Diabetes prediction using machine learning and explainable AI techniques.

Globally, diabetes affects 537 million people, making it the deadliest and the most common non-communicable disease. Many factors can cause a person to get affected by diabetes, like excessive body weight, abnormal cholesterol level, family history, physical inactivity, bad food habit etc. Increased urination is one of the most common symptoms of this disease. People with diabetes for a long time can get several complications like heart disorder, kidney disease, nerve damage, diabetic retinopathy etc. But its risk can be reduced if it is predicted early. In this paper, an automatic diabetes prediction system has been developed using a private dataset of female patients in Bangladesh and various machine learning techniques. The authors used the Pima Indian diabetes dataset and collected additional samples from 203 individuals from a local textile factory in Bangladesh. Feature selection algorithm mutual information has been applied in this work. A semi-supervised model with extreme gradient boosting has been utilized to predict the insulin features of the private dataset. SMOTE and ADASYN approaches have been employed to manage the class imbalance problem. The authors used machine learning classification methods, that is, decision tree, SVM, Random Forest, Logistic Regression, KNN, and various ensemble techniques, to determine which algorithm produces the best prediction results. After training on and testing all the classification models, the proposed system provided the best result in the XGBoost classifier with the ADASYN approach with 81% accuracy, 0.81 F1 coefficient and AUC of 0.84. Furthermore, the domain adaptation method has been implemented to demonstrate the versatility of the proposed system. The explainable AI approach with LIME and SHAP frameworks is implemented to understand how the model predicts the final results. Finally, a website framework and an Android smartphone application have been developed to input various features and predict diabetes instantaneously. The private dataset of female Bangladeshi patients and programming codes are available at the following link: https://github.com/tansin-nabil/Diabetes-Prediction-Using-Machine-Learning.

Blinatumomab maintenance after allogeneic hematopoietic cell transplantation for B-lineage acute lymphoblastic leukemia.

Patients with B-lineage acute lymphoblastic leukemia (ALL) are at high-risk for relapse after allogeneic hematopoietic cell transplantation (HCT). We conducted a single-center phase 2 study evaluating the feasibility of 4 cycles of blinatumomab administered every 3 months during the first year after HCT in an effort to mitigate relapse in high-risk ALL patients. Twenty-one of 23 enrolled patients received at least 1 cycle of blinatumomab and were included in the analysis. The median time from HCT to the first cycle of blinatumomab was 78 days (range, 44 to 105). Twelve patients (57%) completed all 4 treatment cycles. Neutropenia was the only grade 4 adverse event (19%). Rates of cytokine release (5% G1) and neurotoxicity (5% G2) were minimal. The cumulative incidence of acute graft-versus-host disease (GVHD) grades 2 to 4 and 3 to 4 were 33% and 5%, respectively; 2 cases of mild (10%) and 1 case of moderate (5%) chronic GVHD were noted. With a median follow-up of 14.3 months, the 1-year overall survival (OS), progression-free survival (PFS), and nonrelapse mortality (NRM) rates were 85%, 71%, and 0%, respectively. In a matched analysis with a contemporary cohort of 57 patients, we found no significant difference between groups regarding blinatumomab's efficacy. Correlative studies of baseline and posttreatment samples identified patients with specific T-cell profiles as "responders" or "nonresponders" to therapy. Responders had higher proportions of effector memory CD8 T-cell subsets. Nonresponders were T-cell deficient and expressed more inhibitory checkpoint molecules, including T-cell immunoglobulin and mucin domain 3 (TIM3). We found that blinatumomab postallogeneic HCT is feasible, and its benefit is dependent on the immune milieu at time of treatment. This paper is posted on ClinicalTrials.gov, study ID: NCT02807883.

Insightful Valorization of the Biological Activities of Pani Heloch Leaves through Experimental and Computer-Aided Mechanisms.

Pani heloch (Antidesma montanum) is traditionally used to treat innumerable diseases and is a source of wild vegetables for the management of different pathological conditions. The present study explored the qualitative phytochemicals; quantitative phenol and flavonoid contents; in vitro antioxidant, anti-inflammatory, and thrombolytic effects; and in vivo antipyretic and analgesic properties of the methanol extract of A. montanum leaves in different experimental models. The extract exhibited secondary metabolites including alkaloids, flavonoids, flavanols, phytosterols, cholesterols, phenols, terpenoids, glycosides, fixed oils, emodines, coumarins, resins, and tannins. Besides, Pani heloch showed strong antioxidant activity (IC50 = 99.00 µg/mL), while a moderate percentage of clot lysis (31.56%) in human blood and significant anti-inflammatory activity (p < 0.001) was achieved with the standard. Moreover, the analgesic and antipyretic properties appeared to trigger a significant response (p < 0.001) relative to in the control group. Besides, an in silico study of carpusin revealed favorable protein-binding affinities. Furthermore, the absorption, distribution, metabolism, excretion, and toxicity analysis and toxicological properties of all isolated compounds adopted Lipinski's rule of five for drug-like potential and level of toxicity. Our research unveiled that the methanol extract of A. montanum leaves exhibited secondary metabolites that are a good source for managing inflammation, pyrexia, pain, and cellular toxicity. Computational approaches and further studies are required to identify the possible mechanism which responsible for the biological effects.

Curculigo recurvata W.T.Aiton exhibits anti-nociceptive and anti-diarrheal effects in Albino mice and an in silico model.

BACKGROUND: Curculigo recurvata (C. recurvata) is an enthnomedicinally important herb reported to have significant medicinal values. The present study aimed to explore the in vivo and in silico anti-nociceptive and anti-diarrheal effects of a C. recurvate rhizome methanol extract (Me-RCR). METHODS: The analgesic effects of Me-RCR were assessed using acetic acid-induced writhing and the formalin-induced flicking test. The drugs were administered intraperitoneally (IP) at doses of 200 and 400 mg/kg body weight (bw). Anti-diarrheal activity was evaluated by assessing intestinal motility, hypersecretion, and fecal score in mice at oral doses of 200 and 400 mg/kg·bw. Computer facilitated analyses for anti-nociceptive and anti-diarrheal activities of three isolated compounds from C. recurvata were undertaken to identify the best-fit phytoconstituents. RESULTS: The Me-RCR showed significant (P < .05) peripheral anti-nociception at the highest dose. The extract inhibited both early and late phases of nociception in the formalin-induced writhing test. In the castor oil-induced diarrhoea model, the extract significantly (P < .05) prolonged the onset time of diarrhoea, inhibited percentage of diarrhoea, and decreased both the volume and weight of intestinal contents. Rates of intestinal fluid accumulation inhibition were (33.61 ± 1.00)% and (46.44 ± 0.89)% at Me-RCR doses of 200 and 400 mg/kg·bw, respectively. Moreover, a significant (P < .05) reduction in gastrointestinal motility was observed. An absorption, distribution, metabolism, excretion and/or toxicity (ADME/T) test showed that the selected compounds yielded promising results, satisfying Lipinski's rule of five for predicting drug-like potential. Notably, of the three phytoconstituents curculigine and isocurculigine possessed the highest affinity for the COX-1 and COX-2. Isocurculigine was also identified as the most effective anti-diarrheal compound in the computer-facilitated model. CONCLUSION: An extract of the plant C. recurvata showed potential analgesic and anti-diarrheal activity due to the presence of one or more active secondary metabolite(s).

Coverage and cost of a large oral cholera vaccination program in a high-risk cholera endemic urban population in Dhaka, Bangladesh.

A feasibility study of an oral cholera vaccine was carried out to test strategies to reach high-risk populations in urban Mirpur, Dhaka, Bangladesh. The study was cluster randomized, with three arms: vaccine, vaccine plus safe water and hand washing practice, and no intervention. High risk people of age one year and above (except pregnant woman) from the two intervention arms received two doses of the oral cholera vaccine, Shanchol™. Vaccination was conducted between 17th February and 16th April 2011, with a minimum interval of fourteen days between two doses. Interpersonal communication preceded vaccination to raise awareness amongst the target population. The number of vaccine doses used, the population vaccinated, left-out, drop out, vaccine wastage and resources required were documented. Fixed outreach site vaccination strategy was adopted as the mode of vaccine delivery. Additionally, mobile vaccination sites and mop-up activities were carried out to reach the target communities. Of the 172,754 target population, 141,839 (82%) and 123,666 (72%) received complete first and second doses of the vaccine, respectively. Dropout rate from the first to the second dose was 13%. Two complete doses were received by 123,661 participants. Vaccine coverage in children was 81%. Coverage was significantly higher in females than in males (77% vs. 66%, P<0.001). Vaccine wastage for delivering the complete doses was 1.2%. The government provided cold-chain related support at no cost to the project. Costs for two doses of vaccine per-person were US$3.93, of which US$1.63 was spent on delivery. Cost for delivering a single dose was US$0.76. We observed no serious adverse events. Mass vaccination with oral cholera vaccine is feasible for reaching high risk endemic population through the existing national immunization delivery system employed by the government.