Thermal Images and Features for Joint Inflammation Detection Across 14 Anatomical Regions

A dataset supporting research on texture-enhanced thermography for joint inflammation detection. It contains 2,754 thermal images in CSV format across 14 anatomical regions, along with patient metadata and extracted statistical features. The data was collected by researchers including Dima Bykhovsky and published in 2026.

Use Cases

• Training machine learning models for inflammation detection based on first-order temperature statistics.
• Developing image analysis algorithms for medical thermography based on texture descriptors like entropy and kurtosis.
• Validating consumer-grade thermal cameras for clinical use based on MRI-validated patient sub-cohorts.
• Reproducing statistical analyses from the associated 2026 research paper using the provided feature tables.

Strengths

• Contains 2,754 thermal images across 14 anatomical regions.
• Includes 17 feature tables with first-order statistics and texture descriptors.
• Provides patient metadata including age, sex, clinical group, and environmental conditions.
• Data is de-identified and ethics-approved by Meir Medical Center's Institutional Review Board.

Limitations

• Row count for patient and feature tables is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred from the README after download.

Provenance

Source

figshare deposit by Dima Bykhovsky.

Collection Method

Collected as part of a clinical study using consumer-grade thermal cameras.

Freshness

Last updated 2026-05-27 11:45:47.

Geography

Data collection was approved by Meir Medical Center in Israel.