Wisdom_DS: 500 Panoramic Dental Radiographs with Segmentation Masks and Diagnostic Labels

500 high-resolution panoramic dental radiographs (2440 × 1280 px) with pixel-level segmentation masks and structured clinical labels. The dataset, created by Sifat Ishmam Parisa and last updated in April 2026, is designed to facilitate research in dental image analysis and automated diagnosis of third molar teeth. It includes utility scripts for data visualization and manipulation.

Use Cases

• Automated semantic segmentation of third molars based on pixel-level masks.
• Multi-class classification of third molars based on clinical labels like Archer's and Winter's classification.
• Benchmarking deep learning models for pathology assessment and surgical planning based on diagnostic criteria.
• Localization of third molars in panoramic radiographs based on aligned image-mask pairs.

Strengths

• Includes 500 full-HD panoramic radiographs with corresponding pixel-level segmentation masks.
• Provides structured diagnostic labels covering multiple clinical classification systems for third molars.
• Dataset size is 106.6 MB, offering a medium-scale, focused collection for model training.

Limitations

• Row count for the classification label CSV is unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Geographic and demographic source bias inherent to the data collection is not specified.

Provenance

Source

Sifat Ishmam Parisa via figshare

Collection Method

Collection of clinically approved panoramic radiographs and annotations; specific gathering method not detailed.

Freshness

Last updated 2026-04-12 09:47:37; freshness should be verified.