Development of an ensemble CNN model with explainable AI for the classification of gastrointestinal cancer.

The implementation of AI assisted cancer detection systems in clinical environments has faced numerous hurdles, mainly because of the restricted explainability of their elemental mechanisms, even though such detection systems have proven to be highly effective. Medical practitioners are skeptical about adopting AI assisted diagnoses as due to the latter's inability to be transparent about decision making processes. In this respect, explainable artificial intelligence (XAI) has emerged to provide explanations for model predictions, thereby overcoming the computational black box problem associated with AI systems. In this particular research, the focal point has been the exploration of the Shapley additive explanations (SHAP) and local interpretable model-agnostic explanations (LIME) approaches which enable model prediction explanations. This study used an ensemble model consisting of three convolutional neural networks(CNN): InceptionV3, InceptionResNetV2 and VGG16, which was based on averaging techniques and by combining their respective predictions. These models were trained on the Kvasir dataset, which consists of pathological findings related to gastrointestinal cancer. An accuracy of 96.89% and F1-scores of 96.877% were attained by our ensemble model. Following the training of the ensemble model, we employed SHAP and LIME to analyze images from the three classes, aiming to provide explanations regarding the deterministic features influencing the model's predictions. The results obtained from this analysis demonstrated a positive and encouraging advancement in the exploration of XAI approaches, specifically in the context of gastrointestinal cancer detection within the healthcare domain.

Re-visiting the COVID-19 analysis using the class of high ordered integer-valued time series models with harmonic features.

The COVID-19 series is obviously one of the most volatile time series with lots of spikes and oscillations. The conventional integer-valued auto-regressive time series models (INAR) may be limited to account for such features in COVID-19 series such as severe over-dispersion, excess of zeros, periodicity, harmonic shapes and oscillations. This paper proposes alternative formulations of the classical INAR process by considering the class of high-ordered INAR models with harmonic innovation distributions. Interestingly, the paper further explores the bivariate extension of these high-ordered INARs. South Africa and Mauritius' COVID-19 series are re-scrutinized under the optic of these new INAR processes. Some simulation experiments are also executed to validate the new models and their estimation procedures.

An accelerated infliximab induction regimen reduces the need for early colectomy in patients with acute severe ulcerative colitis.

BACKGROUND & AIMS: Administration of infliximab to patients with acute severe ulcerative colitis (ASUC) (rescue therapy) can reduce the rate of early colectomy (within 12 months), but long-term rates of colectomy are the same as those of the pre-biologic era for these patients. The half-life of infliximab is shorter in patients with ASUC than in patients with non-severe UC, so more frequent dosing might be required to produce a therapeutic effect. METHODS: We performed a retrospective analysis of 50 hospitalized patients who received infliximab for steroid-refractory ASUC at a single academic center from September 2005 through 2013. In 2011 an accelerated dosing strategy for infliximab was introduced; we compared outcomes of standard and accelerated dosing regimens. One group of patients (n = 35) were placed on a standard dosing regimen for infliximab and then given the drug at 0, 2, and 6 weeks and then every 8 weeks thereafter. A second group (n = 15) were placed on an accelerated regimen and received 3 induction doses of infliximab within a median period of 24 days. Rates of colectomy were compared between the groups during induction and follow-up periods. RESULTS: There were no differences between groups in median baseline levels of C-reactive protein, albumin, or hemoglobin. The rate of colectomy during induction therapy was significantly lower with the accelerated regimen (6.7%, 1 of 15) than with the standard regimen (40%, 14 of 35) (Fisher exact test, P = .039). The standard regimen was associated with shorter time to colectomy (log-rank test, P = .042). Among patients who completed induction therapy, subsequent need for colectomy was similar between the groups during the follow-up period. Multivariate analysis showed that factors independently associated with successful induction therapy were level of albumin (g/L) when the treatment began (P = .003) and the accelerated dosing regimen (P = .03). CONCLUSIONS: In patients with ASUC, an accelerated infliximab induction strategy reduces the need for early colectomy. An intensified infliximab dosing strategy in response to clinical or laboratory signs of breakthrough inflammation merits consideration in prospective studies.

Gastroparesis and Parkinson's disease: a systematic review.

Some of the gastrointestinal (GI) symptoms commonly experienced by patients with Parkinson's disease (PD) have been attributed to gastroparesis; however, the precise prevalence and relevance of gastric emptying delay in PD is unclear. The definition of gastroparesis varies; currently the most widely accepted definition (from the National Institute of Diabetes and Digestive and Kidney Diseases Gastroparesis Clinical Research Consortium) is the presence of appropriate symptoms (including nausea, retching, vomiting, stomach fullness, and inability to finish a meal) for ≥ 12 weeks, together with delayed gastric emptying on scintigraphy and the absence of any obstructive lesions on upper GI endoscopy. In PD patients, gastroparesis has the potential to affect nutrition and quality of life, as well as the absorption of PD medications, including L-dopa. This reduced absorption of L-dopa has implications for the control of the PD motor symptoms for which it is administered. We performed a systematic review of the literature on gastroparesis in PD with the aim of developing an evidence-based approach to its management. Based on this review, we conclude that while gastric emptying has been reported to be frequently delayed in PD, the existing data do not permit definitive conclusions concerning its true prevalence, relationship to the underlying disease process, relevance to PD management, or the optimal therapy of related GI symptoms. Further study of these important issues is, therefore, required.

Motility disorders in the patient with neurologic disease.

Gastrointestinal symptoms are common in the patient with chronic neurologic disease and may loom large in terms of impact on quality of life and on nutrition and mobility. A knowledge of the range of gastrointestinal disorders associated with a given neurologic disease, together with an understanding of the risks and benefits of various therapeutic options and approaches, should aid gastroenterologists in their efforts to contribute to the care of these patients. In most instances a multidisciplinary team (neurologist/neurosurgeon, gastroenterologist, nutritionist, therapist, specialist nurse) aware of the wishes and needs of the family and their carers and mindful of the nature and the natural history of the underlying disease process are best placed to assess and manage these problems.

Biliary sphincter of Oddi dysfunction: response rates after ERCP and sphincterotomy in a 5-year ERCP series and proposal for new practical guidelines.

INTRODUCTION: The diagnosis and management of sphincter of Oddi dysfunction (SOD) is challenging. AIMS: Retrospective review of endoscopic retrograde cholangiopancreatographies (ERCPs) performed for biliary SOD between 2003 and 2007 and proposal for new management guidelines. METHODS: Patients were identified from the ERCP database. All patients had a history of typical biliary pain and were classified into three groups - group 1: patients with dilated common bile duct and abnormal laboratory investigations at time of pain; group 2: either of the above; group 3: none of the above. All patients underwent ERCP and endoscopic sphincterotomy (ES). Sphincter of Oddi manometry was not carried out. Patients were contacted 18 months after procedure with regard to relief of pain and complications. RESULTS: Seventy-two patients were included - 16.7% male and 83.3% female with an average age of 54.2 years. Ampullary stenosis was found in 76.2% of patients in group 1 and 70% in group 2 compared with 30% in group 3. In group 1, 90.5% of patients reported complete relief of pain compared with 75 and 50% in groups 2 and 3, respectively. Among them 55.6% of patients had previous cholecystectomy. Overall 85% of cholecystectomized patients and 81.3% of patients with an intact gallbladder had complete response to ES. There were no complications in all three groups. DISCUSSION: Our findings are similar to other studies documenting response rates to ES in manometry proven SOD. Our simpler guidelines are more applicable to clinical practice. Group 3 remains the most challenging group. Risk and benefits of ERCP and ES should be discussed with patients in centres where sphincter of Oddi manometry is unavailable, noting that 50% of patients respond favourably to ES.