ChestXray14 Preprocessed: NIH Chest X-Rays with CLAHE Enhancement

Preprocessed cache of the NIH ChestX-ray14 dataset, a large collection of chest radiographs. The data has been processed with Contrast Limited Adaptive Histogram Equalization (CLAHE) and resized to a uniform 448-pixel resolution. Patient-disjoint splits are provided to prevent data leakage in machine learning experiments.

Use Cases

• Train image classification models based on the 14 labeled thoracic pathologies.
• Benchmark model performance using the provided patient-disjoint data splits.
• Develop computer-aided diagnosis tools based on preprocessed, standardized X-ray images.
• Experiment with transfer learning using a large, publicly available medical imaging dataset.

Strengths

• Data is preprocessed with CLAHE and resized to 448px, providing a consistent input format for models.
• Patient-disjoint splits are included, a critical feature for preventing data leakage in medical ML.

Limitations

• Description metadata is limited; actual data quality, row count, and file formats require manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.

Provenance

Source

NIH ChestX-ray14 dataset

Collection Method

Preprocessing (CLAHE, resizing) and splitting of the original NIH data.

Time Range

null

Freshness

Last update date is unknown; freshness unverified.

Geography

null