Fang Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Fang Li has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Xinrong Chen
Fudan University
2 shared publications - Bin Jiang
China University of Petroleum (East China)
1 shared publication Chaolei He
Shanghai University of Electric Power
1 shared publication- Chenggang Yan
Hangzhou Dianzi University, Hangzhou, China
1 shared publication Chenglong Song
Shanghai University of Electric Power
1 shared publicationChufan Xiao
1 shared publication
Latest Publications
- 23 December 2025
- Published by Springer Nature in Multimedia Systems
- Vol. 32 (1) , 53
A hybrid transformer architecture with Mamba for medical image report generation
- 23 December 2025
- Published by Springer Nature in Multimedia Systems
- Vol. 32 (1) , 53
A hybrid transformer architecture with Mamba for medical image report generation
- Open Access
- 30 September 2025
- Published by Journal of Internet Technology in 網際網路技術學刊
- Vol. 26 (5) , 607-616
Diagnosis of Fetal Congenital Heart Disease Based on Deep Learning
Fetal congenital heart disease is the most common birth defect. In early diagnosis, the use of echocardiography is an important means of diagnosis, but due to the unique structure of the fetal heart, there are still many challenges in the early screening process. Hence, this study proposes a diagnosis model called ConvNeXt based on Attention Mechanism and Transfer Learning (ConvNeXt-AMTL) for congenital heart disease, which utilizes a large-kernel convolutional neural network to extract features from fetal echocardiography, and using attention mechanisms to focus and optimize key features. At the same time, in order to alleviate the problem that image data samples are too few to train the model well, this study proposes to use transfer learning to train the model. Numerous experiments have shown that the proposed model can efficiently diagnose fetal congenital heart disease, achieving an accuracy of 98.8% on the test set, effectively promoting prenatal screening of fetal congenital heart disease.
- 30 September 2025
- Published by Journal of Internet Technology in 網際網路技術學刊
- Vol. 26 (5) , 607-616
Open AccessDiagnosis of Fetal Congenital Heart Disease Based on Deep Learning
Fetal congenital heart disease is the most common birth defect. In early diagnosis, the use of echocardiography is an important means of diagnosis, but due to the unique structure of the fetal heart, there are still many challenges in the early screening process. Hence, this study proposes a diagnosis model called ConvNeXt based on Attention Mechanism and Transfer Learning (ConvNeXt-AMTL) for congenital heart disease, which utilizes a large-kernel convolutional neural network to extract features from fetal echocardiography, and using attention mechanisms to focus and optimize key features. At the same time, in order to alleviate the problem that image data samples are too few to train the model well, this study proposes to use transfer learning to train the model. Numerous experiments have shown that the proposed model can efficiently diagnose fetal congenital heart disease, achieving an accuracy of 98.8% on the test set, effectively promoting prenatal screening of fetal congenital heart disease.
- 31 October 2024
- Published by Springer Nature
- p. 316-324
Fetal Congenital Heart Disease Diagnosis Based on CBAM-Enhanced ResNet-50
- 31 October 2024
- Published by Springer Nature
- p. 316-324
Fetal Congenital Heart Disease Diagnosis Based on CBAM-Enhanced ResNet-50
- 7 March 2024
- Published by World Scientific Pub Co Pte Ltd in Journal of Mechanics in Medicine and Biology
- Vol. 24 (02)
INTRAOPERATIVE SURGICAL NAVIGATION BASED ON LASER SCANNER FOR IMAGE-GUIDED ORAL AND MAXILLOFACIAL SURGERY
In oral and maxillofacial surgery, computer-assisted navigation technologies have been widely used to achieve intraoperative positioning. The traditional methods mainly rely on the experience of doctors and the difference between the locations of key points in the surgical area and the preoperative planning, which have certain limitations. In this paper, a new intraoperative surgical navigation framework based on mobile laser scanner is proposed, which ensures that surgery is performed accurately according to the preoperative planning. The framework mainly includes two parts. First, the real-time surface reconstruction of the anatomy should be realized during the operation. Second, the acquired image is matched to the planned image in real time. Although the most common method of surface reconstruction is to render the volume directly from raw data or render the surface from the segmented data using computed tomography/magnetic resonance (CT/MR) data, this method is too complicated for performing the real-time operation during surgery. Furthermore, a new surface registration technique is proposed for image-guided oral and maxillofacial surgery based on the point sets. To improve the registration accuracy and robustness, the point sets are modeled by Mixed Student’s t-Distribution model. In the experiments, the point sets of CT data are from 10 patients with craniomaxillofacial diseases and the surface point set is from the LRS. The TRE of 10 data was less than 1[ ]mm. Compared with the paired-point registration method and Iterative Closest Point algorithm, the results demonstrated better performance of the proposed method, the surgical situation can be displayed in real time during the surgical process, and any differences from the surgical plan can also be reflected.
- 7 March 2024
- Published by World Scientific Pub Co Pte Ltd in Journal of Mechanics in Medicine and Biology
- Vol. 24 (02)
INTRAOPERATIVE SURGICAL NAVIGATION BASED ON LASER SCANNER FOR IMAGE-GUIDED ORAL AND MAXILLOFACIAL SURGERY
In oral and maxillofacial surgery, computer-assisted navigation technologies have been widely used to achieve intraoperative positioning. The traditional methods mainly rely on the experience of doctors and the difference between the locations of key points in the surgical area and the preoperative planning, which have certain limitations. In this paper, a new intraoperative surgical navigation framework based on mobile laser scanner is proposed, which ensures that surgery is performed accurately according to the preoperative planning. The framework mainly includes two parts. First, the real-time surface reconstruction of the anatomy should be realized during the operation. Second, the acquired image is matched to the planned image in real time. Although the most common method of surface reconstruction is to render the volume directly from raw data or render the surface from the segmented data using computed tomography/magnetic resonance (CT/MR) data, this method is too complicated for performing the real-time operation during surgery. Furthermore, a new surface registration technique is proposed for image-guided oral and maxillofacial surgery based on the point sets. To improve the registration accuracy and robustness, the point sets are modeled by Mixed Student’s t-Distribution model. In the experiments, the point sets of CT data are from 10 patients with craniomaxillofacial diseases and the surface point set is from the LRS. The TRE of 10 data was less than 1[ ]mm. Compared with the paired-point registration method and Iterative Closest Point algorithm, the results demonstrated better performance of the proposed method, the surgical situation can be displayed in real time during the surgical process, and any differences from the surgical plan can also be reflected.
- 7 October 2023
- Published by World Scientific Pub Co Pte Ltd in Journal of Mechanics in Medicine and Biology
- Vol. 23 (09)
ISC-TRANSUNET: MEDICAL IMAGE SEGMENTATION NETWORK BASED ON THE INTEGRATION OF SELF-ATTENTION AND CONVOLUTION
In the current medical image segmentation network, the combination of CNN and Transformer has become a mainstream trend. However, the inherent limitations of convolution operation in CNN and insufficient information interaction in Transformer affect the segmentation performance of the network. To solve these problems, an integrated self-attention and convolution medical image segmentation network (ISC-TransUNet) is proposed in this paper. The network consists of encoder, decoder and jump connection. First, the encoder uses a hybrid structure of BoTNet and Transformer to capture more comprehensive image information and reduce additional computing overhead. Then, the decoder uses an upper sampler cascaded by multiple DUpsampling upper blocks to accurately recover the pixel-level prediction. Finally, the feature fusion of encoder and decoder at different resolutions is realized by ResPath jump connection, which reduces the semantic difference between encoder and decoder. Through experiments on the Synapse multi-organ segmentation dataset, compared with the baseline model TransUNet, Dice similarity coefficient of ISC-TransUNet was improved by 1.13%, Hausdorff distance was reduced by 2.38%, and weight was maintained. The experimental results show that the network can effectively segment tissues and organs in medical images, which has important theoretical significance and application value for intelligent clinical diagnosis and treatment.
- 7 October 2023
- Published by World Scientific Pub Co Pte Ltd in Journal of Mechanics in Medicine and Biology
- Vol. 23 (09)
ISC-TRANSUNET: MEDICAL IMAGE SEGMENTATION NETWORK BASED ON THE INTEGRATION OF SELF-ATTENTION AND CONVOLUTION
In the current medical image segmentation network, the combination of CNN and Transformer has become a mainstream trend. However, the inherent limitations of convolution operation in CNN and insufficient information interaction in Transformer affect the segmentation performance of the network. To solve these problems, an integrated self-attention and convolution medical image segmentation network (ISC-TransUNet) is proposed in this paper. The network consists of encoder, decoder and jump connection. First, the encoder uses a hybrid structure of BoTNet and Transformer to capture more comprehensive image information and reduce additional computing overhead. Then, the decoder uses an upper sampler cascaded by multiple DUpsampling upper blocks to accurately recover the pixel-level prediction. Finally, the feature fusion of encoder and decoder at different resolutions is realized by ResPath jump connection, which reduces the semantic difference between encoder and decoder. Through experiments on the Synapse multi-organ segmentation dataset, compared with the baseline model TransUNet, Dice similarity coefficient of ISC-TransUNet was improved by 1.13%, Hausdorff distance was reduced by 2.38%, and weight was maintained. The experimental results show that the network can effectively segment tissues and organs in medical images, which has important theoretical significance and application value for intelligent clinical diagnosis and treatment.
- Open Access
Mixed-UNet: Refined Class Activation Mapping for Weakly-Supervised Semantic Segmentation with Multi-scale Inference
Deep learning techniques have shown great potential in medical image processing, particularly through accurate and reliable image segmentation on magnetic resonance imaging (MRI) scans or computed tomography (CT) scans, which allow the localization and diagnosis of lesions. However, training these segmentation models requires a large number of manually annotated pixel-level labels, which are time-consuming and labor-intensive, in contrast to image-level labels that are easier to obtain. It is imperative to resolve this problem through weakly-supervised semantic segmentation models using image-level labels as supervision since it can significantly reduce human annotation efforts. Most of the advanced solutions exploit class activation mapping (CAM). However, the original CAMs rarely capture the precise boundaries of lesions. In this study, we propose the strategy of multi-scale inference to refine CAMs by reducing the detail loss in single-scale reasoning. For segmentation, we develop a novel model named Mixed-UNet, which has two parallel branches in the decoding phase. The results can be obtained after fusing the extracted features from two branches. We evaluate the designed Mixed-UNet against several prevalent deep learning-based segmentation approaches on our dataset collected from the local hospital and public datasets. The validation results demonstrate that our model surpasses available methods under the same supervision level in the segmentation of various lesions from brain imaging.
Open AccessMixed-UNet: Refined Class Activation Mapping for Weakly-Supervised Semantic Segmentation with Multi-scale Inference
Deep learning techniques have shown great potential in medical image processing, particularly through accurate and reliable image segmentation on magnetic resonance imaging (MRI) scans or computed tomography (CT) scans, which allow the localization and diagnosis of lesions. However, training these segmentation models requires a large number of manually annotated pixel-level labels, which are time-consuming and labor-intensive, in contrast to image-level labels that are easier to obtain. It is imperative to resolve this problem through weakly-supervised semantic segmentation models using image-level labels as supervision since it can significantly reduce human annotation efforts. Most of the advanced solutions exploit class activation mapping (CAM). However, the original CAMs rarely capture the precise boundaries of lesions. In this study, we propose the strategy of multi-scale inference to refine CAMs by reducing the detail loss in single-scale reasoning. For segmentation, we develop a novel model named Mixed-UNet, which has two parallel branches in the decoding phase. The results can be obtained after fusing the extracted features from two branches. We evaluate the designed Mixed-UNet against several prevalent deep learning-based segmentation approaches on our dataset collected from the local hospital and public datasets. The validation results demonstrate that our model surpasses available methods under the same supervision level in the segmentation of various lesions from brain imaging.