PET/CT Feature Clusters for 289 Fever of Unknown Origin Patients

A retrospective cohort study of 289 adult patients with fever or inflammation of unknown origin who underwent 18F-FDG PET/CT imaging. The dataset includes extracted features on organ involvement and SUVmax values, used for unsupervised clustering to identify imaging phenotypes. The data was authored by Sule Ceylan and last updated on 2026-05-29.

Use Cases

• Benchmark unsupervised clustering algorithms based on PET/CT imaging features like organ involvement and SUVmax.
• Explore relationships between imaging phenotypes and clinical etiologies like infections or rheumatologic diseases.
• Validate the use of consensus clustering approaches for identifying patient subgroups in medical data.

Strengths

• Includes data from 289 patients with a mean age of 59 ± 17 years.
• Features were extracted for multiple organ systems, including reticuloendothelial, parenchymal, and large vessels.
• Analysis employed four unsupervised algorithms and a consensus clustering approach for robust results.
• Results include statistical comparisons with final diagnoses using Adjusted Rand Index and Mutual Information.

Limitations

• The dataset is very small at 14.7 KB, indicating limited scope.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

figshare, author Sule Ceylan.

Collection Method

Retrospective cohort study where data were extracted from patient PET/CT scans.

Freshness

Last updated 2026-05-29 04:50:49; freshness should be verified.