Artificial intelligence for discrimination of Crohn's disease and gastrointestinal tuberculosis: A systematic review.

BACKGROUND AND AIM: Discrimination of gastrointestinal tuberculosis (GITB) and Crohn's disease (CD) is difficult. Use of artificial intelligence (AI)-based technologies may help in discriminating these two entities. METHODS: We conducted a systematic review on the use of AI for discrimination of GITB and CD. Electronic databases (PubMed and Embase) were searched on June 6, 2022, to identify relevant studies. We included any study reporting the use of clinical, endoscopic, and radiological information (textual or images) to discriminate GITB and CD using any AI technique. Quality of studies was assessed with MI-CLAIM checklist. RESULTS: Out of 27 identified results, a total of 9 studies were included. All studies used retrospective databases. There were five studies of only endoscopy-based AI, one of radiology-based AI, and three of multiparameter-based AI. The AI models performed fairly well with high accuracy ranging from 69.6-100%. Text-based convolutional neural network was used in three studies and Classification and regression tree analysis used in two studies. Interestingly, irrespective of the AI method used, the performance of discriminating GITB and CD did not match in discriminating from other diseases (in studies where a third disease was also considered). CONCLUSION: The use of AI in differentiating GITB and CD seem to have acceptable accuracy but there were no direct comparisons with traditional multiparameter models. The use of multiple parameter-based AI models have the potential for further exploration in search of an ideal tool and improve on the accuracy of traditional models.

Diagnostic and Therapeutic Challenges in the Management of Acute Massive Overt Bleeding of Jejunal Gastrointestinal Stromal Tumours: Case Series.

INTRODUCTION: Jejunal gastrointestinal stromal tumours (GIST) are rare mesenchymal tumours. Acute massive overt bleeding from jejunal GIST is very rare and poses both diagnostic and therapeutic challenges in emergent conditions. METHODS: A case series with retrospective analysis of prospectively maintained database of patients presenting with acute massive overt bleeding secondary to histologically proven jejunal GIST was done. Clinical characteristics, endoscopic and imaging diagnostic features, histological findings, surgical procedures and outcomes in these patients were studied. RESULTS: Three patients were included in this case series. Mean age of presentation was 49.0 years with two male and one female patient. All three patients presented with melena and hemodynamic instability, resuscitated with adequate blood transfusions. Routine endoscopic assessment were inconclusive. Multiphasic Computed Tomographic Angiography (CTA) revealed hypodense hypervascular mass in jejunum in all three patients. One patient was unresponsive to blood transfusion and underwent emergency exploratory laparotomy. One patient underwent laparoscopic resection and reconstruction. Mean length of hospital stay was 5.3 days. Histopathological examination confirmed jejunal GIST in all three patients with microscopically negative resection margins. Two patients were disease free till 18-month follow-up and the one patient lost to follow-up after 1 year. CONCLUSION: Multiphasic CTA is a single-step diagnostic tool for localisation of bleed and assessment of tumour characteristics in emergent conditions. Surgical resection is the mainstay of treatment for both control of bleed and to provide oncologically clear resection margins.

Perforations from Endoscopic Dilation of Corrosive Strictures in Adults: A Systematic Review and Meta-Analysis.

BACKGROUND AND AIMS: Endoscopic dilation is an important therapeutic option for treatment of corrosive strictures. Its safety is unclear with variable perforation rates in previous studies. We aimed to evaluate its safety with regard to perforations and the effect of dilator type to perforation risk. METHODS: A systematic review of published literature from inception to April 24, 2021, using PubMed and EMBASE databases was conducted. Studies in adult subjects (mean age ≥ 18 years) reporting perforation rates of endoscopic dilation of corrosive esophageal and/or gastric strictures were included. Pooled cumulative perforation rates were computed as primary outcome. Secondary outcomes included, perforation with each dilator subtypes, surgical or conservative modes of management and mortality. Random effect meta-analysis was used to estimate the frequency of each of these outcomes. Variables were reported as percentages with 95% CI. RESULTS: A total of 712 subjects (N) who underwent 4840 dilations (n) were noted in the 15 studies that were included. Of which, eight were retrospective, while the remaining seven were prospective. On meta-analysis, the cumulative pooled perforation rate was 1% (1-2%) of the number of dilations (n%). The perforation rates with SG (1%, 0-3%) and balloon (1%, 0-5%) dilators were similar (p value < 0.01). 45/64 (59%, 11-94%) perforations were subjected to surgery while the remaining 14/64 (41%, 6-89%) was managed conservatively. Choice of management strategy was biased to the norms of the treating team. About 3/712 patients (0%, 0-2%) succumbed following perforation. CONCLUSION: Perforation from endoscopic dilation of corrosive strictures occurs rarely, and there is no significant difference in perforation risk related to the type of dilator.

COMPARISON OF NON-ENDOSCOPIC SCORES FOR THE PREDICTION OF OUTCOMES IN PATIENTS OF UPPER GASTROINTESTINAL BLEED IN AN EMERGENCY OF A TERTIARY CARE REFERRAL HOSPITAL: A PROSPECTIVE COHORT STUDY.

BACKGROUND: Traditionally peptic ulcer disease was the most common cause of upper gastrointestinal (UGI) bleed but with the changing epidemiology; other etiologies of UGI bleed are emerging. Many scores have been described for predicting outcomes and the need for intervention in UGI bleed but prospective comparison among them is scarce. OBJECTIVE: This study was planned to determine the etiological pattern of UGI bleed and to compare Glasgow Blatchford score, Pre-Endoscopy Rockall score, AIMS65, and Modified Early Warning Score (MEWS) as predictors of outcome. METHODS: In this prospective cohort study 268 patients of UGI bleed were enrolled and followed up for 8 weeks. Glasgow Blatchford score, Endoscopy Rockall score, AIMS65, and MEWS were calculated for each patient, and the area under the receiver operating characteristic (AUC-ROC) curve for each score was compared. RESULTS: The most common etiology for UGI bleed were gastroesophageal varices 150 (63.55%) followed by peptic ulcer disease 29 (12.28%) and mucosal erosive disease 27 (11.44%). Total 38 (15.26%) patients had re-bleed and 71 (28.5%) patients died. Overall, 126 (47%) patients required blood component transfusion, 25 (9.3%) patients required mechanical ventilation and 2 (0.74%) patients required surgical intervention. Glasgow Blatchford score was the best in predicting the need for transfusion (cut off - 10, AUC-ROC= 0.678). Whereas AIMS65 with a score of ≥2 was best in predicting re-bleed (AUC-ROC=0.626) and mortality (AUC-ROC=0.725). CONCLUSION: Gastrointestinal bleed was most commonly of variceal origin at our tertiary referral center in Northern India. AIMS65 was the best & simplest score with a score of ≥2 for predicting re-bleed and mortality.

Comparação dos escores não endoscópicos para a previsão de resultados em doentes com sangramento gastrointestinal alto na emergência de um hospital de referência em cuidados terciários: um estudo de coorte prospectivo / Comparison of non-endoscopic scores for the prediction of outcomes in patients of upper gastrointestinal bleed in an emergency of a tertiary care referral hospital: a prospective cohort study

ABSTRACT BACKGROUND: Traditionally peptic ulcer disease was the most common cause of upper gastrointestinal (UGI) bleed but with the changing epidemiology; other etiologies of UGI bleed are emerging. Many scores have been described for predicting outcomes and the need for intervention in UGI bleed but prospective comparison among them is scarce. OBJECTIVE: This study was planned to determine the etiological pattern of UGI bleed and to compare Glasgow Blatchford score, Pre-Endoscopy Rockall score, AIMS65, and Modified Early Warning Score (MEWS) as predictors of outcome. METHODS: In this prospective cohort study 268 patients of UGI bleed were enrolled and followed up for 8 weeks. Glasgow Blatchford score, Endoscopy Rockall score, AIMS65, and MEWS were calculated for each patient, and the area under the receiver operating characteristic (AUC-ROC) curve for each score was compared. RESULTS: The most common etiology for UGI bleed were gastroesophageal varices 150 (63.55%) followed by peptic ulcer disease 29 (12.28%) and mucosal erosive disease 27 (11.44%). Total 38 (15.26%) patients had re-bleed and 71 (28.5%) patients died. Overall, 126 (47%) patients required blood component transfusion, 25 (9.3%) patients required mechanical ventilation and 2 (0.74%) patients required surgical intervention. Glasgow Blatchford score was the best in predicting the need for transfusion (cut off - 10, AUC-ROC= 0.678). Whereas AIMS65 with a score of ≥2 was best in predicting re-bleed (AUC-ROC=0.626) and mortality (AUC-ROC=0.725). CONCLUSION: Gastrointestinal bleed was most commonly of variceal origin at our tertiary referral center in Northern India. AIMS65 was the best & simplest score with a score of ≥2 for predicting re-bleed and mortality.

RESUMO CONTEXTO: Tradicionalmente, a doença úlcera péptica era a causa mais comum de sangramento digestivo alto, mas com a mudança da epidemiologia, outras etiologias do sangramento do trato digestivo alto estão emergindo. Muitas pontuações têm sido descritas para prever resultados e a necessidade de intervenção na hemorragia gastrointestinal superior, mas a comparação prospectiva entre elas é escassa. OBJETIVO: Este estudo foi planeado para determinar o padrão etiológico de pacientes com hemorragia digestiva alta e comparar os escores de Glasgow Blatchford, o Rockall pré-endoscopia, o AIMS65 e o Early Warning modificado (MEWS) como preditores do resultado. MÉTODOS: Neste estudo prospetivo de coorte, 268 pacientes com sangramento digestivo alto foram acompanhados durante 8 semanas. Os escores Glasgow Blatchford, Rockall pré-endoscopia, AIMS65 e MEWS foram calculados para cada paciente, e a área sob a curva (AUC-ROC) para cada pontuação foi comparada. RESULTADOS: A etiologia mais comum para a hemorragia gastrointestinal alta foi varizes gastroesofágicas 150 (63,55%), seguida de úlcera péptica 29 (12,28%) e de doença erosiva de mucosa 27 (11,44%). No total, 38 (15,26%) doentes voltaram a sangrar e 71 (28,5%) doentes morreram. No total, 126 (47%) doentes necessitaram de transfusão de componentes sanguíneos, 25 (9,3%) necessitaram de ventilação mecânica e 2 (0,74%) destes doentes necessitaram de intervenção cirúrgica. O escore de Glasgow Blatchford foi o melhor na previsão da necessidade de transfusão (corte - 10, AUC-ROC =0,678). Enquanto o AIMS65 com uma pontuação de ≥2 foi o melhor na previsão de ressangramento (AUC-ROC =0,626) e mortalidade (AUC-ROC =0,725). CONCLUSÃO: O sangramento gastrointestinal alto mais comum é de origem varicosa em centro de referência terciária. O AIMS65 é o melhor escore simples, com uma pontuação de ≥2 para prever o ressangramento e a mortalidade.