Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Rakesh Kumar Saroj has a total of 51 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Prevalence and academic impact of allergic rhinitis among medical students in Sikkim

    by Garima Talukdar, Santosh Prasad Kesari, Rakesh Kumar Saroj, Krishanu Rudra Pal, Himangshu Hazarika

  • Open Access

    Early risk prediction in endometrial cancer using biomarker and machine learning models: a systematic review and meta-analysis protocol

    by Priya Giri, Rakesh Kumar Saroj

    Globally, endometrial cancer (EC) is among the most prevalent gynaecological cancers, with rising incidence driven by demographic and metabolic changes. Accurate early prognostic assessment is essential to guide personalized treatment and improve long-term outcomes. Traditional clinicopathological prognostic methods-including FIGO (International Federation of Gynaecology and Obstetrics) staging, tumor grade, and histological subtype remain the current standard but have limitations in detecting early recurrence risk. In recent years, biomarker-based prognostic signatures and machine-learning (ML) guided risk models, including radiomics and multi-omics approaches, have emerged as promising tools for improving prognostic accuracy. However, no systematic review and meta-analysis have comprehensively evaluated whether these biomarker or ML-based prognostic models offer improved early risk prediction compared with conventional methods in EC. We will undertake a comprehensive search of MEDLINE, Scopus, IEEE Xplore, and Web of Science from 2010 to 2025. Two reviewers will independently perform title/abstract screening, full-text assessment, data extraction, and quality appraisal, with arbitration by a third reviewer where necessary. Studies developing, validating, or evaluating biomarker-based, radiomics-based and ML-guided prognostic models in human EC patients will be eligible for inclusion. Purely laboratory-based biomarker discovery studies that do not develop or evaluate a prognostic model in human participants will be excluded. Outcomes of interest include measures of prognostic performance such as AUC (area under the curve), C-index (concordance index), time-dependent AUC, calibration metrics, hazard ratios for survival outcomes, and decision analytic measures. Risk of bias will be assessed using PROBAST (Prediction model Risk of Bias Assessment Tool) and QUIPS (Quality in Prognosis Studies), and certainty of evidence will be summarized using GRADE (Grading of Recommendations Assessment, Development and Evaluation). When feasible, meta-analysis of performance metrics will be performed using random-effects models with restricted maximum likelihood (REML) estimation, with subgroup analysis restricted to histological subtype when sufficient data permits. This systematic review will synthesize current evidence on the prognostic performance of biomarker and ML-guided prognostic models in EC, comparing them with conventional clinicopathological risk assessment methods. Findings will help clarify whether emerging model types offer meaningful improvements in the early prognostic accuracy, provide insight into methodological limitations of existing models, and identify gaps in evidence requiring future research. This review will support clinicians, researchers, and policymakers seeking to integrate precision-based prognostic tools into EC care. PROSPERO CRD420251230895.


  • Open Access

    Predictive Modeling of Under-Five Mortality Determinants using Machine Learning Techniques

    by Rakesh Saroj, Pawan Yadav, Vizovonuo Visi

    Background Under-five mortality is one of the major public health issues and directly influences the population’s health, social development, and economic status of countries. Thus, early detection is essential to find what efficient prevention can take to save it. Therefore, this study will explain how machine-learning techniques can help predict the important determinants of under-five mortality in India. Methods This study used data from the National Family Health Survey-V of India. We performed the tenfold cross-validation to assess the model’s capability in the dataset. The decision tree, random forest, logistic regression, neural network, ridge regression, k-nearest neighbor, and naive Bayes models were used in under-five mortality data, and metrics like confusion matrix, accuracy, recall, precision, F1-score, Cohen kappa, and the area under receiver operative characteristics (AUROC) were used to assess the predictive power of the models. The chi-square scores, recursive feature elimination, extra tree classifier, random forest importance, sequential feature selector, and traditional logistic regression were used to predict the important features(factors) of under-five mortality. All computational algorithms were done with the help of SPSS-27 and Jupiter notebook (inbuilt Python 3.3) software. Result The result reveals that the random forest model was the best predictive model compared to other ML models for under-five mortality. The Random Forest model’s precision was estimated to be 98.88% for all factors and 96.25% for important selected variables. After that, neural network accuracy was 96.52%, and accuracy was 94.83% with important variables. Traditional logistic regression accuracy was 93.99% and 93.51%, respectively. The number of living children, breastfeeding status, birth in the last five years, children ever born, time, antenatal care, region, size of children, number of household members, and birth order, were important factors of under-five mortality after using the feature selection methods. Conclusions This is the first study of India to use machine learning approaches to find the important ML predictive model and determine the causative factors for under-five mortality. The random forest model predicted the most important factors with the highest accuracy of under-five mortality. This machine-learning approach can be used as reference concepts to understand students, non-computing professionals, healthcare professionals, and decision-makers in various real-world situations and application areas, particularly from the technical point of view.

  • Open Access

    Investigation of an Acute Gastrointestinal Illness Outbreak Linked to Drinking Water in a Higher Educational Institute in East Sikkim, India

    by Karma G Dolma, Madhuchhanda Das, Shanmuga S Saravanabhavan, Rachana Khati, Goutam Chowdhury, Jayden L Bhutia, Vizovonuo Visi, Rakesh K Saroj, Thandavarayan Ramamurthy

    Introduction An acute gastrointestinal illness outbreak was reported in a higher educational institution among students and faculties in East Sikkim, India, from January to February 2023. The investigation was conducted to identify the source of the infection and causative pathogens and prevent the spread of the outbreak. Methods We defined a case as three or more loose stools in 24 hours, abdominal pain, or vomiting with the onset of symptoms between January 16 and February 16, 2023. Active surveillance was conducted by reviewing the affected individuals at the campus and patient registers at the dispensary, where cases were treated. Stool samples, rectal swabs, water samples, and suspected food samples were collected for microbiological testing using conventional culture, multiplex polymerase chain reaction (PCR), and kit-based real-time PCR methods. Results Out of 1,850 residents, 106 (5.7%) were affected by gastrointestinal symptoms like diarrhea, vomiting, etc. The attack rate for females was 23 (1.24%) and for males was 83 (4.49%). The most affected individual median age was 21 years (range: 2-51 years). From the laboratory investigations, most of the cases demonstrated polymicrobial etiologies. Gastroenteritis pathogens like Campylobacter, astrovirus, androtavirusdiarrheagenic Escherichia coli (DEC) (EAEC, EIEC, ETEC, EPEC, and EAEC), Shigella,etc., were detected in the suspected samples. The environmental investigation indicated the presence of rusted and leaky water pipes and sewage pipelines, along with ineffective chlorination of the water plant. Conclusions Based on epidemiological and laboratory investigations, it is conjectured that sewage and fecal contamination of drinking water and poor maintenance of the water distribution system most likely caused the outbreak described in this study. Basic treatment modalities, adequate chlorination, and periodic inspection of the water system were suggested, which controlled the outbreak to a greater extent.


  • Prediction and Classification Machine Learning Models in Diabetic Studies

    by Rakesh Kumar Saroj, Kanchan Yadav

    Diabetes is one of the major health problems globally. Diabetes is a metabolic disorder that causes various complications such as stroke, chronic kidney disease, foot ulcer, retinopathy, diabetic ketoacidosis, and many more. Timely predictions and classifying of the diabetes problem will help to prevent diabetes. There is various machine-learning approach that can be used to predict and classify diabetes problem in the general population. The diabetes prediction and classification process can help to further improve the disease. This paper will discuss the machine learning applications and steps to understanding the concepts. The purpose of this paper framework is to early prediction and classification of diabetes to reduce stress and save human life, and financial burden.

  • Open Access

    Predictive Modelling of Under-Five Mortality Determinants Using Machine Learning Techniques

    by Rakesh Kumar Saroj, Pawan Kumar Yadav, Vizovonuo Visi

    Background Under-five mortality is one of the major public health issues and directly influences the population’s health, social development, and economic status of countries. Thus, early detection is essential to find what efficient prevention can take to save it. Therefore, this study will explain how machine-learning techniques can help predict the important determinants of under-five mortality in India. Methods This study used data from the National Family Health Survey-V of India. We performed the tenfold cross-validation to assess the model’s capability in the dataset. The decision tree, random forest, logistic regression, neural network, ridge regression, k-nearest neighbor, and naive Bayes models were used in under-five mortality data, and metrics like confusion matrix, accuracy, recall, precision, F1-score, Cohen kappa, and the area under receiver operative characteristics (AUROC) were used to assess the predictive power of the models. The chi-square scores, recursive feature elimination, extra tree classifier, random forest importance, sequential feature selector, and traditional logistic regression were used to predict the important features(factors) of under-five mortality. All computational algorithms were done with the help of SPSS-27 and Jupiter notebook (inbuilt Python 3.3) software. Result The result reveals that the random forest model was the best predictive model compared to other ML models for under-five mortality. The Random Forest model’s precision was estimated to be 98.88% for all factors and 96.25% for important selected variables. After that, neural network accuracy was 96.52%, and accuracy was 94.83% with important variables. Traditional logistic regression accuracy was 93.99% and 93.51%, respectively. The number of living children, breastfeeding status, birth in the last five years, children ever born, time, antenatal care, region, size of children, number of household members, and birth order, were important factors of under-five mortality after using the feature selection methods. Conclusions This is the first study of India to use machine learning approaches to find the important ML predictive model and determine the causative factors for under-five mortality. The random forest model predicted the most important factors with the highest accuracy of under-five mortality. This machine-learning approach can be used as reference concepts to understand students, non-computing professionals, healthcare professionals, and decision-makers in various real-world situations and application areas, particularly from the technical point of view.


  • Open Access

    Development and validation of a novel scale for antiretroviral therapy readiness among pregnant women in urban Zambia with newly diagnosed HIV infection

    by Mwangelwa Mubiana-Mbewe, Samuel Bosomprah, Rakesh Kumar Saroj, Jillian Kadota, Aybuke Koyuncu, Kusanthan Thankian, Michael J. Vinikoor

    Background: Women who are newly diagnosed with HIV infection during pregnancy may not be ready to immediately initiate lifelong antiretroviral therapy (ART; called Option B +) as is recommended. Lack of “readiness” drives early disengagement from care and undermines prevention of HIV transmission to infants. Several studies have shown high early attrition of women initiating ART in pregnancy. Although poor ART uptake and adherence have been attributed to various factors including stigma, disclosure issues and structural issues, there is no standard way of determining which pregnant woman will face challenges and therefore need additional support. We developed and validated a novel ART readiness tool in Lusaka, Zambia. Methods: The aim of this study was to develop and validate a tool that could be used to assess how ready a newly diagnosed pregnant woman living with HIV would be to initiate ART on the day of diagnosis. Using a mixed method design, we conducted this study in three public-setting health facilities in Lusaka, Zambia. Informed by qualitative research and literature review, we identified 27 candidate items. We assessed content validity using expert and target population judgment approaches. We administered the 27-item questionnaire to 454 newly diagnosed pregnant women living with HIV, who were enrolled into a randomized trial (trials number NCT02459678). We performed item reduction analysis and used Cronbach’s alpha coefficient of 0.70 as threshold for reliability. Results: A total of 454 pregnant women living with HIV enrolled in the study between March 2017 and December 2017; 452 had complete data for analysis. The correlation coefficient between the 27 items on the completed ART readiness scale ranged from 0.31 to 0.70 while item discrimination index ranged from -0.01 to 2.38. Sixteen items were selected for the final scale, representing three domains, which we classified as “internalized and anticipated HIV stigma”, “partner support” and “anticipated structural barriers”. Conclusion: We developed and validated a tool that could be used to assess readiness of newly diagnosed women living with HIV to initiate ART. This ART readiness tool could allow clinics to tailor limited resources to pregnant women living with HIV needing additional support to initiate and remain on ART.


  • Open Access

    Machine Learning Algorithms for understanding the determinants of under-five Mortality

    by Rakesh Kumar Saroj, Pawan Kumar Yadav, Rajneesh Singh, Obvious.N. Chilyabanyama

    Background: Under-five mortality is a matter of serious concern for child health as well as the social development of any country. The paper aimed to find the accuracy of machine learning models in predicting under-five mortality and identify the most significant factors associated with under-five mortality. Method: The data was taken from the National Family Health Survey (NFHS-IV) of Uttar Pradesh. First, we used multivariate logistic regression due to its capability for predicting the important factors, then we used machine learning techniques such as decision tree, random forest, Naïve Bayes, K- nearest neighbor (KNN), logistic regression, support vector machine (SVM), neural network, and ridge classifier. Each model’s accuracy was checked by a confusion matrix, accuracy, precision, recall, F1 score, Cohen’s Kappa, and area under the receiver operating characteristics curve (AUROC). Information gain rank was used to find the important factors for under-five mortality. Data analysis was performed using, STATA-16.0, Python 3.3, and IBM SPSS Statistics for Windows, Version 27.0 software. Result: By applying the machine learning models, results showed that the neural network model was the best predictive model for under-five mortality when compared with other predictive models, with model accuracy of (95.29% to 95.96%), recall (71.51% to 81.03%), precision (36.64% to 51.83%), F1 score (50.46% to 62.68%), Cohen’s Kappa value (0.48 to 0.60), AUROC range (93.51% to 96.22%) and precision-recall curve range (99.52% to 99.73%). The neural network was the most efficient model, but logistic regression also shows well for predicting under-five mortality with accuracy (94% to 95%)., AUROC range (93.4% to 94.8%), and precision-recall curve (99.5% to 99.6%). The number of living children, survival time, wealth index, child size at birth, birth in the last five years, the total number of children ever born, mother’s education level, and birth order were identified as important factors influencing under-five mortality. Conclusion: The neural network model was a better predictive model compared to other machine learning models in predicting under-five mortality, but logistic regression analysis also shows good results. These models may be helpful for the analysis of high-dimensional data for health research.


  • Open Access

    Development and validation of a novel scale for antiretroviral therapy readiness among pregnant women in urban Zambia with newly diagnosed HIV infection

    by Mwangelwa Mubiana-Mbewe, Samuel Bosomprah, Rakesh K. Saroj, Jillian Kadota, Aybuke Koyuncu, Thankaian Kusanathan, Michael J. Vinikoor

    BackgroundPregnant women who are newly diagnosed with HIV infection may not be ready to initiate lifelong antiretroviral therapy (ART; called Option B+) on the day of HIV diagnosis as is recommended. This drives early disengagement from care and undermines prevention of HIV transmission to infants. Several studies have shown high early attrition of women initiating ART in pregnancy. Although various reasons including stigma, disclosure issues and structural issues have been proposed as contributing to poor ART uptake and adherence, there is no standard way of determining which pregnant woman will face challenges and therefore needs additional support. We developed and validated a novel Option B+ readiness tool in Lusaka, Zambia. MethodsThe aim of this study was to develop and validate a tool that could be used to assess how ready a newly diagnosed HIV positive pregnant woman would be to initiate ART on the day of diagnosis. Using a mixed method design, we conducted this study in three public-setting health facilities in Lusaka, Zambia. Informed by qualitative research and literature review, we identified 27 candidate items. We assessed content validity using expert and target population judgment approaches. We administered the 27-item questionnaire to 454 newly diagnosed HIV positive pregnant women, who were enrolled into a randomized trial (trials number NCT02459678). We performed item reduction analysis and used Cronbach’s alpha coefficient of 0.70 as threshold for reliability. ResultsA total of 454 HIV-infected pregnant women enrolled in the study between March 2017 and December 2017; 452 had complete data for analysis. The correlation coefficient between the 27 items on the completed ART readiness scale ranged from 0.31 to 0.70 while item discrimination index ranged from -0.01 to 2.38. Sixteen items were selected for the final scale, representing three domains, which we classified as “internalized and anticipated HIV stigma”, “partner support” and “anticipated structural barriers”. ConclusionWe developed and validated a tool that could be used to assess readiness of newly diagnosed HIV positive women to initiate ART. This ART readiness tool could allow clinics to tailor limited resources to pregnant women living with HIV needing additional support to initiate and remain on ART.