covid 19 image data  (Kaggle Inc)

Journal: Scientific Reports

Article Title: Federated learning with swarm intelligence for efficient and secure medical image analysis

doi: 10.1038/s41598-026-50882-8

Figure Lengend Snippet: Cumulative \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varepsilon$$\end{document} over communication rounds for the COVID-19 dataset with target \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varepsilon = 1.9$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\delta = 10^{-5}$$\end{document} . The RDP accountant tracks multiple \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\alpha$$\end{document} orders and yields the final \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\varepsilon$$\end{document} at round 75.

Article Snippet: This study utilizes publicly available datasets, specifically the COVID-19 image data, Monkeypox image data, and Breast Cancer Wisconsin (Diagnostic) data, all of which were collected and shared according to relevant ethical guidelines and regulations as verified by their respective providers: The COVID-19 dataset was made available by Paul Timothy Mooney on Kaggle for research purposes and has been widely used in the academic community for developing diagnostic algorithms.

Techniques:

Journal: Scientific Reports

Article Title: Federated learning with swarm intelligence for efficient and secure medical image analysis

doi: 10.1038/s41598-026-50882-8

Figure Lengend Snippet: Performance analysis for Hospital Model A (Client A) across COVID-19, Monkeypox, and Breast Cancer datasets. The radar charts display accuracy, precision, recall, specificity, and F1-score metrics for different CNN-optimization combinations.

Article Snippet: This study utilizes publicly available datasets, specifically the COVID-19 image data, Monkeypox image data, and Breast Cancer Wisconsin (Diagnostic) data, all of which were collected and shared according to relevant ethical guidelines and regulations as verified by their respective providers: The COVID-19 dataset was made available by Paul Timothy Mooney on Kaggle for research purposes and has been widely used in the academic community for developing diagnostic algorithms.

Techniques:

Journal: Scientific Reports

Article Title: Federated learning with swarm intelligence for efficient and secure medical image analysis

doi: 10.1038/s41598-026-50882-8

Figure Lengend Snippet: Performance evaluation for Hospital Model B (Client B) showing superior COVID-19 detection capabilities with VGG16 + PSO and exceptional Monkeypox classification using Inception + PSO across all evaluation metrics.

Article Snippet: This study utilizes publicly available datasets, specifically the COVID-19 image data, Monkeypox image data, and Breast Cancer Wisconsin (Diagnostic) data, all of which were collected and shared according to relevant ethical guidelines and regulations as verified by their respective providers: The COVID-19 dataset was made available by Paul Timothy Mooney on Kaggle for research purposes and has been widely used in the academic community for developing diagnostic algorithms.

Techniques:

Journal: Scientific Reports

Article Title: Federated learning with swarm intelligence for efficient and secure medical image analysis

doi: 10.1038/s41598-026-50882-8

Figure Lengend Snippet: Hospital Model C (Client C) performance analysis demonstrating exceptional consistency across all medical datasets, with VGG16 + PSO excelling in COVID-19 detection and Inception + PSO achieving superior Monkeypox classification accuracy.

Article Snippet: This study utilizes publicly available datasets, specifically the COVID-19 image data, Monkeypox image data, and Breast Cancer Wisconsin (Diagnostic) data, all of which were collected and shared according to relevant ethical guidelines and regulations as verified by their respective providers: The COVID-19 dataset was made available by Paul Timothy Mooney on Kaggle for research purposes and has been widely used in the academic community for developing diagnostic algorithms.

Techniques:

Journal: Scientific Reports

Article Title: Federated learning with swarm intelligence for efficient and secure medical image analysis

doi: 10.1038/s41598-026-50882-8

Figure Lengend Snippet: Hospital Model D (Client D) performance evaluation showing exceptional COVID-19 detection with VGG16 + PSO achieving 96.71% accuracy and superior Monkeypox classification using Inception + PSO with balanced precision and recall metrics.

Article Snippet: This study utilizes publicly available datasets, specifically the COVID-19 image data, Monkeypox image data, and Breast Cancer Wisconsin (Diagnostic) data, all of which were collected and shared according to relevant ethical guidelines and regulations as verified by their respective providers: The COVID-19 dataset was made available by Paul Timothy Mooney on Kaggle for research purposes and has been widely used in the academic community for developing diagnostic algorithms.

Techniques:

Journal: Frontiers in Digital Health

Article Title: Guideline-based strategies to identify severe cytokine release syndrome in COVID-19 and cancer immunotherapy using large-scale electronic health records

doi: 10.3389/fdgth.2025.1625889

Figure Lengend Snippet: Definition of the cohorts of interests, inclusion and exclusion criteria. From the Optum® COVID-19 data, subjects with active COVID-19 or who received TCE treatment (blinatumomab) were isolated. The first COVID-19 infection or TCE treatment was considered a triggering event. Subjects with missing information on gender or age were excluded. Subjects with pre-existing comorbidities that share traits with CRS at least 7 days before the triggering event were excluded. A significant number of COVID-19 patients had no reported data 30 days around the triggering event and were excluded. Patients diagnosed with Sepsis within 30 days after onset were also excluded. Three cohorts of interest were used for subsequent CRS case identification: ‘COVID-19 adult cohort’, ‘TCE adult cohort’ and ‘TCE pediatric cohort’.

Article Snippet: Using the Optum® de-identified COVID-19 Electronic Health Record data set (‘Optum® COVID-19 data’), we compared the occurrence of severe CRS following those three case definition strategies on 2.5 million patients with COVID-19 infection and 171 patients treated with the TCE blinatumomab.

Techniques: Isolation, Infection

Journal: Frontiers in Digital Health

Article Title: Guideline-based strategies to identify severe cytokine release syndrome in COVID-19 and cancer immunotherapy using large-scale electronic health records

doi: 10.3389/fdgth.2025.1625889

Figure Lengend Snippet: CRS grading algorithm (decision tree) on EHR datasets following the ASTCT grading guideline. (A) The latest and commonly used ASTCT CRS grading , keeping identical wording as in the original publication. (B) From the list of reported patient features within a time window of 30 days after the triggering event (TCE administration or COVID-19 diagnosis), patients are first graded into ‘grade N+’ and then separated into definite grades: grade 1+ includes patients with fever ≥38 °C (‘strict’ definition) or those with potentially mitigated fever by corticosteroid or cytokine blocker (anti-IL1 or anti-IL6) therapy (‘mitigated’ definition). Grade 2+ to 4+ are defined based on the grade-defining interventions: grade 4+: CPAP or invasive ventilation or use of multiple vasopressors; grade 3+ (one vasopressor or non-CPAP ventilation); and grade 2+ (evidence for hypoxia or hypotension). Notably, we assumed that the use of vasopressors or ventilation indicated hypoxia or hypotension, even if the reported cardiovascular or respiratory parameters were within the reference range. Patients without grade 2+ were classified as “definite” grades if lab values were in range, “probable” grades if hypoxia or hypotension were not measured, or as non-classifiable. We proposed a definition for CRS grade 2+ (or grade 3+) positive and negative (i.e., control) cohorts. SaO2 = arterial oxygen saturation, SBP = systolic blood pressure SpO2 = peripheral oxygen saturation, PaO2 = partial arterial oxygen pressure, PvO2 = venous oxygen tension, and DBP = diastolic blood pressure.

Article Snippet: Using the Optum® de-identified COVID-19 Electronic Health Record data set (‘Optum® COVID-19 data’), we compared the occurrence of severe CRS following those three case definition strategies on 2.5 million patients with COVID-19 infection and 171 patients treated with the TCE blinatumomab.

Techniques: Biomarker Discovery, Control

Journal: Frontiers in Digital Health

Article Title: Guideline-based strategies to identify severe cytokine release syndrome in COVID-19 and cancer immunotherapy using large-scale electronic health records

doi: 10.3389/fdgth.2025.1625889

Figure Lengend Snippet: Identified patients following different implementations of the consensus ASTCT CRS grading on the COVID-19 and TCE cohorts. (A) Grading into N + groups, from grade 1+ to grade 4+, depending on the implementations: strict, extended, extended+mitigations. (B) Breakdown of the cohorts by grades based on the ‘extended+mitigations’ implementation, including details of definite and probable grading, as well as ambiguous patients who show symptoms of a higher grade but do not qualify for grade 1, and deceased patients with (probable grade 5) or without (not gradable) grade 2+ features. Notably, sepsis patients have been excluded, and the cohort includes patients who could be CRS positive or negative. Therefore, the percentages are calculated within the “usable cohort” of patients who did not experience sepsis, but these numbers should be adjusted to include the entire cohort, including sepsis cases, if prevalence needs to be determined.

Article Snippet: Using the Optum® de-identified COVID-19 Electronic Health Record data set (‘Optum® COVID-19 data’), we compared the occurrence of severe CRS following those three case definition strategies on 2.5 million patients with COVID-19 infection and 171 patients treated with the TCE blinatumomab.

Techniques: