MOPG-7: A Multi-Clinic Dental Panoramic Radiograph Dataset with Expert YOLO Labels

Description

MOPG-7 is a publicly available, multi-clinic dental imaging database that aims at furthering studies on artificial intelligence (AI), computer vision, and computer-aided diagnosis (CAD) based on panoramic dental radiographs (orthopantomogram or OPG). This database includes 2,095 completely anonymized panoramic dental radiographs gathered retrospectively from four separate dental centers in Bangladesh such as Sonia Nursing Home, Tangail (1,539 radiographs); Ibn Sina Diagnostic & Consultation Center, Dhaka (533 radiographs); Niramoy Diagnostic Center, Tangail (114 radiographs); and Health City Diagnostic Center, Gaibandha (15 radiographs). Every image comes with a bounding box annotation file in YOLO format (.txt) that has been validated by experts, making it easier for use with modern object detection models like Ultralytics YOLO. The initial bounding boxes were created by a licensed dentist and reviewed independently by another experienced dental professional. A rigorous process of quality control followed after that to remove 127 bounding boxes that were not up to the mark, resulting in 9,834 bounding boxes being used. Dataset Classes The dataset includes annotations for seven clinically relevant dental categories: Missing Teeth: 2,609 annotations Dental Crown: 1,984 annotations Root Canal: 1,956 annotations Caries: 1,410 annotations Wisdom Teeth: 869 annotations Broken Down Teeth: 795 annotations Healthy Teeth: 211 annotations Dataset Contents The released dataset includes: Panoramic dental radiographs (.png) YOLO bounding-box annotation files (.txt) Class definition file (classes.txt) Documentation (README.md) describing the dataset structure, annotation format, and usage instructions Potential Research Applications MOPG-7 was built to enable a broad range of uses for research and learning purposes such as multi-class dental object detection, localization of dental abnormalities, CAD, deep learning for medical imaging, computer vision research, transfer learning and building foundation models, XAI, medical image analysis, object detection algorithm benchmarking, AI-enabled dental diagnosis, dental AI learning, and reproducibility research. Benchmark Performance In order to have an effective baseline, YOLOv11m was trained and tested on MOPG-7 to give a precision score of 90.2%, recall score of 91.5%, mAP@0.5 of 93.0%, and mAP@0.5:0.95 of 59.2%. Of the seven classes, the Wisdom Teeth and Dental Crown were those that performed best, while Missing Teeth and Caries posed more difficulties because of anatomical differences.

Institutions

• Daffodil International University

  Dhaka Division, Dhaka

Categories

Licence