Brain Tumor Segmentation Metrics from BraTS2023 Validation

A 5.5 KB Excel file containing metric values for brain tumor segmentation from a validation set of 188 MRI scans. The dataset was created by Muthulakshmi Kirubakaran and last updated on 2026-05-18. It corresponds to a study proposing a GA-ResUNetGAN model for segmenting tumor core, whole tumor, and enhancing tumor subregions.

Use Cases

• Benchmarking segmentation model performance based on reported Dice scores for tumor subregions.
• Analyzing the effectiveness of hybrid architectures like GA-ResUNetGAN mentioned in the description.
• Comparing segmentation quality and boundary delineation against baseline methods such as 3D UNet and SwinUNETR.

Strengths

• Validation set size of 188 scans is explicitly stated.
• Reported Dice scores (0.94, 0.96, 0.91) for tumor subregions provide concrete performance metrics.
• Dataset is associated with the BraTS2023 challenge, a known benchmark in the field.

Limitations

• Row count is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• The 5.5 KB file size suggests a very limited scope, likely containing only summary metrics, not raw image data.

Provenance

Source

BraTS2023 challenge dataset.

Collection Method

Likely derived from model validation outputs of a proposed GA-ResUNetGAN architecture.

Freshness

Last updated 2026-05-18 17:47:15; freshness should be verified.