@kimskarad.in
Medicine
Kims deemed to be University Karad
MD medicine
Medicine
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Et al. Uddhav T. Kumbhar
Science Research Society
With the growing integration of Internet of Things (IoT) technologies and Artificial Intelligence (AI) in healthcare, it is crucial to prioritize transparency and interpretability in the decision-making process. This paper presents a novel framework that utilizes Explainable AI (XAI) to improve the interpretability of predictive healthcare models. The proposed system integrates feature importance-based methodologies with the Local Interpretable Model-agnostic Explanations (LIME) technique to offer a comprehensive comprehension of the predictive and preventive healthcare recommendations. The framework commences by conducting an in-depth examination of the present condition of Internet of Things (IoT) in the healthcare sector, as well as the importance of predictive and preventive healthcare. The literature review examines the difficulties related to the comprehensibility of artificial intelligence (AI) in the healthcare field and presents feature importance-based approaches and LIME as potential remedies. The focus is on the hybrid approach that combines these techniques, as it has the potential to offer precise predictions while also ensuring a strong level of interpretability. The methodology section delineates the procedure for gathering healthcare data and IoT sensor data, subsequently followed by preprocessing measures such as data cleansing and feature engineering. The predictive models undergo a process of selection, training, and evaluation, with the primary objective of attaining a notable accuracy level of 0.961. This text provides a detailed explanation of how the combination of feature importance-based approaches and LIME improves the transparency and interpretability of the model. An extensive case study is provided to illustrate the implementation of the suggested framework in an actual situation. The results and evaluation section showcases the exceptional precision of 0.961, as well as enhanced interpretability scores and decreased computational time in comparison to the baseline XAI models. The discussion section juxtaposes the suggested hybrid approach with conventional models, examines ethical considerations, and investigates the scalability and generalizability of the framework. To conclude, the paper provides a concise overview of the findings and implications of the Explainable AI-Powered IoT Systems for Predictive and Preventive Healthcare framework. This hybrid approach demonstrates high accuracy, improved interpretability, and efficient computational performance, making it a promising advancement in personalized health management and wellness optimization. This research adds to the expanding collection of literature on Explainable Artificial Intelligence (XAI) in the healthcare sector, thus opening up possibilities for future research avenues and practical applications in this domain.
KailasD Datkhile, PratikP Durgawale, VirendraC Patil, VasantV Devkar, SatishV Kakade, SarjeraoA Dabane, and VijaykumarS Wader
Medknow
Background: Type II diabetes mellitus (T2DM) is increasing at an alarming rate in urban as well as rural parts of India. Change in life-style, dietary habits, aging, environmental factors contribute to developing this disorder, T2DM is a polygenic disorder which involves large number of genes interacting with each other and the environment to manifest itself phenotypically. Aims and Objectives: In the present study, we investigated the association of Vitamin D receptor (VDR) and Paraoxonase I (PON I) gene polymorphisms with T2DM patients among rural population in South western Maharashtra. Materials and Methods: The study groups included healthy non-diabetic control group (n = 120), non-obese diabetic group (n = 120), obese diabetic group (n = 120) from the same geographical region. The single nucleotide polymorphisms (SNPs) amongst these groups were studied using Polymerase chain reaction- restriction fragment length polymorphism (PCR-RFLP) for VDR and PON I genes. Results and Conclusion: Amongst the studied genes VDR SNP rs1544410 [OR 0.24 (0.1-0.63); P 0.005] and Paraoxonase I SNP rs854560 (OR 0.47 (0.26-0.84); P 0.015) had a protective effect against T2DM for non-obese individuals. Whereas only VDR SNP rs1544410 [OR 0.025 (0.003-0.91); P 0.001] had a protective effect in case of obese individuals. The findings suggest greater gene-gene interaction and gene-environment interactions influence the phenotypic outcomes of genetic polymorphisms.
Virendra C Patil, Suresh J. Bhosale, Ashok Kshirsagar, Harsha V. Patil, and Sanjay S. Patil
VM Media SP. zo.o VM Group SK
Background: Laboratory parameters play a key role in triaging, predicting disease course, severity and may determine prognosis COVID-19 patients. Material and methods: Aim and Objectives: To study the relation of clinical and laboratory parameters (total WBC count, neutrophil: lymphocyte ratio, serum ferritin, serum D-dimer, serum LDH, CRP, ESR) with severity and outcome of Corona Virus Disease (COVID-19) confirmed by real-time RT-PCR. Sample size: It was a time-bound study conducted over 3 months (1st April to 30th June, 2020). A total of 206 patients will be included in this study satisfying the inclusion criteria. Study design: This was a prospective, observational and non-interventional study conducted on patients with laboratory-confirmed COVID-19 admitted in a tertiary care teaching hospital. Statistical Analysis: Data will be analyzed for mean, percentage, standard deviation, and chi-square test for quantitative data by using appropriate statistical tests using INSTAT software version 8.0 (trial version) and p-value < 0.05 will be considered statistically significant. Results: Total of 206 patients of both genders were included in the present study. Total 141 (68.44%) patients were males and 65 (31.55%) patients were females (Chi Sq. 56.07;DF:1;p < 0.001). Among all the groups according to the severity of illness, 'D' group was the most common group (n = 99;45.06%). Age > 60 years (17.48%), obesity (13.11%), hypertension (10.19%), COPD (5.83%), and diabetes mellitus (5.83%) were the most frequent risk factors or comorbidities associated with COVID-19 disease. Many patients had multiple risk factors in the present study. The majority (3/4th) of the patients were in C and D group (moderate) with co-morbidities and about 1/4th were in the severe group. Total 5 (2.43%) patients with COVID-19 patients succumbed to death with an overall case fatality rate of 2.43%. The case fatality rate was significantly higher among the patients with risk factors or comorbidities (p = 0.0124). Late presentation, associated comorbidity, advancing age, High level of ferritin, D-dimer, CRP, PaO2/FiO(2) ratio less than 100 at the time of admission were associated with mortality. The terminal event in patients who have succumbed was bradycardia followed by cardiorespiratory arrest. The cause of death was ARDS with bilateral extensive pneumonia. Conclusions: Late presentation, associated comorbidity like diabetes mellitus, advancing age, High level of ferritin, D-dimer, CRP, PaO2/FiO(2) ratio less than 100 at the time of admission were associated with mortality. The terminal event in patients who have succumbed was bradycardia followed by cardiorespiratory arrest. The cause of death was ARDS with bilateral extensive pneumonia.
Pradip C Warghane, V. C. Patil V. C. Patil, Aparna Patange, and Shilpa Patil
Oriental Scientific Publishing Company
Background and Objectives: Anemia is the most common condition in older population. Failure to diagnose anemia may lead to delayed diagnosis of treatable conditions. Therefore, this study was conducted to study the clinical, biochemical, and hematological profile of elderly anemic patients in a tertiary care center. Materials and Method: This prospective, observational study was conducted in 207 elderly patients with hemoglobin (Hb) <12 g/dL in females and <13 g/dL in males. A detailed history was obtained and thorough clinical examination and symptom analysis was done in all the patients. Hematological investigations including peripheral smear and biochemical investigations were conducted. The morphological pattern of anemia was analyzed based on red cell indices and peripheral smear. Pearson’s Chi-square test was used to test the association between presenting symptoms and anemia grade. P<0.05 was considered statistically significant. Results: Majority of the patients presented with moderate anemia, and the most common presenting symptoms were generalized weakness and anorexia. A significant association was observed between severity of anemia and fatigue (p=0.006), headache (p=0.04), and dyspnea (p=0.010). Patients with pallor (P=0.01) and nail changes (P=0.0005) had significantly severe anemia. All the hematological parameters were within the reference range. Normocytic anemia was the most common type of morphological anemia among elderly patients . Conclusion: The incidence of anemia was quite high among elderly patients with chronic diseases and malignancies. Hence, it is very important to diagnose the cause of anemia by detailed investigations before initiating the required therapy.
Pratik P. Durgawale, Kailas D. Datkhile, Virendra C. Patil, Vasant V. Devkar, Sarjerao A. Dabane, Vijaykumar S. Wader, and Satish V. Kakade
Oriental Scientific Publishing Company
The most commonly found type of diabetes in India is type II diabetes mellitus (T2DM), which is characterized by decrease in insulin secretion and decrease in insulin sensitivity. Several environmental factors, genetic factors, socio-economic factors, life style, dietary habits have contributed to the surge of T2DM cases in India. Numerous genes involved in lipid metabolism are likely to be candidates as the markers for obesity and T2DM. In the present study, single nucleotide polymorphism (SNP) of two genes namely Apolipoprotein A5 (APOA5) and Lipoprotein lipase (LPL) involved in triglyceride metabolism were investigated using polymerase chain reaction- restriction fragment length polymorphism (PCR-RFLP). The control group comprised of non-obese, non-diabetic subjects (n=120) and T2DM cases were divided into obese (n=120), and non-obese (n=120) groups based on their body mass index (BMI). The demographic features between the control and cases were compared using Chi-square distribution. The genotype frequencies of control and cases were compared using analysis of variance (ANOVA) and binary logistic regression analysis (Odds’ ratio (OR) and adjusted Odds’ ratio). It was observed that APOA5 rs3135506 (OR = 0.46 (0.27-0.79); p = 0.007) was negatively associated, while APOA5 rs662799 (OR = 2.22 (1.28-3.84); p = 0.006) was significantly associated in non-obese diabetic patients. APOA5 rs3135506 (OR = 0.03 (0.01-0.06); p < 0.001) was negatively associated and rs662799 (OR = 4.68 (1.47-14.93); p = 0.01) was significantly associated in obese diabetic patients. Both LPL SNPs (rs285 and rs320) were found not to be associated with T2DM. The association of Apo A5 variants with T2DM may be because of post transcriptional inhibition leading to reduced Apo A5 expression or these alleles may be in linkage disequilibrium with alleles which directly affect the functioning of APOA5. The observations indicated that T2DM is a multi-factorial disease with a large number of gene-gene and gene-environment interactions.
AniketBhaskar Avhad and VirendraChandrashekhar Patil
Medknow
Background: Diabetes is an important risk factor of cardiomyopathy which can eventually lead to acute or chronic heart failure. Many factors such as obesity, hyperglycemia, suboptimal glycemic control, and hyperlipidemia predispose as well as aggravate long-term complications in diabetes mellitus (DM). Aim: The present study was conducted to assess the relation of glycosylated hemoglobin (HbA1c) and obesity indices (body mass index [BMI], waist circumference [WC]) with diastolic dysfunction in participants with type 2 DM (T2DM). Methods: This was a cross-sectional study conducted at the teaching hospital during a 1-year period. All 100 consecutive T2DM participants were investigated for HbA1c, blood sugars and lipid profile parameters, and echocardiography. Standardized transthoracic echocardiographic examination was performed (American Society of Echocardiography) using Vivid S6 ultrasound machine with 4–5 MHz probe. Left ventricular diastolic dysfunction (LVDD) was labeled if 3 or more of these variables were abnormal: “Septal e <7 cm/s, lateral e <10 cm/s, E/e' >14, LA volume index >34 mL/m2, and Peak TR velocity >2.8 m/s.” Results: Out of the total 100 diabetics, 37% had diastolic dysfunction (Grade I: 30 [81.08%] and Grade II: 7 [18.91%]). The peak early transmitral filling wave velocity (E) in diabetics with obesity was low as compared to diabetics with normal weight (44 ± 11 cm/s vs. 52 ± 7.5 cm/s). Higher BMI was associated with higher peak late transmitral velocity (A) (61 ± 14 cm/s vs. 38 ± 6.24 cm/s). The mean of E/A ratio was significantly lower in diabetics with obesity. The E/e' ratio was higher in diabetics with obesity (14.8 ± 2.8 vs. 12.86 ± 2.14). The mean of WC was 90.18 ± 6.82 cm in participants of diastolic dysfunction, whereas it was 86.46 ± 6.42 cm in those without dysfunction. Diastolic dysfunction was present in participants with a suboptimal glycemic control, HbA1c ≥ 6.% (40%), with longer duration of diabetes (62.16%) (P = 0.003). Thus, a higher BMI was associated with higher diastolic dysfunction (P = 0.027). Conclusions: The present study reveals burden of LVDD in asymptomatic T2DM participants with its significant association with age, obesity indices (WC and BMI), HbA1c levels, and longer duration of diabetes. This highlights the magnitude of obesity and suboptimal glycemic control that should be considered in the management of cardiac complications in the form of development of diastolic dysfunction. It will benefit in triage for the optimal management of diabetes.
VirendraChandrashekhar Patil, HarshaVirendra Patil, and ShivajiT Mohite
Medknow
Background: Acinetobacter, a nonfermenting Gram-negative coccobacilli, have emerged as significant pathogens causing multidrug-resistant (MDR) ventilator-associated pneumonia (VAP). Metallo-beta-lactamase (MBL)-producing Acinetobacter spp. have become an emerging therapeutic concern worldwide due to the MDR isolates. Aim and Objectives: Phenotypic detection of MBL producing MDR Acinetobacter isolates in patients with VAP and to study the antibiotic susceptibility pattern of MBL-producing isolates. Materials and Methods: This was a prospective observational and noninterventional study conducted on patients with VAP over a period of 2 years. This study was conducted at a tertiary care teaching hospital in the intensive care unit. A total of 164 MBL-producing MDR AB isolates were included in the study. MBL was detected by imipenem-EDTA double-disc synergy test (DDST), imipenem-EDTA combined disc synergy test (CDST-IPM), and MBL-E test. Results: A total of 188 samples were enrolled for the study, fulfilling the inclusion criteria of VAP. Total MDR Acinetobacter spp. isolates were 188 (76.42%) of them, 164 (87.23%) were MBL producing and 24 (12.76%) were nonMBL (P < 0.002). Total 11.17% and 88.83% MDR VAP due to Acinetobacter spp. were early-onset VAP and Late-onset VAP, respectively (P < 0.001). Late-onset VAP due to MDR Acinetobacter spp. was predominant in the present study caused by Acinetobacter spp. Of total 188 MDR Acinetobacter isolates, 156 (82.98%) were Acinetobacter baumannii, 15 (7.98%) were Acinetobacter iwoffii, 9 (4.79%) were Acinetobacter calcoacetiucs, 5 (2.66%) were Acinetobacter hemotyticus, and 3 (1.59%) were ABC complex, predominated by A. baumannii (P < 0.001). Of total 188 MDR Acinetobacter spp. 164 (87.23%) were putative MBL producing and 24 (12.67%) were nonMBL Acinetobacter spp. Of 164 MBL-producing isolates, 141 (85.98%) were detected by the DDST method, and 23 (14.02%) were DDST negative. Total 146 (89.02%) MDR Acinetobacter spp. were detected by a combined disc test-IMP test. A total of 152 (92%) MDR Acinetobacter spp. were detected by MBLe-Test. All MBL-producing MDR Acinetobacter spp. isolates (164) were resistant to piperacillin (PI), piperacillin + tazobactam (PIT), ciprofloxacin (CIP), ceftazidime (CAZ), cefepime (CPM), imipenem (IMP), and meropenem (MRP). The tigecycline (21.34%) resistance was significantly less compared to all other antibiotics. Conclusions: The present study highlighted the burden of MDR MBL producing Acinetobacter spp. in patients with VAP. About three-fourth of patients with VAP had MDR Acinetobacter spp. Eighty percent were MDR Acinetobacter spp. were MBL producers. MDR Acinetobacter isolates, including MBL producer, were significantly higher in late-onset VAP. The ability of phenotypic identification of Acinetobacter spp. for MBL producer among imipenem-EDTA double-disc synergy test (DDST), CDST-IPM and MBL-E Test were comparable. All MBL-producing MDR Acinetobacter spp. isolates were resistant to PI, Ciprofloxacin, CAZ, CPM, IMP, and MRP. The tigecycline resistance was significantly less (1/5th). The study of antibiotic sensitivity patterns and screening for MBL production among A. baumannii isolates is essential for controlling Acinetobacter infections. The judicious use of antimicrobial therapy and combined approaches of rotational antibiotic therapy is strongly suggested.
Purushottam Narayandas Toshniwal and Virendra C Patil
GP Innovations Pvt. Ltd.
The present observational study was conducted to evaluate the cardiological manifestations in chronic alcoholic in relation with Electrocardiographic and 2-Dimensional Transthoracic Echocardiogram and Doppler findings. In Present study we observed most common 2-D echocardiographic finding being Grade 1 Diastolic Dysfunction (i.e. 28.35%), followed by Left ventricular Dysfunction (i.e.7.87%) being second most common finding, followed by mild Pulmonary Hypertension (i.e. 5.51%). In present study, the observed most common Electrocardiographic finding being sinus rhythm ( i.e. 34.6%) followed by second most common finding being atrial fibrillation (i.e. 13.38%) followed by sinus tachycardia ( i.e. 11.18%) so most common abnormal elctro-cardiographic finding is Atrial fibrillation. The present study was crosssectional, descriptive, hospital based observational study to assess cardiac function in chronic alcoholic by use of Electrocardiogram and Two Dimensional Transthoracic Echocardiogram and Doppler function. Duration of Study is 18 months and Study Setting is conducted in patients admitted in wards and ICU under Department of Medicine, Krishna Hospital and Medical Research Centre, Karad. There is Study Population is all alcoholic patients admitted in wards and ICU during the period from October 2016 to March 2018 under Department of Medicine, Krishna Hospital and Medical Research Centre Karad with alcohol consumption duration more than 5 years.
VirendraC Patil, AniketB Avhad, AkshayR Kulkarni, and KaustubhA Pandere
Medknow
12 years