Introduction

Artificial intelligence (AI) characterization (CADx) systems for colonic polyps exhibit high diagnostic accuracy. However, they may only support, not replace, clinical decision-making and little is known of this human-AI interaction – even less so with trainee endoscopists. This issue is difficult to explore, but data from a study investigating CADx-supported trainee endoscopists compared to AI-blinded experts could give insight.

Aims & Methods

This analysis draws on data from a prospective observational study comparing the diagnostic performance of CADx-supported trainee endoscopists (using GI Genius®, Medtronic) to expert endoscopists blinded to AI support. Two key scenarios were explored: (a) instances in which the CADx system produced a ‘no prediction’ output, and (b) cases where trainees actively overruled AI-generated diagnoses. These were assessed to examine the impact of AI absence and human-AI discordance on diagnostic accuracy.

Results

The baseline cohort included 11 trainees performing 225 colonoscopies, mean age 63.8 years (SD 12.7), 48.9% male patients. Baseline ADR was 57.8%. Of 630 resected lesions (median size 3 mm, IQR 2;4), 40.0% were in the rectosigmoid, and 53.6% were non-adenomatous. CADx failed to provide a prediction in 19.2% (n=121) of cases. These ‘no prediction’ polyps were smaller (median 3 mm, p=0.048) and more often located in the right colon (72% vs. 58%, p=0.005. In the absence of AI support, trainee diagnostic accuracy declined significantly by 12.8% (63.6% vs. 76.4%, p=0.007). Interestingly, expert endoscopists—despite being blinded to AI—also performed worse on these polyps (64.7% vs. 73.9%, p=0.045), suggesting an inherent difficulty in classifying these lesions.
Trainees overruled AI suggestion leading to discordant diagnoses in 9.0% (57/630) of polyps, comprised of 20 (35.1%) non-adenomatous, and 37 (64.9%) adenomatous lesions. Trainees incorrectly overruled AI in 59.6% (34/57), which mostly consisted of downgrading lesions to NICE 1. Discordant polyps did not differ in size (3 mm vs. 3 mm, p=0.413), but were more often located proximally (77.2% vs. 58.1%, p=0.005). The difference in false negative rate (FNR) between trainees and AI was 62.16% (81.07% vs. 18.92%, p<0.0001), demonstrating a risk of missed adenomas when trainees override AI. Blinded expert endoscopists achieved identical overall diagnostic accuracy in those polyps as the AI system (59.6%). However, their error profiles differed – In this subset, blinded experts and AI performed equally in overall accuracy (59.6%), although with differing error profiles: AI showed a numerically lower FNR (18.9% vs. 35.1%, Δ16.2%, p=0.17), offset by a higher false positive rate (80.0% vs. 50.0%).

Conclusion

CADx systems generate ‘no prediction’ outputs in a substantial proportion of polyps, significantly impairing trainee diagnostic performance. These cases may represent inherently more challenging lesions. Despite CADx’s high accuracy, trainees overrule AI in almost every tenth polyp, most often incorrectly downstaging polyps. Structured training on appropriate use of CADx, emphasizing both its capabilities and limitations, may enhance clinical effectiveness. Further research on human-AI interaction and trust is warranted, as these factors will be critical for successful integration of CADx into routine endoscopic practice.

Disclosure

The original study (AC-CADx) received technical support by Medtronic. Medtronic was not involved in the design of the study or analysis of the data, neither in the original study nor this exploratory analysis.

Introduction

Alcohol-induced acute pancreatitis (AP) is a common form of AP, while alcoholic hepatitis (AH) is a severe complication of alcoholic liver disease. Maléth et al. demonstrated that ethanol and fatty acids reduce cystic fibrosis transmembrane conductance regulator (CFTR) expression and activity, worsening alcohol-induced AP. Previously, we showed that ethanol and fatty acids disturb the CFTR-PMCA interaction, leading to disruped intracellular Ca²⁺ balance, exacerbating pancreatitis and potentially hepatitis. Previously Berrocal et al. described that MB reverses the inhibitory effect of Tau on PMCA, suggesting that its beneficial effect is based on the preservation of endogenous Ca²⁺-ATPase activity. Methylene blue (MB) is known for its antioxidant and mitochondrial protective properties, which can also influence intracellular Ca²⁺ regulation. Here we hypothesize that potentiating PMCA with methylene blue could have a protective effect in alcohol-induced AP and AH.

Aims & Methods

This study investigates MB’s effect on alcohol-induced AP severity and its underlying protective mechanisms in alcohol-related diseases. The effects of MB were assessed in mouse pancreatic (MPO), mouse liver (MLO), and human pancreatic ductal organoid (HPO) cultures. Microfluorimetric measurements were performed using an inverted fluorescence microscope. The polarized expression of ion channels was analyzed via immunofluorescence (IF). Gene expression changes were examined using qPCR, while mitochondrial membrane potential was assessed with TMRM and JC-10 staining. Cytotoxicity was evaluated using the CellTiter-Glo® 3D assay. Alcohol-induced AP and AH were modeled in wild-type FVB/N mice. AP was induced by repeated ethanol (EtOH) and palmitoleic acid (POA) injections, while AH was provoked using a 5% ethanol-fed Lieber-DeCarli diet.

Results

Fluorescent Ca²⁺ measurements showed that MB significantly potentiated PMCA activity following EtOH and PA exposure in MPOs, MLOs, and HPOs compared to EtOH-PA-treated groups. Mitochondrial membrane potential measurements confirmed MB’s protective effect. Gene expression analysis revealed no significant differences in CFTR, PMCA4, or ORAI1 expression between EtOH-PA and MB-treated groups. IF staining of PMCA4 demonstrated a defined apical localization in EtOH-PA + MB-treated groups. In experimental AH, MB’s effects were evaluated using Oil Red O and Trichrome staining, along with serum ALAT, ASAT, and triglyceride (TG) levels. In alcohol-induced AP, MB treatment significantly reduced serum amylase levels, oedema, leukocyte infiltration, and necrosis scores compared to the EtOH-POA group.

Conclusion

MB treatment mitigates the harmful effects of EtOH and PA exposure by enhancing PMCA activity, preserving mitochondrial function, and promoting PMCA4’s apical localization. Although no significant changes were observed in CFTR, PMCA4, or ORAI1 gene expression, histological and biochemical analyses highlight MB’s potential in alleviating alcohol-induced AP, as evidenced by reduced oedema, leukocyte infiltration, necrosis scores, and serum amylase levels. These findings suggest MB as a promising therapeutic candidate for protecting cellular and mitochondrial integrity in alcohol-related diseases.

Disclosure

I declare no conflict of interest

Introduction

Autoimmune pancreatitis (AIP) is a rare, immune-mediated form of chronic pancreatitis, classified into three subtypes. Type 1 AIP, the pancreatic manifestation of IgG4-related disease, often presents with multisystem involvement and frequent relapse (1,2). Type 2 AIP is pancreas-limited, tends to affect younger individuals, and is commonly associated with inflammatory bowel disease (IBD) (2,3). A third form—checkpoint inhibitor-related pancreatitis (CIP or Type 3 AIP)—has been recently described in patients receiving immune checkpoint inhibitors, typically asymptomatic but may result in pancreatic atrophy (4).
Precise subtyping is critical due to differences in clinical behavior, relapse risk, and treatment response (1,2,3). We present a single-center series comparing clinical, radiologic, and histopathologic characteristics of Type 1 and Type 2 AIP.

Aims & Methods

We retrospectively reviewed 20 patients diagnosed with AIP between 2020–2024 at a tertiary center. Classification followed International Consensus Diagnostic Criteria (1). Diagnostic evaluation included endoscopic ultrasound (EUS), serum IgG4 levels, and contrast-enhanced CT/MRI/MRCP (1,2). EUS-guided fine-needle biopsy (FNB) was performed where malignancy could not be excluded. Histopathology differentiated lymphoplasmacytic sclerosing pancreatitis (LPSP, Type 1) from idiopathic duct-centric pancreatitis (IDCP, Type 2) (2). Data on extra-pancreatic involvement, steroid response, and relapse were collected.

Results

Fifteen patients were diagnosed with Type 1 AIP and five with Type 2. Type 1 patients had a mean age of 63 years and were predominantly male (80%). Obstructive jaundice was the most common presentation. All had elevated serum IgG4. EUS revealed diffusely hyperechoic parenchyma and long-segment ductal narrowing. Histology confirmed LPSP. Extra-pancreatic involvement was observed in 60%, most commonly IgG4 cholangitis. All responded to corticosteroids; 40% relapsed within one year and received rituximab (5).
Type 2 patients had a mean age of 38, with equal gender distribution. Presenting symptoms included abdominal pain and acute pancreatitis. Serum IgG4 was normal in all cases. EUS revealed focal hypoechoic lesions with short-segment narrowing. Histology confirmed IDCP. Two patients (40%) had ulcerative colitis. All achieved remission with steroids, and no relapses were observed (2,3).

Conclusion

This series emphasizes the distinct clinical courses of AIP subtypes. Type 1 AIP represents a systemic IgG4-related disease with high relapse risk, often requiring long-term immunosuppression (5). Type 2 AIP affects younger patients, remains confined to the pancreas, and shows sustained remission after steroid therapy (2,3). Accurate classification using EUS, MRCP, IgG4, and histology is crucial for appropriate management and long-term outcomes (1,2).

References

1. Hart PA, Zen Y, Chari ST. Recent Advances in Autoimmune Pancreatitis. Gastroenterology. 2015;149(1):39–51.

2. de Pretis N, Crinò SF, Amodio A, et al. Type 2 autoimmune pancreatitis: Clinical features and long-term follow-up in an Italian cohort. United European Gastroenterol J. 2024;12(1):50–59.
   

3. Lohr JM, Dominguez-Munoz E, Rosendahl J, et al. United European Gastroenterology evidence-based guidelines for the diagnosis and therapy of chronic pancreatitis (HaPanEU). United European Gastroenterol J.2017;5(2):153–199.

4. Lamarca A, Dhani N, Hubner RA, et al. Immune checkpoint inhibitor-induced autoimmune pancreatitis (Type 3 AIP): Emerging entity with distinct features. Ann Oncol. 2023;34(3):385–393.
   

5. Ghazale A, Chari ST, Zhang L, et al. Treatment of relapsing autoimmune pancreatitis with immunomodulators and rituximab: The Mayo Clinic experience. Clin Gastroenterol Hepatol. 2012;10(6):640–646.