Introduction

The use of artificial intelligence (AI) systems for decision support has recently gained great popularity in the medicine.¹ However, recent studies have shown possible limitations emerging from the interaction between humans and AI. Among these, the term “automation bias (AB)” refers to human errors deriving from over-relying on AI’s false alarms such that the human changes a correct initial interpretation.² AB is poorly investigated in the medical field and was never tested in the gastroenterological setting.

Aims & Methods

The aim of the study was to investigate the impact of AI on human diagnostic performance in gastrointestinal endoscopy, including AB.
Ninety high-quality endoscopic videoclips (16 seconds) extracted from small bowel (SB) capsule endoscopy (N=60) and proctosigmoidoscopy (N=30) were edited with the addition of a mock AI system designed with diagnostic accuracy, sensitivity and specificity of 80%, 86% and 75%. The ground truth was settled on the agreement of 9 established experts about the presence of SB lesions with significant hemorrhagic potential according to the Saurin classification (30 videoclips), inflammatory lesions relevant for the diagnosis of SB Crohn's disease according to the Lewis score (30 videoclips) and assessing the Ulcerative Colitis (UC) Endoscopic Index of Severity (30 videos). The videoclips were randomly divided into three sessions and submitted around Europe to fully trained endoscopist, trainees and endoscopic technicians. Participants were instructed to the accuracy measures of AI annotations and scored each videoclip first without, then with AI support. Participants’ performances with and without AI were calculated. Subgroup analyses were made after stratification by participants’ experience level in each endoscopic setting and test performances.

Results

A total of 12.660 videoclips (both with and without AI) were visualized in 422 sessions (182, 118 and 122 participants completed the first, second and third session, respectively). A strong and significant difference between the respondents' performance (N=281) when they were unaided by AI (average accuracy (M)=0.78, [0.77, 0.79] and when they were supported by the AI (M=0.81, [0.80, 0.82], paired t-test, t(276)=11.2, p<0.001) was found. Size effect and the evidence in favour of a significant positive effect of AI were large, 0.67 (Cohen's D) and very large (Bayes Factor>100), respectively. Interestingly, no significant difference between trainees and specialists both in regard to their performance, either unaided, M=0.77 [0.74, 0.79] vs M=0.78, [0.77, 0.79] or aided M=0.80 [0.78, 0.82] vs M=0.81 [0.80, 0.82] and their accuracy improvement (M=0.03, [0.02, 0.05] vs M=0.03, [0.02, 0.04] (unaided t-value=-0.89, p=0.37; aided t-value=-0.69, p=0.49, difference t-value=0.54, p=0.59) were found. A strong leveller effect emerged, with low-performers improving their accuracy much more than top-performers (average improvement 5%, [0.04, 0.06] vs 1%, [0.00, 0.02], paired T test, t=6.53, p<0.001). Nine percent of the final decisions involved disregarding correct AI advice in favour of incorrect human judgment, while AB occurred in 2% of cases.

Conclusion

In this study, the application of AI to assess SB bleeding, SB inflammatory lesions, and UC activity brings significant improvement in diagnostic accuracy, with the benefit appearing to be greater in the low-performers. However, in rare cases, AI may induce diagnostic errors due to AB. Future studies directly evaluating the impact of AI in real-life clinical settings are needed to confirm these data.

References

1. Cabitza F, Campagner A, Angius R, Natali C, Reverberi C. 2023. AI Shall Have No Dominion: on How to Measure Technology Dominance in AI-supported Human decision-making. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems (CHI '23). Association for Computing Machinery, New York, NY, USA, Article 354, 1–20. https://doi.org/10.1145/3544548.3581095
2. Reverberi, C., Rigon, T., Solari, A. et al. Experimental evidence of effective human–AI collaboration in medical decision-making. Sci Rep 12, 14952 (2022). https://doi.org/10.1038/s41598-022-18751-2

Introduction

The incidence and severity of metabolic dysfunction-associated steatotic liver disease (MASLD) are notably higher in individuals diagnosed with type 2 diabetes mellitus (T2DM). However, studies evaluating the progression of MASLD in patients with T2DM and normal serum aminotransferase (ALT) levels are limited.

Aims & Methods

This 5-year prospective cohort study aimed to evaluate the progression of MASLD in T2DM patients with normal ALT. People with type 2 diabetes who had MASLD (defined by controlled attenuation parameter (CAP) >250 dB/min) and ALT levels that stayed normal (defined by ALT ≤40 IU/L for at least two occasions spanning at least six months) at Rajavithi Hospital, Bangkok, from November 2016 to 2017 were included. We excluded patients with viral hepatitis and alcohol consumption exceeding moderate levels. Transient elastography (TE) and CAP measurements were performed at baseline and ≥5 years of follow-up. The cutoffs to define significant and advanced fibrosis were TE >8 kPa and TE >12 kPa, respectively. All patients were instructed to reduce their weight and abstain from herbal remedies; medications for comorbidities were prescribed without limitations, subject to the physician's assessment. (most patients were under endocrinologist care.)

Results

The study included 84 patients, 57% of whom were female and had a mean age of 58.5 years. Of these, 62% were obese, with a mean BMI of 28.6 kg/m2. At baseline, the median TE was 6.92±3.35 kPa; 57 (68%) had no significant fibrosis, 22 (26%) had significant fibrosis, and 5 (6%) had advanced fibrosis, and after a 5-year follow-up, the median TE was 6.54±3.57 kPa; 69 (82%) had no significant fibrosis, 15 (18%) had significant fibrosis, and 2 (2%) had advanced fibrosis. During the follow-up, fibrosis decreased in 45 (54%), remained stable in 6 (7%), and increased in 33 (39%) patients, while a small proportion of patients experienced changes in the stages of fibrosis: 1.2% progressed from significant to advanced fibrosis; 17.4% regressed from advanced to significant fibrosis or transitioned from significant to no fibrosis; 12.8% remained stable with significant and advanced fibrosis; and the majority (66.3%) remained without significant fibrosis. A significant factor associated with fibrosis progression was alcohol consumption (p=0.013), while the use of angiotensin receptor blockers (p=0.042) was associated with fibrosis regression. Significant factors associated with steatosis progression included high baseline BMI (p=0.013), an increase in BMI ≥ 5% (p=0.04), high fasting plasma glucose, and LDL cholesterol levels, while a decrease in BMI ≥ 5% (p=0.002) the use of calcium channel blockers (p=0.038) and coffee consumption >3 cups per day (p=0.031) were associated with steatosis regression. ALT elevation during the follow-up occurred in a small proportion of patients, and it was not significantly associated with fibrosis or steatosis progression.

Conclusion

The majority of MASLD patients with T2DM and normal ALT maintained a relatively stable liver disease during 5 years of follow-up. Obesity, weight gain, alcohol consumption, and inadequate metabolic control were all associated with the progression of MASLD, whereas weight loss, coffee consumption, angiotensin receptor blockers, and calcium channel blockers were all associated with the improvement of MASLD.

Disclosure

I have nothing to disclose

Introduction

Capsule endoscopy (CE) is commonly used as the initial exam in situations of suspected mid-gastrointestinal bleeding, after normal upper and lower endoscopy. Although the assessment of the small bowel is the primary focus of CE, detection of upstream or downstream vascular lesions may also be clinically significant. This study aimed to develop and test a Convolutional Neural Network (CNN)-based model for panendoscopic automatic detection of vascular lesions during CE.

Aims & Methods

A multicentric retrospective study was conducted, based on 1022 CE exams. We used a total of 35655 frames, from seven types of CE devices (PillCam Colon 1, PillCam Crohn's, PillCam SB1, PillCam SB3, OMOM HD, Olympus, Mirocam), of which 11091 were considered to have vascular lesions (angiectasia or varices) after triple validation. We divided data into a training and a validation set, the latter being used to evaluate the model’s performance. At time of division, all frames from a given patient were assigned to the same dataset. Our primary outcome measures were sensitivity, specificity, accuracy, positive predictive value (PPV), negative predictive value (NPV) and an area under the precision-recall curve (AUC-PR).

Results

A total of 35655 frames were used, with 11091 including vascular lesions. The validation data set contained 2433 frames with normal mucosa and 933 vascular lesions.
Sensitivity and specificity were 86.4% and 98.3%, respectively. PPV was 95.2%, while the NPV was 95.0%. Overall accuracy was 95.0%. The AUC-PR value was 0.96. The CNN processed 115 frames per second.

Conclusion

The study presents promising results for the application of AI deep learning models in panendoscopic automatic detection of vascular lesions in various gastrointestinal topographies. The methodological details, including the patient-split design and the use of frames from different types of CE devices, suggest that the model could be effectively used in routine clinical practice independently of the device brand. This study highlights the potential of panendoscopic evaluation of all gastrointestinal tract and the need for further research in this area. Overall, the high diagnostic performance of this CNN in multibrand devices addresses an important issue of technological interoperability, allowing it to be replicated in multiple technological settings. However, the study has some limitations, including a small number of frames and a retrospective design, which may introduce selection bias, therefore prospective and multicentric studies are needed before introducing this deep learning model into clinical practice. This is the first proof-of concept AI deep learning model developed for panendoscopic automatic detection of vascular lesions during CE.

References

Triester, S.L., et al., A meta-analysis of the yield of capsule endoscopy compared to other diagnostic modalities in patients with obscure gastrointestinal bleeding. Am J Gastroenterol, 2005. 100(11): p. 2407-18.
Liao, Z., et al., Indications and detection, completion, and retention rates of small-bowel capsule endoscopy: a systematic review. Gastrointest Endosc, 2010. 71(2): p. 280-6.
Rondonotti, E., et al., Small-bowel capsule endoscopy and device-assisted enteroscopy for diagnosis and treatment of small-bowel disorders: European Society of Gastrointestinal Endoscopy (ESGE) Technical Review. Endoscopy, 2018. 50(4): p. 423-446.
Ribeiro, T.F.-U.I.C.H., Colon capsule endoscopy and artificial intelligence: a perfect match por panendoscopy, in Artificial Intelligence in Capsule Endoscopy, C.H. Mascarenhas M, Macedo G, Editor. 2023, Academic Press.
Tsuboi, A., et al., Artificial intelligence using a convolutional neural network for automatic detection of small-bowel angioectasia in capsule endoscopy images. Dig Endosc, 2020. 32(3): p. 382-390.
Saurin, J.C., et al., Diagnostic value of endoscopic capsule in patients with obscure digestive bleeding: blinded comparison with video push-enteroscopy. Endoscopy, 2003. 35(7): p. 576-84.
Mascarenhas, M., et al., Deep learning and colon capsule endoscopy: automatic detection of blood and colonic mucosal lesions using a convolutional neural network. Endosc Int Open, 2022. 10(2): p. E171-e177

Clinical Case Summary

Introduction: Superior Mesenteric Vein (SMV) Thrombosis accounts for only 5% of all cases of intestinal ischemia. As opposed to acute mesenteric ischemia secondary to an arterial thrombus which has a more drastic presentation, SMV thrombus typically presents with non-specific signs and symptoms. Rarely, when venous congestion leads to bowel wall edema which can consequently head to intestinal obstruction.
Case presentation: We present the case of a 47-year-old male presenting with severe generalized abdominal pain and vomiting. Abdominal CT findings were suggestive of small bowel obstruction and venous thrombosis. CT angiogram confirmed SMV thrombus which was visualized as severe narrowing of the SMV at the midabdomen a few centimeters above the umbilicus, with consequent congestion and decreased contrast enhancement proximal to the area of narrowing. The patient was initially managed conservatively with anticoagulation, however the patient did not improve clinically. Since there were no signs of infarction or peritonitis that would warrant immediate surgical intervention, an alternative approach using endovascular technique was attempted. The patient underwent SMV stenting which subsequently relieved the obstruction.

References

1. Treatment of superior mesenteric and portal vein thrombosis with direct thrombolytic infusion via an operatively placed mesenteric catheter; https://pubmed.ncbi.nlm.nih.gov/1527918
2. A systematic review of the management of acute superior mesenteric vein thrombosis in adults https://mednexus.org/doi/full/10.1097/JP9.0000000000000051
3. Naffouj S, Al-Shammari M, Salgia R. Treatment of colonic varices with a superior mesenteric venous stent: a case report describing a unique approach. Gastroenterol Rep (Oxf). 2021 Feb 4;9(6):597-600. doi: 10.1093/gastro/goab003. PMID: 34925858; PMCID: PMC8677559
4. Vekariya P, Daneti D, Senthamizhselvan K, et al. (June 14, 2021) Acute Intestinal Obstruction as an Initial Presentation of Mesenteric Venous Thrombosis. Cureus 13(6): e15652. doi:10.7759/cureus.15652

Disclosure

None

Introduction

Urgency is not an IBD Patient Reported Outcome (PRO) in the STRIDE II but it has been assessed at baseline and as an outcome measure in recent pivotal RCTs with newer biologicals and targeted synthetic small molecules in both Crohn’s disease CD) and ulcerative colitis (UC). However, although it is reasonable to assume that UC patients or CD patients with rectal involvement may complain for severe urgency, urgency has been assessed even in patients with Crohn’s ileitis (iCD).

Aims & Methods

The aim of this study was to assess whether urgency is reported by iCD patients with flares or quiescent disease in a Real-World (RW) setting. This prospective, pilot, RW study included CD patients with solely L1 phenotype. L1 location was documented by ileocolonoscopy, histology and, imaging (MRE, US, or capsule endoscopy). Patients with colonic and/or rectal disease, perianal involvement and complicated ileal disease were excluded. Patients were interviewed by office visits during disease flare [HBI>5, CRP <5mg/dl, faecal calprotectin (FC)>250μg/g faecal tissue and SES-CD for terminal ileum >3 if colonoscopy was performed within a month prior to the visit) and during clinical remission (HBI<5, CRP <5mf/dl and FC <200μg/g and ileal SES-CD=0, in a recent ileocolonoscopy). The presence of and severity of urgency was recorded by a Symptom Rating Score of 0-10 (absent to very severe). The presence and severity of abdominal pain, the number and consistency of stools, nocturnal stools, blood and/or mucus in the stools, and the feeling of inadequate evacuation were sought by a structed questionnaire.

Results

During a 15-month period, 85 iCD patients [45 males, median age (range) 33 (18-65) years, median (range) disease duration 9 (0.6-25) years] were enrolled in the study. Patients were receiving a wide range of treatments, including anti-TNF agents (n=35), ustekinumab (N=25), vedolizumab (N=5), azathioprine (n=10), budesonide (n=5) and mesalazine (n=15). Fifty patients had a flare, 65 were in remissionand 30 patients were interviewed sequentially in flare and remission of disease. All patients in disease flare reported abnormal classical PROs of CD. Moderate to severe urgency [median score 5, range 4-8] was reported by 5/50 (10%) patients with disease flare, 8/65 (12%) patients with disease remission and 2/30 (6.5%) patients who were interviewed during flare and remission of disease. 3/50 patients with urgency during a flare of CD did not report urgency during disease remission. Six of the 8 patients who were interviewed only with iCD in remission and reported urgency had concurrently typical Irritable Bowel Symptoms (IBS, abdominal pain relieved by defecation, alternating diarrhoea/constipation, inadequate bowel emptying).

Conclusion

This prospective, pilot, real-world study shows that the reliability of urgency as classical PRO in patients with iCD is questioned and raises doubts whether urgency should be included in the spectrum of PROs for patients with iCD both in RCTs and in RW studies. Validation of the terminology in different ethnic languages may also pose additional barriers in patient understanding and reporting the symptom in the presence of additional IBS-type symptoms. Large studies are needed to assess if urgency is present in patients with iCD and how reliably is reported especially in multinational RCTs.

Disclosure

Dr GJ Mantzaris: Research grants: AbbVie, Genesis, Merck Sharp & Dohme, Takeda; advisor/speaker: AbbVie, Aenorasis, Dr Falk, Ferring, Hospira, Janssen, Merck Sharp & Dohme, MYLAN, Pfizer, Takeda, Vianex.
Dr A Loukou: Nothing to declare