Shijie Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Shijie Zhang has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
Bosha Jia
Department of Orthodontics, Tianjin Medical University School and Hospital of Stomatological & Tianjin Key Laboratory of Oral Soft and Hard Tissues Restoration and Regeneration, Tianjin, P. R. China
1 shared publicationCong Lin
School of Intelligent Systems Science and Engineering, Jinan University, Zhuhai, China
1 shared publicationFang Zheng
School of Medical Laboratory, Tianjin Medical University, Tianjin, China
1 shared publicationHanshan Zhang
The Australian National University, Canberra, Australia
1 shared publicationJianing Li
Department of Orthodontics, Tianjin Medical University School and Hospital of Stomatological & Tianjin Key Laboratory of Oral Soft and Hard Tissues Restoration and Regeneration, Tianjin, P. R. China
1 shared publicationLei Li
Clinical Laboratory, Tianjin Chest Hospital, Tianjin, China
1 shared publication
Latest Publications
- Open Access
- 26 February 2026
- Published by Springer Nature in Scientific Data
- Vol. 13 (1)
A dataset of midpalatal suture maturation stage in cone-beam computed tomography
In the field of orthodontics, the degree of fusion of the midpalatal suture (MPS) is a crucial factor in determining the most appropriate maxillary expansion technique. This study analyzed 600 cone-beam computed tomography (CBCT) scans from patients aged 4 to 25, covering all MPS maturity stages (35 in stage A, 113 in stage B, 186 in stage C, 163 in stage D, and 103 in stage E). A custom model integrated 3D convolutional neural networks (3D CNNs) for processing image data and fully connected networks for handling tabular data to classify MPS stages. Key factors including gender, age, cervical vertebra staging, dental age, palatal morphology, and MPS bone density ratio were evaluated for their correlation with MPS fusion. Ethical approval was secured, and images were quality-assessed by orthodontists and a maxillofacial imaging physician. The model shows great performance in the diagnosis with AUC over 0.95. The model aids in selecting appropriate expanders, favoring traditional rapid palatal expansion (RME) for unclosed MPS and bone-anchored RME or surgery for fused cases.
- 26 February 2026
- Published by Springer Nature in Scientific Data
- Vol. 13 (1)
Open AccessA dataset of midpalatal suture maturation stage in cone-beam computed tomography
In the field of orthodontics, the degree of fusion of the midpalatal suture (MPS) is a crucial factor in determining the most appropriate maxillary expansion technique. This study analyzed 600 cone-beam computed tomography (CBCT) scans from patients aged 4 to 25, covering all MPS maturity stages (35 in stage A, 113 in stage B, 186 in stage C, 163 in stage D, and 103 in stage E). A custom model integrated 3D convolutional neural networks (3D CNNs) for processing image data and fully connected networks for handling tabular data to classify MPS stages. Key factors including gender, age, cervical vertebra staging, dental age, palatal morphology, and MPS bone density ratio were evaluated for their correlation with MPS fusion. Ethical approval was secured, and images were quality-assessed by orthodontists and a maxillofacial imaging physician. The model shows great performance in the diagnosis with AUC over 0.95. The model aids in selecting appropriate expanders, favoring traditional rapid palatal expansion (RME) for unclosed MPS and bone-anchored RME or surgery for fused cases.
- Open Access
- 7 February 2023
- Published by Springer Nature in Scientific Reports
- Vol. 13 (1) , 1-12
A deep learning model for detection of leukocytes under various interference factors
The accurate detection of leukocytes is the basis for the diagnosis of blood system diseases. However, diagnosing leukocyte disorders by doctors is time-consuming and requires extensive experience. Automated detection methods with high accuracy can improve detection efficiency and provide recommendations to inexperienced doctors. Current methods and instruments either fail to automate the identification process fully or have low performance and need suitable leukocyte data sets for further study. To improve the current status, we need to develop more intelligent strategies. This paper investigates fulfilling high-performance automatic detection for leukocytes using a deep learning-based method. We established a new dataset more suitable for leukocyte detection, containing 6273 images (8595 leukocytes) and considering nine common clinical interference factors. Based on the dataset, the performance evaluation of six mainstream detection models is carried out, and a more robust ensemble model is proposed. The mean of average precision (mAP) @IoU = 0.50:0.95 and mean of average recall (mAR)@IoU = 0.50:0.95 of the ensemble model on the test set are 0.853 and 0.922, respectively. The detection performance of poor-quality images is robust. For the first time, it is found that the ensemble model yields an accuracy of 98.84% for detecting incomplete leukocytes. In addition, we also compared the test results of different models and found multiple identical false detections of the models, then provided correct suggestions for the clinic.
- 7 February 2023
- Published by Springer Nature in Scientific Reports
- Vol. 13 (1) , 1-12
Open AccessA deep learning model for detection of leukocytes under various interference factors
The accurate detection of leukocytes is the basis for the diagnosis of blood system diseases. However, diagnosing leukocyte disorders by doctors is time-consuming and requires extensive experience. Automated detection methods with high accuracy can improve detection efficiency and provide recommendations to inexperienced doctors. Current methods and instruments either fail to automate the identification process fully or have low performance and need suitable leukocyte data sets for further study. To improve the current status, we need to develop more intelligent strategies. This paper investigates fulfilling high-performance automatic detection for leukocytes using a deep learning-based method. We established a new dataset more suitable for leukocyte detection, containing 6273 images (8595 leukocytes) and considering nine common clinical interference factors. Based on the dataset, the performance evaluation of six mainstream detection models is carried out, and a more robust ensemble model is proposed. The mean of average precision (mAP) @IoU = 0.50:0.95 and mean of average recall (mAR)@IoU = 0.50:0.95 of the ensemble model on the test set are 0.853 and 0.922, respectively. The detection performance of poor-quality images is robust. For the first time, it is found that the ensemble model yields an accuracy of 98.84% for detecting incomplete leukocytes. In addition, we also compared the test results of different models and found multiple identical false detections of the models, then provided correct suggestions for the clinic.