insights: How Is AI Revolutionizing Small-Bowel Endoscopy?

The role of capsule endoscopy and enteroscopy in managing various small-bowel pathologies is well-established. However, their broader application has been hampered mainly by their lengthy reading times. As a result, there is a growing interest in employing artificial intelligence (AI) in these diagnostic and therapeutic procedures, driven by the prospect of overcoming some major limitations and enhancing healthcare efficiency, while maintaining high accuracy levels. In the past two decades, the applicability of AI to gastroenterology has been increasing, mainly because of the strong imaging component. Nowadays, there are a multitude of studies using AI, specifically using convolutional neural networks, that prove the potential applications of AI to these endoscopic techniques, achieving remarkable results. These findings suggest that there is ample opportunity for AI to expand its presence in the management of gastroenterology diseases and, in the future, catalyze a game-changing transformation in clinical activities. This review provides an overview of the current state-of-the-art of AI in the scope of small-bowel study, with a particular focus on capsule endoscopy and enteroscopy.

Lavage, Simethicone, and Prokinetics-What to Swallow with a Video Capsule

The development of new capsules now allows endoscopic diagnosis in all segments of the gastrointestinal tract and comes with new needs for differentiated preparation regimens. Although the literature is steadily increasing, the results of the conducted trials on preparation are sometimes conflicting. The ingestion of simethicone before gastric and small bowel capsule endoscopy for prevention of air bubbles is established. The value of a lavage before small bowel capsule endoscopy (SBCE) is recommended, although not supported by all studies. Ingestion in the morning before the procedure seems useful for the improvement of mucosa visualization. Lavage after swallowing of the capsule seems to improve image quality, and in some studies also diagnostic yield. Prokinetics has been used with first generation capsules to shorten gastric transit time and increase the rate of complete small bowel visualization. With the massively prolonged battery capacity of the new generation small bowel capsules, prokinetics are only necessary in significantly delayed gastric emptying as documented by a real-time viewer. Lavage is crucial for an effective colon capsule or pan-intestinal capsule endoscopy. Mainly high or low volume polyethylene glycol (PEG) is used. Apart from achieving optimal cleanliness, propulsion of the capsule by ingested boosts is required to obtain a complete passage through the colon within the battery lifetime. Boosts with low volume sodium picosulfate (NaP) or diatrizoate (gastrografin) seem most effective, but potentially have more side effects than PEG. Future research is needed for more patient friendly but effective preparations, especially for colon capsule and pan-intestinal capsule endoscopy.

Development and Verification of a Deep Learning Algorithm to Evaluate Small-Bowel Preparation Quality

Capsule endoscopy (CE) quality control requires an objective scoring system to evaluate the preparation of the small bowel (SB). We propose a deep learning algorithm to calculate SB cleansing scores and verify the algorithm’s performance. A 5-point scoring system based on clarity of mucosal visualization was used to develop the deep learning algorithm (400,000 frames; 280,000 for training and 120,000 for testing). External validation was performed using additional CE cases (n = 50), and average cleansing scores (1.0 to 5.0) calculated using the algorithm were compared to clinical grades (A to C) assigned by clinicians. Test results obtained using 120,000 frames exhibited 93% accuracy. The separate CE case exhibited substantial agreement between the deep learning algorithm scores and clinicians’ assessments (Cohen’s kappa: 0.672). In the external validation, the cleansing score decreased with worsening clinical grade (scores of 3.9, 3.2, and 2.5 for grades A, B, and C, respectively, p < 0.001). Receiver operating characteristic curve analysis revealed that a cleansing score cut-off of 2.95 indicated clinically adequate preparation. This algorithm provides an objective and automated cleansing score for evaluating SB preparation for CE. The results of this study will serve as clinical evidence supporting the practical use of deep learning algorithms for evaluating SB preparation quality.

Scoring systems in clinical small-bowel capsule endoscopy: all you need to know!

Capsule endoscopy (CE) emerged out of the pressing clinical need to image the small bowel (SB) in cases of midgut bleeding and provide an overall comfortable and reliable gastrointestinal (GI) diagnosis 1. Since its wider adoption in clinical practice, significant progress has been made in several areas including software development, hardware features and clinical indications, while innovative applications of CE never cease to appear 2 3. Currently, several manufacturers provide endoscopic capsules with more or less similar technological features 4. Although there is engaging and continuous academic and industry-fueled R&D, promising furtherment of CE technology 4 5, the current status of clinical CE remains that of by and large an imaging modality. Clinical relevance of CE images is cornerstone in the decision-making process for medical management. In one of the larger to date SB CE studies, 4,206 abnormal images were detected in 3,280 patients 6. Thus, CE leads to the identification of a large amount of potential pathology, some of which are pertinent (or relevant) while some (probably the majority) are not.

Soon artificial intelligence (AI) is likely to carry out several roles currently performed by humans; in fact, we are witnessing only the first stages of a transition in the clinical adoption of AI-based solutions in several aspects of gastroenterology including CE 7. Until then though, human-based decision-making profoundly impacts patient care and – although not suggested in the updated European Society of Gastrointestinal Endoscopy (ESGE) European curriculum 8 9 – it should be an integral part of CE training. Frequently, interpretation of CE images by experts or at least experienced readers differs. In a tandem CE reading study, expert review of discordant cases revealed a 50 % (13/25 discordant results) error rate by experienced readers, corresponding (in 5/13 cases) to ‘over-classification’ of an irrelevant abnormality 10. Another comparative study showed an ‘over-classification’ of such irrelevant abnormalities in ~10 % of CE readings 11. One thing which has been for a while on the table – in relation to optimizing and/or standardizing CE reporting and subsequent decision-making – is the need for reproducible scoring systems and for a reliable common language among clinicians responsible for further patient’s management.

Over the years, several of these scoring systems were developed while others appear in the wake of software and hardware improvements aiming to replace and/or complement their predecessors. This review presents a comprehensive account of the currently available classification/scoring systems in clinical CE spanning from predicting the bleeding potential of identified SB lesions (with emphasis on vascular lesions), and the individual rebleeding risk; scoring systems for the prediction of SB lesions in patients with obscure gastrointestinal bleeding (OGlB), having the potential to improve patient selection and rationalize the use of enteroscopy, with better allocation of resources, optimized diagnostic workflow and tailored treatment. This review also includes scores for reporting the inflammatory burden, the cleansing level that underscores confidence in CE reporting and the mass or bulge question in CE. Essentially, the aim is to become a main text for reference when scoring is required and facilitate the inclusion of -through readiness of access- one of the other in the final report.

Development of a deep learning-based software for calculating cleansing score in small bowel capsule endoscopy

A standardized small bowel (SB) cleansing scale is currently not available. The aim of this study was to develop an automated calculation software for SB cleansing score using deep learning. Consecutively performed capsule endoscopy cases were enrolled from three hospitals. A 5-step scoring system based on mucosal visibility was trained for deep learning in the training set. Performance of the trained software was evaluated in the validation set. Average cleansing score (1.0 to 5.0) by deep learning was compared to clinical grading (A to C) reviewed by clinicians. Cleansing scores decreased as clinical grading worsened (scores of 4.1, 3.5, and 2.9 for grades A, B, and C, respectively, P < 0.001). Adequate preparation was achieved for 91.7% of validation cases. The average cleansing score was significantly different between adequate and inadequate group (4.0 vs. 2.9, P < 0.001). ROC curve analysis revealed that a cut-off value of cleansing score at 3.25 had an AUC of 0.977. Diagnostic yields for small, hard-to-find lesions were associated with high cleansing scores (4.3 vs. 3.8, P < 0.001). We developed a novel scoring software which calculates objective, automated cleansing scores for SB preparation. The cut-off value we suggested provides a standard criterion for adequate bowel preparation as a quality indicator.

Poor Quality of Small Bowel Capsule Endoscopy Images Has a Significant Negative Effect in the Diagnosis of Small Bowel Malignancy

Background and Aims

Capsule endoscopy (CE) is a visual modality; hence, diagnosis relies on image quality. We studied the contribution of image parameters to visualization quality and their effect on diagnostic certainty of small bowel (SB) lesions.

Methods

Five clear CE images of common SB pathology – two vascular lesions, two inflammatory, one polyp – were processed for three image parameters to simulate poor SB conditions: opacity (color-matched to luminal content; 10–90%, 10% increments); blurriness (radius 1–10 pixels; one pixel increments); and contrast (−50-50%; 10% increments). Nine expert readers evaluated whether images were adequate for diagnosis. Points where perception of image quality changed significantly were determined for each parameter. Three further sets of SBCE images (vascular, inflammatory, and neoplastic lesions; nine images/set) were processed for four points/parameters. Twenty experienced/expert CE readers reviewed these images.

Results

The negative effects of opacity in diagnostic certainty were mostly evident in images of neoplasia; images of vascular and inflammatory lesions were less affected. Similar results were observed with increasing blur radius, simulating movement, and poor focus. The proportions of readers finding vascular and inflammatory images adequate for diagnosis did not drop significantly at wider blur radii, while images of neoplasia were quickly deemed inadequate. Low contrast had a greater negative effect than high, most consistently in neoplastic lesions.

Conclusion

Poor visualization quality in all parameters affected mostly neoplastic lesions. Software to increase contrast and sharpen images can improve visualization quality; smart frame rate adaptation could improve the number of high-quality frames obtained. Thoroughness in SB cleansing is most important when there is a suspicion of neoplasia.

Evaluation of the sensitivity of the Express View function in the Mirocam® capsule endoscopy software

Background: A new computer algorithm called Express-View has recently been introduced by Mirocam, but data concerning its application and efficacy are scarce.Objective: To evaluate the lesion detection rate, per-patient sensitivity and the diagnostic accuracy using Express-View.Methods: All patients who performed CE between January 2018 and June 2019, whose indication was obscure gastrointestinal bleeding (OGIB) and with findings on CE, were included. Lesions identified in conventional reading were selected and considered as reference.Results: Eighty-nine patients were included, 50.6% male, with a mean age of 68.4 years-old (±12.3). The Express-View mode detected 85.5% of lesions previously detected by conventional reading (524 out of 613). There were 89 missed lesions, mainly erosions or ulcers (44.9%) and angioectasias (38.2%). The lesion detection rate was found to be lower in the jejunum and ileum compared to extra-small bowel locations and duodenum (p = .04). Although Express-View had a per-patient sensitivity for all lesions of 56.2% and a per-patient sensitivity for all clinically significant lesions of 83.1%, it achieved a diagnostic accuracy of 91%.Conclusions: The per-patient sensitivity for all lesions was shown to be below expectations, although the lesion detection rate, the per-patient sensitivity for all clinically significant lesions, and the diagnostic accuracy were shown to be higher.

Evaluation of the Usefulness of Virtual Chromoendoscopy with Different Color Modes in the MiroCam® System for Characterization of Small Bowel Lesions

Background

Virtual chromoendoscopy (VC) in small bowel capsule endoscopy can improve the visualization and characterization of different small bowel lesions (SBLs). There are few studies of its usefulness in the Given® system, and there is no evidence yet of its utility in the MiroCam® system.

Aim

The aim of this study was to evaluate whether VC can improve the characterization of SBLs with the MiroCam® system.

Methods

Twenty-two patients were selected, in which 100 elementary lesions were identified, including erosions (n = 45), ulcers (n = 17), and angioectasias (n = 38). For each lesion identified, images were captured without chromoendoscopy (normal image [NI]) and with chromoendoscopy modes 1 (color mode [CM] 1), 2 (CM2), and 3 (CM3). A score from 1 to 4 was assigned to each image, in which a better evaluation of the characteristics and limits of the lesion was classified in ascending order, where 1 is the worst and 4 the best evaluation. The scores of the various modes were compared with Kendall's tau-c coefficient.

Results

The average scores attributed to the photographs in NI, CM1, CM2, and CM3 were 3.83, 2.89, 1.85, and 1.43, respectively (tau-c = -0.75, p < 0.001). Evaluating the elementary lesions independently, the average scores for modes NI, CM1, CM2, and CM3 were 3.83, 2.92, 1.86, and 1.38 (tau-c = -0.77, p < 0.001) for erosions, respectively; 3.87, 2.96, 1.76, and 1.40 (tau-c = -0.80, p < 0.001) for ulcers, respectively; and 3.81, 2.82, 1.87, and 1.50 (tau-c = -0.71, p < 0.001) for angioectasias, respectively.

Conclusions

VC using the CMs available in the MiroCam® system has not proven useful for a better assessment of any of the SBLs.

Randomized controlled trial of 3 days fasting and oral senna, combined with mannitol and simethicone, before capsule endoscopy

BACKGROUND AND STUDY AIMS

The approach to small bowel preparation before capsule endoscopy (CE) is still suboptimal.

PATIENTS AND METHODS

One hundred eighty patients were randomly allocated to 3 groups. Patients in Group A took 250 mL 20% mannitol and 1 L 0.9% saline orally at 05:00 hours on the day of the procedure. In Group B the same preparation was taken at 20:00 on the day before, and at 05:00 on the day of CE; in addition, 20 mL oral simethicone was taken 30 minutes before CE. Group C was treated identically to Group B, except that the patients fasted for 3 days and took 3 g senna orally 3 times daily before CE. The length of bowel containing green luminal contents was assessed by ImageJ software and bowel cleanliness was evaluated by computed assessment of the cleansing score.

RESULTS

Cleansing of the whole small bowel and the distal small bowel were significantly different between the 3 groups (χ = 22.470, P = .000; χ = 17.029, P = .000, respectively). There were also significant differences between the 3 groups in the length of small bowel and specifically the length of the distal small bowel containing green luminal contents (χ = 12.390, P = .000, χ = 15.141, P = .000, respectively), but not with regard to the proximal small bowel (χ = 0.678, P = .509).

CONCLUSIONS

Three days fasting and oral senna, combined with 20% mannitol and simethicone, before CE, can reduce the effects of bile on the small bowel and improve small bowel cleansing, especially in the distal small intestine.