Search

Your search keyword '"Fan, Mengran"' showing total 23 results
Start Over Author "Fan, Mengran"
23 results on '"Fan, Mengran"'

2. Microscopic fine-grained instance classification through deep attention

Author

Fan, Mengran, Chakrabort, Tapabrata, Chang, Eric I-Chao, Xu, Yan, and Rittscher, Jens

Subjects
Computer Science - Computer Vision and Pattern Recognition
Abstract

Fine-grained classification of microscopic image data with limited samples is an open problem in computer vision and biomedical imaging. Deep learning based vision systems mostly deal with high number of low-resolution images, whereas subtle detail in biomedical images require higher resolution. To bridge this gap, we propose a simple yet effective deep network that performs two tasks simultaneously in an end-to-end manner. First, it utilises a gated attention module that can focus on multiple key instances at high resolution without extra annotations or region proposals. Second, the global structural features and local instance features are fused for final image level classification. The result is a robust but lightweight end-to-end trainable deep network that yields state-of-the-art results in two separate fine-grained multi-instance biomedical image classification tasks: a benchmark breast cancer histology dataset and our new fungi species mycology dataset. In addition, we demonstrate the interpretability of the proposed model by visualising the concordance of the learned features with clinically relevant features.

Published
2020

3. Predicting the future risk of lung cancer: development, and internal and external validation of the CanPredict (lung) model in 19·67 million people and evaluation of model performance against seven other risk prediction models

Author

Gleeson, Fergus, Baldwin, David, Batchkala, George, Buchanan, James, Burchardt, Judith, Chakraborty, Rohan, Chana, Ravi, Chen, Yan, Coupland, Carol, Crichton, Charles, Davies, Jim, Devaraj, Anand, Fan, Mengran, Hippisley-Cox, Julia, Koleva-Kolarova, Rositsa, Lee, Richard, Liao, Weiqi, Nair, Arjun, Pickup, Lyndsey, Powell, Anne, Rittscher, Jens, Shadmaan, Amied, Shanmugam, Kandavel, Stokes, Elizabeth, Verrill, Clare, Watkins, Johnathan, Wordsworth, Sarah, Coupland, Carol A C, Baldwin, David R, and Gleeson, Fergus V

Published
2023
Full Text
View/download PDF

4. Attention mechanisms for fine-grained biomedical image analysis

Author

Fan, Mengran, Rittscher, Jens, and Chakraborty, Rohan

Subjects
616.07, Biomedical image analysis
Abstract

Problem: Recently, deep convolutional neural networks have greatly improved our ability to develop robust image analysis methods, especially those tailored to biomedical scenarios. Despite important advances in computer vision, it is usually challenging to achieve the same success on specific biomedical datasets due to certain specific challenges, such as limited size datasets, expensive annotations, high-resolution information, multiple discernible biomedical objects, and lack of interpretability. Therefore, it is still in high demand to investigate automatic image analysis methods with high interpretability and accuracy to transform high-resolution biomedical image data into meaningful quantitative information with weak annotations and limited availability of training data. Methodology: Considering the challenges of biomedical datasets and the corresponding technical limitations, this thesis aims to incorporate global context-aware information to enable automated image analysis systems to provide reliable diagnostics in clinical applications. The thesis first introduces a hybrid method equipped with U-Net and a global probabilistic model for the segmentation of dense cell populations. This model jointly exploits shape priors and global cellular interactions to compensate for the limitations of standard deep networks. Next, this thesis focuses on attention-based, interpretable fine-grained classification methods for biomedical images that is capable of jointly capturing local high-resolution detail and global background information. Outcomes: In this thesis, a number of fine-grained feature learning methods were proposed for high-resolution biomedical images. The original contributions are as follows: (1) A global probabilistic model for dense cell nucleus segmentation. (2) A gated attention mechanism for simultaneous localisation of multiple discriminative instances. (3) A multi-task learning scheme for fine-grained feature learning. (4) Two new clinical datasets for biomedical image classification.

Published
2020

7. Predicting the future risk of lung cancer: development, and internal and external validation of the CanPredict (lung) model in 19·67 million people and evaluation of model performance against seven other risk prediction models

Author

Liao, Weiqi, primary, Coupland, Carol A C, additional, Burchardt, Judith, additional, Baldwin, David R, additional, Gleeson, Fergus V, additional, Hippisley-Cox, Julia, additional, Gleeson, Fergus, additional, Baldwin, David, additional, Batchkala, George, additional, Buchanan, James, additional, Chakraborty, Rohan, additional, Chana, Ravi, additional, Chen, Yan, additional, Coupland, Carol, additional, Crichton, Charles, additional, Davies, Jim, additional, Devaraj, Anand, additional, Fan, Mengran, additional, Koleva-Kolarova, Rositsa, additional, Lee, Richard, additional, Liao, Weiqi, additional, Nair, Arjun, additional, Pickup, Lyndsey, additional, Powell, Anne, additional, Rittscher, Jens, additional, Shadmaan, Amied, additional, Shanmugam, Kandavel, additional, Stokes, Elizabeth, additional, Verrill, Clare, additional, Watkins, Johnathan, additional, and Wordsworth, Sarah, additional

Published
2023
Full Text
View/download PDF

10. 双层花状光学涡旋晶格产生及特性研究

Author

韩冰 Han Bing, 樊海豪 Fan Haihao, 凡梦然 Fan Mengran, 景晓楠 Jing Xiaonan, 温冰艳 Wen Bingyan, and 李新忠 Li Xinzhong

Subjects
Atomic and Molecular Physics, and Optics, Electronic, Optical and Magnetic Materials
Published
2023

12. Association of Gut Microbiota during Early Pregnancy with Risk of Incident Gestational Diabetes Mellitus

Author

Hu, Ping, primary, Chen, Xiuyi, additional, Chu, Xufeng, additional, Fan, Mengran, additional, Ye, Yi, additional, Wang, Yi, additional, Han, Maozhen, additional, Yang, Xue, additional, Yuan, Jiaying, additional, Zha, Li, additional, Zhao, Bin, additional, Yang, Chun-Xia, additional, Qi, Xiao-Rong, additional, Ning, Kang, additional, Debelius, Justin, additional, Ye, Weimin, additional, Xiong, Bo, additional, Pan, Xiong-Fei, additional, and Pan, An, additional

Published
2021
Full Text
View/download PDF

15. GPU-Accelerated Flexible Molecular Docking

Author

Fan, Mengran, primary, Wang, Jian, additional, Jiang, Huaipan, additional, Feng, Yilin, additional, Mahdavi, Mehrdad, additional, Madduri, Kamesh, additional, Kandemir, Mahmut T., additional, and Dokholyan, Nikolay V., additional

Published
2021
Full Text
View/download PDF

17. Association Of Gut Microbiota During Early Pregnancy With Risk Of Incident Gestational Diabetes Mellitus: A Nested Case-Control Study

Author

Hu, Ping, primary, Chen, Xiuyi, additional, Chu, Xufeng, additional, Fan, Mengran, additional, Ye, Yi, additional, Wang, Yi, additional, Han, Maozhen, additional, Yang, Xue, additional, Yuan, Jiaying, additional, Zha, Li, additional, Zhao, Bin, additional, Yang, Chunxia, additional, Qi, Xiaorong, additional, Ning, Kang, additional, Debelius, Justin, additional, Ye, Weimin, additional, Xiong, Bo, additional, Pan, Xiongfei, additional, and Pan, An, additional

Published
2020
Full Text
View/download PDF

23. Automated Annotator: Capturing Expert Knowledge for Free.

Author

Elmes S, Chakraborti T, Fan M, Uhlig H, and Rittscher J

Subjects
Automation, Data Curation, Histology
Abstract

Deep learning enabled medical image analysis is heavily reliant on expert annotations which is costly. We present a simple yet effective automated annotation pipeline that uses autoencoder based heatmaps to exploit high level information that can be extracted from a histology viewer in an unobtrusive fashion. By predicting heatmaps on unseen images the model effectively acts like a robot annotator. The method is demonstrated in the context of coeliac disease histology images in this initial work, but the approach is task agnostic and may be used for other medical image annotation applications. The results are evaluated by a pathologist and also empirically using a deep network for coeliac disease classification. Initial results using this simple but effective approach are encouraging and merit further investigation, specially considering the possibility of scaling this up to a large number of users.

Published
2021
Full Text
View/download PDF

Catalog

Books, media, physical & digital resources
See catalog results