Yuan Ni
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Yuan Ni has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Guotong Xie
Ping An Technology, Shenzhen, China
4 shared publications - Bin Lv
Ping An Technology, 12F Building B PingAn IFC No. 1-3 Xinyuan South Road, Beijing, 100027, PR China
2 shared publications - Dan Song
Department of Ophthalmology, Peking University International Hospital, No.1 Shengmingyuan Road, Zhongguancun Life Science Park, Changping District, Beijing, PR China
2 shared publications - Guangfeng Liu
Department of Ophthalmology, Peking University International Hospital, No.1 Shengmingyuan Road, Zhongguancun Life Science Park, Changping District, Beijing, PR China
2 shared publications - Guanzheng Wang
Ping An Technology, 12F Building B PingAn IFC No. 1-3 Xinyuan South Road, Beijing, 100027, PR China
2 shared publications Bin Lv
Ping An Technology, Shanghai, China
1 shared publication
Latest Publications
- Open Access
- 1 February 2025
- Published by Elsevier in Photodiagnosis and Photodynamic Therapy
- Vol. 52, 104511
Age and gender-related changes in choroidal thickness: Insights from deep learning analysis of swept-source OCT images
- 1 February 2025
- Published by Elsevier in Photodiagnosis and Photodynamic Therapy
- Vol. 52, 104511
Open AccessAge and gender-related changes in choroidal thickness: Insights from deep learning analysis of swept-source OCT images
- Open Access
- 7 January 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-11
Evaluation of an acne lesion detection and severity grading model for Chinese population in online and offline healthcare scenarios
Accurate acne severity grading is crucial for effective clinical treatment and timely follow-up management. Although some artificial intelligence methods have been developed to automate the process of acne severity grading, the diversity of acne image capture sources and the various application scenarios can affect their performance. Therefore, it’s necessary to design special methods and evaluate them systematically before introducing them into clinical practice. To develop and evaluate a deep learning-based algorithm that could accurately accomplish acne lesion detection and severity grading simultaneously in different healthcare scenarios. We collected 2,157 facial images from two public and three self-built datasets for model development and evaluation. An algorithm called AcneDGNet was constructed with a feature extraction module, a lesion detection module and a severity grading module. Its performance was evaluated in both online and offline healthcare scenarios. Experimental results on the largest and most diverse evaluation datasets revealed that the overall performance for acne severity grading achieved accuracies of 89.5% in online scenarios and 89.8% in offline scenarios. For follow-up visits in online scenarios, the accuracy for detecting the changing trends reached 87.8%, with a total counting error of 1.91 ± 3.28 for all acne lesions. Additionally, the prospective evaluation demonstrated that AcneDGNet was not only much more accurate for acne grading than junior dermatologists but also comparable to the accuracy of senior dermatologists. These findings indicated that AcneDGNet can effectively assist dermatologists and patients in the diagnosis and management of acne, both in online and offline healthcare scenarios.
- 7 January 2025
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 1-11
Open AccessEvaluation of an acne lesion detection and severity grading model for Chinese population in online and offline healthcare scenarios
Accurate acne severity grading is crucial for effective clinical treatment and timely follow-up management. Although some artificial intelligence methods have been developed to automate the process of acne severity grading, the diversity of acne image capture sources and the various application scenarios can affect their performance. Therefore, it’s necessary to design special methods and evaluate them systematically before introducing them into clinical practice. To develop and evaluate a deep learning-based algorithm that could accurately accomplish acne lesion detection and severity grading simultaneously in different healthcare scenarios. We collected 2,157 facial images from two public and three self-built datasets for model development and evaluation. An algorithm called AcneDGNet was constructed with a feature extraction module, a lesion detection module and a severity grading module. Its performance was evaluated in both online and offline healthcare scenarios. Experimental results on the largest and most diverse evaluation datasets revealed that the overall performance for acne severity grading achieved accuracies of 89.5% in online scenarios and 89.8% in offline scenarios. For follow-up visits in online scenarios, the accuracy for detecting the changing trends reached 87.8%, with a total counting error of 1.91 ± 3.28 for all acne lesions. Additionally, the prospective evaluation demonstrated that AcneDGNet was not only much more accurate for acne grading than junior dermatologists but also comparable to the accuracy of senior dermatologists. These findings indicated that AcneDGNet can effectively assist dermatologists and patients in the diagnosis and management of acne, both in online and offline healthcare scenarios.
- Open Access
- 21 May 2024
- Published by Elsevier in Photodiagnosis and Photodynamic Therapy
- Vol. 47, 104218
Evaluation of choroid vascular layer thickness in wet age-related macular degeneration using artificial intelligence
- 21 May 2024
- Published by Elsevier in Photodiagnosis and Photodynamic Therapy
- Vol. 47, 104218
Open AccessEvaluation of choroid vascular layer thickness in wet age-related macular degeneration using artificial intelligence
- 29 May 2023
- Published by Springer Nature in Knowledge and Information Systems
- Vol. 65 (10) , 4335-4358
Multi-task entity linking with supervision from a taxonomy
- 29 May 2023
- Published by Springer Nature in Knowledge and Information Systems
- Vol. 65 (10) , 4335-4358
Multi-task entity linking with supervision from a taxonomy
- Open Access
- 30 July 2021
- Published by Springer Nature in BMC Medical Informatics and Decision Making
- Vol. 21 (S2) , 1-8
CapsTM: capsule network for Chinese medical text matching
Background: Text Matching (TM) is a fundamental task of natural language processing widely used in many application systems such as information retrieval, automatic question answering, machine translation, dialogue system, reading comprehension, etc. In recent years, a large number of deep learning neural networks have been applied to TM, and have refreshed benchmarks of TM repeatedly. Among the deep learning neural networks, convolutional neural network (CNN) is one of the most popular networks, which suffers from difficulties in dealing with small samples and keeping relative structures of features. In this paper, we propose a novel deep learning architecture based on capsule network for TM, called CapsTM, where capsule network is a new type of neural network architecture proposed to address some of the short comings of CNN and shows great potential in many tasks.Methods: CapsTM is a five-layer neural network, including an input layer, a representation layer, an aggregation layer, a capsule layer and a prediction layer. In CapsTM, two pieces of text are first individually converted into sequences of embeddings and are further transformed by a highway network in the input layer. Then, Bidirectional Long Short-Term Memory (BiLSTM) is used to represent each piece of text and attention-based interaction matrix is used to represent interactive information of the two pieces of text in the representation layer. Subsequently, the two kinds of representations are fused together by BiLSTM in the aggregation layer, and are further represented with capsules (vectors) in the capsule layer. Finally, the prediction layer is a connected network used for classification. CapsTM is an extension of ESIM by adding a capsule layer before the prediction layer.Results: We construct a corpus of Chinese medical question matching, which contains 36,360 question pairs. This corpus is randomly split into three parts: a training set of 32,360 question pairs, a development set of 2000 question pairs and a test set of 2000 question pairs. On this corpus, we conduct a series of experiments to evaluate the proposed CapsTM and compare it with other state-of-the-art methods. CapsTM achieves the highest F-score of 0.8666.Conclusion: The experimental results demonstrate that CapsTM is effective for Chinese medical question matching and outperforms other state-of-the-art methods for comparison.
- 30 July 2021
- Published by Springer Nature in BMC Medical Informatics and Decision Making
- Vol. 21 (S2) , 1-8
Open AccessCapsTM: capsule network for Chinese medical text matching
Background: Text Matching (TM) is a fundamental task of natural language processing widely used in many application systems such as information retrieval, automatic question answering, machine translation, dialogue system, reading comprehension, etc. In recent years, a large number of deep learning neural networks have been applied to TM, and have refreshed benchmarks of TM repeatedly. Among the deep learning neural networks, convolutional neural network (CNN) is one of the most popular networks, which suffers from difficulties in dealing with small samples and keeping relative structures of features. In this paper, we propose a novel deep learning architecture based on capsule network for TM, called CapsTM, where capsule network is a new type of neural network architecture proposed to address some of the short comings of CNN and shows great potential in many tasks.Methods: CapsTM is a five-layer neural network, including an input layer, a representation layer, an aggregation layer, a capsule layer and a prediction layer. In CapsTM, two pieces of text are first individually converted into sequences of embeddings and are further transformed by a highway network in the input layer. Then, Bidirectional Long Short-Term Memory (BiLSTM) is used to represent each piece of text and attention-based interaction matrix is used to represent interactive information of the two pieces of text in the representation layer. Subsequently, the two kinds of representations are fused together by BiLSTM in the aggregation layer, and are further represented with capsules (vectors) in the capsule layer. Finally, the prediction layer is a connected network used for classification. CapsTM is an extension of ESIM by adding a capsule layer before the prediction layer.Results: We construct a corpus of Chinese medical question matching, which contains 36,360 question pairs. This corpus is randomly split into three parts: a training set of 32,360 question pairs, a development set of 2000 question pairs and a test set of 2000 question pairs. On this corpus, we conduct a series of experiments to evaluate the proposed CapsTM and compare it with other state-of-the-art methods. CapsTM achieves the highest F-score of 0.8666.Conclusion: The experimental results demonstrate that CapsTM is effective for Chinese medical question matching and outperforms other state-of-the-art methods for comparison.