Introduction

Endoscopic ultrasound (EUS) fusion imaging with cross sectional imaging (CT or MR) has been proposed as a method for co-registration of structures visible through different imaging techniques, with localization and tracking of the position of biopsy tools or therapeutic devices. Although initially developed for transabdominal [1] or endorectal [2] applications, the method seams feasible also for EUS [3], as tested with an initial prototype system [4]. The aim of our paper was to report on the testing of a second generation EUS fusion imaging based on electromagnetic (EM) navigation.

Aims & Methods

The equipment used consisted of Hitachi Noblus US system, Pentax linear scope, Custom made model and Aurora NDI magnetic-navigation. In order to create a porcine model we used an artificial “esophagus” and a porcine stomach. Beyond that, we attached to the model “gallbladder filled with stones” and a “pseudocyst“ with the aim of simulating interventional procedures such as EUS-guided fine needle aspiration / biopsy (EUS-FNA/B) or EUS-guided pseudocyst / gallbladder drainage. Additionally, one EM 6 degrees of freedom (DOF) positioning sensor was placed on the “patient“ chest and scanned by CT. Furthermore, the 3D rendering of the model and registration with the pre-procedure data using the sensor have been performed.The last step was placing a catheter with an EM 6DOF sensor mounted inside, in the working channel of the endoscope close to the distal tip in order to fuse live EUS images with the pre-procedure CT.

Results

The system displayed in real-time both EUS image and corresponding CT cross-section. Small registrations errors due to the differences between the pre-operative data and real model anatomy at the procedure time occurred but they have been overcome. Thus, the rotation correction between EUS scan and corresponding CT cross section can be done by the doctor during endoscopic procedure operating a special designed catheter handle and adjusting the angle of the catheter tip inside the endoscope.

Conclusion

The purpose of incorporating image fusion into pancreatico-biliary pathology is to boost the quality of diagnostic and therapeutic strategies not only by improving and speeding up the learning curve of EUS, but also by enhancing the accuracy of preoperative assessment regarding tumor resectability or by increasing the safety of therapeutic EUS procedures.

References

1. Ewertsen C, Săftoiu A, Gruionu LG, Karstrup S, Nielsen MB. Real-time image fusion involving diagnostic ultrasound. AJR Am J Roentgenol 2013; 200(3): W249-55.
2. Ignee A, Dong Y, Schuessler G, Baum U, Dietrich CF. Endorectal fusion imaging: A description of a new technique. Endosc Ultrasound 2017; 6(4): 241-244.
3. Singh BS, Cazacu IM, Deza CA, Rigaud BS, Saftoiu A, Gruionu G, Guionu L, Brock KK, Koay EJ, Herman JM, Bhutani MS. Image Fusion Involving Real-Time Transabdominal or Endoscopic Ultrasound for Gastrointestinal Malignancies: Review of Current and Future Applications. Diagnostics (Basel). 2022 Dec 19;12(12):3218.
4. Gruionu LG, Saftoiu A, Gruionu G. A novel fusion imaging system for endoscopic ultrasound. Endosc Ultrasound 2016; 5(1): 35-42.

Disclosure

Funding: IDEAR project
Relationship with industry: None

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

Clinical Case Summary

A 45-year-old woman was founded with a raised serum alpha-fetoprotein (AFP) level of 89 ng/ml ( normal range < 8ug/L ) when taking a routine physical examination. During the following year, the serum AFP concentration level was gradually rising, up to 6353 ng/ml. There was no significant loss of weight or loss of appetite, and she denied having any feeling of bloating or abdominal pain.
Although the serum AFP concentration level was abnormally high, the cancer antigen (CA)125, cancer antigen (CA)199, carcinoembryonic antigen (CEA), PIVKA-II, AFP-L3, β-HCG levels were all within normal limits at each check. There was no abnormal findings in the hepatobiliary systems except the modest increased bilirubin (within 2 times). Hepatitis A, hepatitis C, hepatitis E antibody screen was negative. The result of HBV markers were: HBsAg(-), HBsAb(+), HBeAb(-), HBeAb(-), HBcAb(+), While HBV DNA quantification was less than 5.00E+2 IU/ml. Moreover, there was no significant abnormality concerning autoimmune liver disease. During the one-year follow up, several imaging examinations were performed,including abdominal ultrasonography, enhanced computer tomography (CT) , magnetic resonance imaging (MRI) and PET-CT. However, there was no evidence of liver lesion or tumour on all of them, even the PET scan.To check for hepatoid adenocarcinoma, gastroscopy, colonscopy, enhanced chest CT enhanced abdomen CT, and thyroid ultrasound were performed. Nevertheless, no suspicious lesion was found.
One year later, a 2.8×2.6 cm mass was detected in the left lobe of liver in a surveillance ultrasound scan. She underwent laparoscopic hepatocarcinoma resection, and the mass was confirmed moderately differentiated hepatocellular carcinoma by pathology. After surgery, AFP levels took 3 months to normalize.
From this case, we found that a gradual rise elevation in AFP were positive predictors for tumor development, especially hepatocellular carcinoma.

Clinical Case Summary

Caudate lobe hypertrophy is a common finding in cirrhotic liver, although giant variants are rare. Nevertheless, any lesion on a cirrhotic liver warrants consideration for hepatocellular carcinoma (HCC).
This case involves a 28-year-old male patient irregularly followed up at our center for secondary biliary cirrhosis. With a history of intrahepatic bile duct paucity due to Alagille syndrome since neonatal age, he is a current smoker (10 pack-years) with no alcohol abuse. After three years lost to follow-up, the patient presented with a gradual 4-month onset of jaundice, dark urine, and upper right quadrant discomfort. Laboratory findings revealed worsening liver function tests: total bilirubin 11.5 mg/dl (vs 2.5 mg/dl), predominantly conjugated 7.6 mg/dl (vs 1.7 mg/dl), alkaline phosphatases 3.3x ULN (vs 2.6x ULN), gamma-glutamyl transferase 6.4x ULN (vs 4.5x ULN), aspartate aminotransferase 3.8x ULN (vs 1.7x ULN), alanine aminotransferase 1.5x ULN (vs 1.3x ULN).
The 4-phase CT scan revealed a dysmorphic liver with a large mass spanning segments 4, 5, 6, 7, and 8, measuring 130x100mm. The mass exhibited arterial enhancement without late phase washout and no evidence of a capsule. This mass displaced the hepatic veins and abutted the inferior vena cava and portal branches, without signs of invasion. MRI showed a similar appearance with no diffusion restriction. Initially suspected as HCC, the mass was classified as LIRADS-4. Alpha-foetoprotein level was normal. CT-guided biopsy retrieved multiple 41mm samples. Histopathological findings indicated annular fibrosis with histological cholestasis and no malignant cells. Therefore, the established diagnosis was caudate lobe hypertrophy, following a multidisciplinary team discussion and review of radiological features.
Despite this entity's rarity, biopsy seems crucial to rule out HCC, especially with abnormal vascular behavior like arterial enhancement, to avoid overtreatment.

Clinical Case Summary

Renal transplant recipients are at an increased risk of opportunistic infections, and a subset of this population may develop de novo inflammatory bowel disease (IBD). Differentiating the various forms of post-transplant colitis is challenging. Chronic active EBV enteritis is rarely reported and easily misdiagnosed as IBD, which results in death in severe cases.
A 61-year-old man with a history of kidney transplant presented with bloody diarrhea, fever and weight loss. Ileocolonoscopy showed ulcers in the ileum and the right colon. He was empirically started on ganciclovir and mycophenolate mofetil (MMF) was suspended. Colonic biopsies showed unspecific findings. Cytomegalovirus infection, Clostridioides difficile infection, intestinal tuberculosis and lymphoproliferative disease were excluded. The hypothesis of de novo IBD was postulated. Two months later the patient presented with non-bloody diarrhea and fever. Steroid therapy was increased and everolimus was added. A repeat ileocolonoscopy showed pancolitis and ileitis with oval transversal ulcers. Biopsies were positive for EBV by in-situ hybridization, suggesting an EBV-induced enteritis. Serum viral load was 1397 IU/mL. Immunosuppression was reduced and intravenous immunoglobulin therapy was administered. The patient did not respond and developed hemorrhagic shock with indication for salvage surgery (subtotal colectomy). Less than one month later the patient died due to hemorrhagic shock.
Chronic active EBV infective enteritis should be considered for patients who exhibit IBD-like symptoms and showing irregular ulcers in both colon and small intestine. Clinicians should recognize the presence of EBV infection to avoid misdiagnosis. However, chronic immunosuppression and opportunistic infections, such as EBV, are potential triggers of de novo IBD. Management is challenging and still controversial. Janus kinase inhibitors represent a promising treatment strategy as they target the EBV-driven pathway.

References

- Choi C, Yu Q, Deb PQ, Wang W. Rare case of EBV-induced colitis in an immunocompetent individual. BMJ Open Gastroenterol. 2020;7(1):e000360;
- Liu R, Wang M, Zhang L, et al. The clinicopathologic features of chronic active Epstein-Barr virus infective enteritis. Mod Pathol. 2019;32(3):387-395;
- Gioco R, Puzzo L, Patanè M, et al. Post-transplant colitis after kidney transplantation: clinical, endoscopic and histological features. Aging (Albany NY). 2020;12(24):24709-24720;
- Zhang H, Zhao S, Cao Z. Impact of Epstein-Barr virus infection in patients with inflammatory bowel disease. Front Immunol. 2022;13:1001055;
- Lv K, Yin T, Yu M, Chen Z, Zhou Y, Li F. Treatment Advances in EBV Related Lymphoproliferative Diseases. Front Oncol. 2022;12:838817.

Introduction

Microlithiasis and sludge have been considered the largest aetiological drivers within the cohort of idiopathic acute pancreatitis for decades. Subsuming microlithiasis and sludge-induced pancreatitis patients into the idiopathic or gallstone-induced pancreatitis cohorts has so far made adequate risk assessment regarding pancreatitis outcome impossible. It is therefore still unclear whether different therapeutic approaches are needed for microlithiasis and sludge than for gallstone-induced pancreatitis.

Aims & Methods

348 patients with acute biliary pancreatitis treated at a high-volume pancreas centre from 2005 to 2021 were stratified into a total of 7 concrement groups using the new consensus definition for microlithiasis and sludge. The gallstone-pancreatitis cohort was compared with the microlithiasis and sludge cohort and all possible combination concrement groups with regard to pancreatitis disease progression (cholestasis pattern, EUS/ERCP results).

Results

Microlithiasis and sludge-induced pancreatitis as well as combination concrement groups classified according to the revised Atlanta classification did not show a milder course than gallstone-induced pancreatitis (p = 0.63). Microlithiasis (3.6 mg/dl; mean) and sludge (2.6 mg/dl; mean) showed an increase in bilirubin on the day of admission to hospital due to pancreatitis, which was not significantly different from gallstone-induced pancreatitis (3.5 mg/dl mean, normal range < 1.2 mg/dl; p = 0.067). The highest EUS concrement detection and ERCP concrement extraction rate occurred on the day of admission and day 1, respectively. EUS detection rate (day 1): Gallstone AP 44.82 % (n = 13/29), Microlithiasis AP 46 % (n= 7/15), Sludge AP 40.7 % (n = 11/27), p = 0.98; ERCP extraction rate (admission day): Gallstone AP 39.1 % (n = 18/46), Microlithiasis AP 25 % (n = 5/20),Sludge AP 38.8 % (n = 7/18); p = 0.49).

Conclusion

Microlithiasis and sludge achieve a gallstone-equivalent cholestasis pattern and also do not differ in pancreatitis outcome from gallstone-induced acute biliary pancreatitis. In the case of EUS detection of microlithiasis and sludge in the early phase of acute pancreatitis, the concrement should be considered as a trigger rather than a cause of pancreatitis (unless another aetiology can be detected).