Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PRAD: Periapical Radiograph Analysis Dataset and Benchmark Model Development

Zhenhuan Zhou, Yuchen Zhang, Ruihong Xu, Xuansen Zhao, Tao Li

TL;DR

The paper tackles the limited availability of high-quality periapical radiograph data for deep learning by introducing PRAD-10K, a 10,000-image PR dataset with pixel-level annotations across nine categories and disease labels. It also proposes PRNet, a segmentation network that fuses a Multi-scale Wavelet Convolution Network (MWCN) encoder with a Channel Fusion Attention (CFA) mechanism to address PR images' multi-scale challenges. PRNet achieves state-of-the-art performance on PRAD-10K with an average DSC of 84.24%, outperforming recent medical segmentation models and demonstrating robust detection of small structures and lesions. The dataset and benchmarking code are released to accelerate DL-based PR analysis in endodontics, highlighting the value of combining multi-scale wavelet features with cross-scale channel attention for precise PR segmentation and anatomy/lesion identification. The work sets a practical foundation for future expansion to fully supervised, semi-supervised, or multimodal PR analysis in dentistry.

Abstract

Deep learning (DL), a pivotal technology in artificial intelligence, has recently gained substantial traction in the domain of dental auxiliary diagnosis. However, its application has predominantly been confined to imaging modalities such as panoramic radiographs and Cone Beam Computed Tomography, with limited focus on auxiliary analysis specifically targeting Periapical Radiographs (PR). PR are the most extensively utilized imaging modality in endodontics and periodontics due to their capability to capture detailed local lesions at a low cost. Nevertheless, challenges such as resolution limitations and artifacts complicate the annotation and recognition of PR, leading to a scarcity of publicly available, large-scale, high-quality PR analysis datasets. This scarcity has somewhat impeded the advancement of DL applications in PR analysis. In this paper, we present PRAD-10K, a dataset for PR analysis. PRAD-10K comprises 10,000 clinical periapical radiograph images, with pixel-level annotations provided by professional dentists for nine distinct anatomical structures, lesions, and artificial restorations or medical devices, We also include classification labels for images with typical conditions or lesions. Furthermore, we introduce a DL network named PRNet to establish benchmarks for PR segmentation tasks. Experimental results demonstrate that PRNet surpasses previous state-of-the-art medical image segmentation models on the PRAD-10K dataset. The codes and dataset will be made publicly available.

PRAD: Periapical Radiograph Analysis Dataset and Benchmark Model Development

TL;DR

The paper tackles the limited availability of high-quality periapical radiograph data for deep learning by introducing PRAD-10K, a 10,000-image PR dataset with pixel-level annotations across nine categories and disease labels. It also proposes PRNet, a segmentation network that fuses a Multi-scale Wavelet Convolution Network (MWCN) encoder with a Channel Fusion Attention (CFA) mechanism to address PR images' multi-scale challenges. PRNet achieves state-of-the-art performance on PRAD-10K with an average DSC of 84.24%, outperforming recent medical segmentation models and demonstrating robust detection of small structures and lesions. The dataset and benchmarking code are released to accelerate DL-based PR analysis in endodontics, highlighting the value of combining multi-scale wavelet features with cross-scale channel attention for precise PR segmentation and anatomy/lesion identification. The work sets a practical foundation for future expansion to fully supervised, semi-supervised, or multimodal PR analysis in dentistry.

Abstract

Deep learning (DL), a pivotal technology in artificial intelligence, has recently gained substantial traction in the domain of dental auxiliary diagnosis. However, its application has predominantly been confined to imaging modalities such as panoramic radiographs and Cone Beam Computed Tomography, with limited focus on auxiliary analysis specifically targeting Periapical Radiographs (PR). PR are the most extensively utilized imaging modality in endodontics and periodontics due to their capability to capture detailed local lesions at a low cost. Nevertheless, challenges such as resolution limitations and artifacts complicate the annotation and recognition of PR, leading to a scarcity of publicly available, large-scale, high-quality PR analysis datasets. This scarcity has somewhat impeded the advancement of DL applications in PR analysis. In this paper, we present PRAD-10K, a dataset for PR analysis. PRAD-10K comprises 10,000 clinical periapical radiograph images, with pixel-level annotations provided by professional dentists for nine distinct anatomical structures, lesions, and artificial restorations or medical devices, We also include classification labels for images with typical conditions or lesions. Furthermore, we introduce a DL network named PRNet to establish benchmarks for PR segmentation tasks. Experimental results demonstrate that PRNet surpasses previous state-of-the-art medical image segmentation models on the PRAD-10K dataset. The codes and dataset will be made publicly available.

Paper Structure

This paper contains 13 sections, 3 equations, 3 figures, 3 tables.

Table of Contents

  1. Introduction
  2. PRAD-10K
  3. Overview
  4. Collection and Annotation
  5. Method
  6. Overview
  7. Multi-scale Wavelet Convolution Network
  8. Channel Fusion Attention
  9. Experiments
  10. Implementation Details
  11. Comparisons with Other Methods
  12. Ablation experiments
  13. Conclusion

Figures (3)

  • Figure 1: Examples of three dental radiographic imaging modalities. (a) PAN, (b) CBCT and (c) PR.
  • Figure 2: Examples from the PRAD-10K dataset (top left) along with their corresponding pixel-level annotations (bottom left). The right side displays the label distribution.
  • Figure 3: Overall framework of the proposed PRNet. The structure of the decoder blocks are consistent with UNet.