PET/CT Features for 289 Fever of Unknown Origin Patients

Sule Ceylan published a dataset on 2026-05-29 containing results from a retrospective cohort study of 289 adult patients with fever of unknown origin (FUO). The data includes PET/CT imaging features such as organ involvement, SUVmax values, and uptake patterns, which were used for unsupervised clustering analysis. The study aimed to classify patients based on imaging phenotypes and explore their relationship with final clinical diagnoses.

Use Cases

• Benchmark unsupervised clustering algorithms based on multi-feature PET/CT data.
• Analyze imaging phenotype patterns based on reticuloendothelial, parenchymal, and large vessel involvement.
• Investigate the relationship between imaging clusters and clinical etiologies like infections or rheumatologic diseases.
• Validate consensus clustering approaches for medical data integration.

Strengths

• Includes data from 289 patients with a defined mean age of 59 ± 17 years and gender distribution (44.3% women, 55.7% men).
• Features were extracted using a structured protocol covering multiple organ systems and quantitative SUVmax values.
• Analysis employed four distinct unsupervised algorithms and a consensus clustering approach for robust results.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the underlying feature table is unknown, which may limit suitability assessment.
• The dataset is very small at 14.9 KB, indicating limited scope, likely containing summary results rather than raw patient-level data.

Provenance

Source

figshare, author Sule Ceylan.

Collection Method

Retrospective cohort study of adult FUO patients who underwent 18F-FDG PET/CT.

Freshness

Last updated 2026-05-29 05:11:50; freshness should be verified.