FZFan Zhang

Fan Zhang

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Fan Zhang has a total of 8 co-authors. Here are the top co-authors by number of shared publications:

  • Fuli Wu

    Zhejiang University of Technology, Hangzhou, Zhejiang, China

    2 shared publications
  • Fudong Zhu

    Stomatology Hospital, School of Stomatology, Zhejiang University School of Medicine, Zhejiang Provincial Clinical Research Center for Oral Diseases, Key Laboratory of Oral Biomedical Research of Zhejiang Province Cancer Center of Zhejiang University, Engineering Research Center of Oral Biomaterials and Devices of Zhejiang Province Hangzhou China

    1 shared publication
  • Haihua Zhu

    Stomatology Hospital, School of Stomatology, Zhejiang University School of Medicine, Zhejiang Provincial Clinical Research Center for Oral Diseases, Zhejiang Key Laboratory of Oral Biomedical, Engineering Research Center of Oral Biomaterials and Devices of Zhejiang Province, Hangzhou, China

    1 shared publication
  • Haojie Yu

    Key Laboratory of Artificial Organs and Computational Medicine of Zhejiang Province, Key Laboratory of Pollution Exposure and Health Intervention of Zhejiang Province Shulan International Medical College, Zhejiang Shuren University Hangzhou China

    1 shared publication
  • Junhua Zhu

    School/Hospital of Stomatology, Zhejiang Chinese Medical University Hangzhou China

    1 shared publication
  • Pengyi Hao

    Zhejiang University of Technology, Hangzhou, China

    1 shared publication

Latest Publications

  • Automatic segmentation and detection of ectopic eruption of first permanent molars on panoramic radiographs based on nnU‐Net

    by Haihua Zhu, Haojie Yu, Fan Zhang, Zheng Cao, Fuli Wu, Fudong Zhu

    Aim: The purpose of this research was to present an artificial intelligence (AI) model, which can automatically segment and detect ectopic eruption of first permanent molars (EMMs) in early mixed dentition on panoramic radiographs using the no‐new‐Net (nnU‐Net) model.Design: A total of 438 EMMs obtained from 285 panoramic radiographs were included in this study. An AI model based on nnU‐Net was trained to segment and detect EMMs. The performance of the model was evaluated by the intersection over union (IoU), precision, F1‐score, accuracy and FROC. Furthermore, the detecting performance of nnU‐Net was compared with that of three dentists with different years of experience using the McNemar chi‐squared test. The reliability of different dentists was evaluated by intraclass correlation coefficients (ICCs).Results: The nnU‐Net yielded an IoU of 0.834, a precision of 0.845, an F1‐score of 0.902 and an accuracy of 0.990, whereas the dentists yielded a mean IoU of 0.530, a mean precision of 0.539, a mean F1‐score of 0.699 and a mean accuracy of 0.811. The ICC of different dentists was 0.776. The statistical analysis of the McNemar chi‐squared test showed that the nnU‐Net results were statistically significant and superior to those of dentists (p < .05).Conclusion: This study validated an AI model based on nnU‐Net for automatically segmenting and detecting EMMs more consistently and accurately on panoramic radiography.


  • BDU‐net: Toward accurate segmentation of dental image using border guidance and feature map distortion

    by Fan Zhang, Junhua Zhu, Pengyi Hao, Fuli Wu, Yuanna Zheng

    The importance of instance segmentation for each tooth is increasing in dental disease diagnosis and computer‐assisted treatment. However, most existing segmentation methods are mainly concerned with semantic feature extraction, while ignoring complex situations such as blurred boundaries and dislocation of teeth in panoramic radiographs. To address these problems, we propose a dual subnetworks structure based on border guidance and feature map distortion, called BDU‐net. Specifically, we embed the Disout method with the capability to distort features into the encoder to obtain multi‐scale feature maps with excellent generalization. Then, the distorted feature map is submitted to two subnetworks, each having a feature pyramid structure in a shared manner, including a border subnetwork that adjusts the segmentation boundaries and a region subnetwork that generates the region segmentation results. Experiments conducted on the teeth segmentation dataset show that the proposed BDU‐net performs better than the other approaches, especially for the case of teeth missing or dislocated teeth.