Zahra Keshavarz
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
Zahra Keshavarz has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Farid Zand
Anesthesiology and Critical Care Research Center, Shiraz University of Medical Sciences, Shiraz, Iran
1 shared publication Mahkameh Rafiee
HIV/STI Surveillance Research Center, and WHO Collaborating Center for HIV Surveillance, Institute for Futures Studies in Health, Kerman University of Medical Sciences, Kerman, Iran
1 shared publication- Mohammad Hossein Keshavarz
Faculty of Applied Sciences, Malek Ashtar University of Technology, Shahin-shahr, Iran
1 shared publication - Naeimehossadat Asmarian
Anesthesiology and Critical Care Research Center, Shiraz University of Medical Sciences, Shiraz, Iran
1 shared publication Nasiri Mahdi
Health Human Resources Research Center, School of Management & Information Sciences, Shiraz University of Medical Sciences, Shiraz, Iran
1 shared publicationPournik Omid
Department of Community Medicine, School of Medicine, Iran University of Medical Sciences, Tehran, Iran
1 shared publication
Latest Publications
- Open Access
Predictive variables for early detection of sepsis in adults admitted to intensive care units: protocol for a systematic review
Introduction: Sepsis is a life-threatening condition in intensive care units (ICUs), where any delay in diagnosis and treatment can lead to organ dysfunction, prolonged hospital stay and increased mortality. Early identification of sepsis prior to its clinical manifestation may enable timely intervention and improve outcomes. This review aims to identify, synthesise and categorise predictor variables assessed in adult ICU patients prior to sepsis diagnosis. Methods/design: This protocol, in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) guideline, will involve a comprehensive search of PubMed/Medline, Scopus, Embase, Web of Science, IEEE Xplore and Cochrane Library for English-language studies from database inception. The search is planned to be conducted between March 2026 and June 2026. We will include prospective, retrospective, cohort, case–control and cross-sectional studies, and both randomised and non-randomised designs. Additionally, we will consider secondary analysis of electronic health records, clinical registries and routinely acquired clinical data that determine risk factors and predictive variables that will help in the early identification of sepsis using machine learning, artificial intelligence, computational or statistical methods. Two independent reviewers will perform title and abstract screening followed by full-text review and data extraction. The risk of bias will be evaluated using appropriate tools depending on study design: the Newcastle-Ottawa Scale for observational studies, Cochrane Risk of Bias tool for randomised trials and PROBAST for prediction model studies. We will perform a narrative synthesis structured by variable-type, timing and performance metrics (area under the curve, sensitivity, ORs). Subgroup analyses will be conducted based on study design, sepsis definition, type of ICU (medical, surgical, mixed) and geographic area (high-income and low- and middle-income countries). Ethics and dissemination: Findings from this systematic review will be disseminated through publication in a peer-reviewed journal and/or presentation at scientific conferences. The data used will not include individual patient data, so ethical approval is not required. PROSPERO registration number: CRD420251113781.
Open AccessPredictive variables for early detection of sepsis in adults admitted to intensive care units: protocol for a systematic review
Introduction: Sepsis is a life-threatening condition in intensive care units (ICUs), where any delay in diagnosis and treatment can lead to organ dysfunction, prolonged hospital stay and increased mortality. Early identification of sepsis prior to its clinical manifestation may enable timely intervention and improve outcomes. This review aims to identify, synthesise and categorise predictor variables assessed in adult ICU patients prior to sepsis diagnosis. Methods/design: This protocol, in accordance with the Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) guideline, will involve a comprehensive search of PubMed/Medline, Scopus, Embase, Web of Science, IEEE Xplore and Cochrane Library for English-language studies from database inception. The search is planned to be conducted between March 2026 and June 2026. We will include prospective, retrospective, cohort, case–control and cross-sectional studies, and both randomised and non-randomised designs. Additionally, we will consider secondary analysis of electronic health records, clinical registries and routinely acquired clinical data that determine risk factors and predictive variables that will help in the early identification of sepsis using machine learning, artificial intelligence, computational or statistical methods. Two independent reviewers will perform title and abstract screening followed by full-text review and data extraction. The risk of bias will be evaluated using appropriate tools depending on study design: the Newcastle-Ottawa Scale for observational studies, Cochrane Risk of Bias tool for randomised trials and PROBAST for prediction model studies. We will perform a narrative synthesis structured by variable-type, timing and performance metrics (area under the curve, sensitivity, ORs). Subgroup analyses will be conducted based on study design, sepsis definition, type of ICU (medical, surgical, mixed) and geographic area (high-income and low- and middle-income countries). Ethics and dissemination: Findings from this systematic review will be disseminated through publication in a peer-reviewed journal and/or presentation at scientific conferences. The data used will not include individual patient data, so ethical approval is not required. PROSPERO registration number: CRD420251113781.
- Open Access
- 1 January 2020
- Published by IOS Press
- Vol. 272, 387-390
Obstructive Sleep Apnea: A Prediction Model Using Supervised Machine Learning Method
Obstructive Sleep Apnea (OSA) is the most common breathing-related sleep disorder, leading to increased risk of health problems. In this study, we investigated and evaluated the supervised machine learning methods to predict OSA. We used popular machine learning algorithms to develop the prediction models, using a dataset with non-invasive features containing 231 records. Based on the methodology, the CRISP-DM, the dataset was checked and the blanked data were replaced with average/most frequented items. Then, the popular machine learning algorithms were applied for modeling and the 10-fold cross-validation method was used for performance comparison purposes. The dataset has 231 records, of which 152 (65.8%) were diagnosed with OSA. The majority was male (143, 61.9%). The results showed that the best prediction model with an overall AUC reached the Naïve Bayes and Logistic Regression classifier with 0.768 and 0.761, respectively. The SVM with 93.42% sensitivity and the Naïve Bayes of 59.49% specificity can be suitable for screening high-risk people with OSA. The machine learning methods with easily available features had adequate power of discrimination, and physicians can screen high-risk OSA as a supplementary tool.
- 1 January 2020
- Published by IOS Press
- Vol. 272, 387-390
Open AccessObstructive Sleep Apnea: A Prediction Model Using Supervised Machine Learning Method
Obstructive Sleep Apnea (OSA) is the most common breathing-related sleep disorder, leading to increased risk of health problems. In this study, we investigated and evaluated the supervised machine learning methods to predict OSA. We used popular machine learning algorithms to develop the prediction models, using a dataset with non-invasive features containing 231 records. Based on the methodology, the CRISP-DM, the dataset was checked and the blanked data were replaced with average/most frequented items. Then, the popular machine learning algorithms were applied for modeling and the 10-fold cross-validation method was used for performance comparison purposes. The dataset has 231 records, of which 152 (65.8%) were diagnosed with OSA. The majority was male (143, 61.9%). The results showed that the best prediction model with an overall AUC reached the Naïve Bayes and Logistic Regression classifier with 0.768 and 0.761, respectively. The SVM with 93.42% sensitivity and the Naïve Bayes of 59.49% specificity can be suitable for screening high-risk people with OSA. The machine learning methods with easily available features had adequate power of discrimination, and physicians can screen high-risk OSA as a supplementary tool.
- 10 February 2016
- Published by Wiley in Zeitschrift für anorganische und allgemeine Chemie
- Vol. 642 (4) , 335-342
Relation between Electric Spark Sensitivity and Impact Sensitivity of Nitroaromatic Energetic Compounds
Impact and electric spark sensitivities of energetic compounds are two important sensitivity parameters, which are closely related to many accidents in working places. In contrast to electric spark sensitivity, impact sensitivity can be easily measured. A new simple method is introduced to correlate electric spark and impact sensitivities of nitroaromatic compounds. Two correcting functions are used to consider several molecular moieties for reliable prediction of electric spark sensitivity through the measured or estimated impact sensitivity of nitroaromatics. The model is optimized using a set of 28 CHNO polynitroaromatic explosives and then it is tested for some nitroaromatics containing the other atoms such as sulfur. The predicted electric sensitivities of the new method are also compared with the reported results of a new quantum mechanical approach. For 22 CHNO nitroaromatics, quantum mechanical calculations are within ±3.0 J of 18 measured values and more than ±3.0 J for remaining 4 experimental data. Meanwhile, the predicted results of the method are less than ±3.0 J for 28 CHNO nitroaromatics. The root‐mean‐square (rms) deviations of the new model and quantum mechanical are also 1.55 and 2.51 J, respectively.
- 10 February 2016
- Published by Wiley in Zeitschrift für anorganische und allgemeine Chemie
- Vol. 642 (4) , 335-342
Relation between Electric Spark Sensitivity and Impact Sensitivity of Nitroaromatic Energetic Compounds
Impact and electric spark sensitivities of energetic compounds are two important sensitivity parameters, which are closely related to many accidents in working places. In contrast to electric spark sensitivity, impact sensitivity can be easily measured. A new simple method is introduced to correlate electric spark and impact sensitivities of nitroaromatic compounds. Two correcting functions are used to consider several molecular moieties for reliable prediction of electric spark sensitivity through the measured or estimated impact sensitivity of nitroaromatics. The model is optimized using a set of 28 CHNO polynitroaromatic explosives and then it is tested for some nitroaromatics containing the other atoms such as sulfur. The predicted electric sensitivities of the new method are also compared with the reported results of a new quantum mechanical approach. For 22 CHNO nitroaromatics, quantum mechanical calculations are within ±3.0 J of 18 measured values and more than ±3.0 J for remaining 4 experimental data. Meanwhile, the predicted results of the method are less than ±3.0 J for 28 CHNO nitroaromatics. The root‐mean‐square (rms) deviations of the new model and quantum mechanical are also 1.55 and 2.51 J, respectively.