1
|
Luo X, Wang J, Tan C, Dou Q, Han Z, Wang Z, Tasnim F, Wang X, Zhan Q, Li X, Zhou Q, Cheng J, Liao F, Yip HC, Jiang J, Tan RT, Liu S, Yu H. Rapid Endoscopic Diagnosis of Benign Ulcerative Colorectal Diseases With an Artificial Intelligence Contextual Framework. Gastroenterology 2024; 167:591-603.e9. [PMID: 38583724 DOI: 10.1053/j.gastro.2024.03.039] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 05/24/2023] [Revised: 03/22/2024] [Accepted: 03/28/2024] [Indexed: 04/09/2024]
Abstract
BACKGROUND & AIMS Benign ulcerative colorectal diseases (UCDs) such as ulcerative colitis, Crohn's disease, ischemic colitis, and intestinal tuberculosis share similar phenotypes with different etiologies and treatment strategies. To accurately diagnose closely related diseases like UCDs, we hypothesize that contextual learning is critical in enhancing the ability of the artificial intelligence models to differentiate the subtle differences in lesions amidst the vastly divergent spatial contexts. METHODS White-light colonoscopy datasets of patients with confirmed UCDs and healthy controls were retrospectively collected. We developed a Multiclass Contextual Classification (MCC) model that can differentiate among the mentioned UCDs and healthy controls by incorporating the tissue object contexts surrounding the individual lesion region in a scene and spatial information from other endoscopic frames (video-level) into a unified framework. Internal and external datasets were used to validate the model's performance. RESULTS Training datasets included 762 patients, and the internal and external testing cohorts included 257 patients and 293 patients, respectively. Our MCC model provided a rapid reference diagnosis on internal test sets with a high averaged area under the receiver operating characteristic curve (image-level: 0.950 and video-level: 0.973) and balanced accuracy (image-level: 76.1% and video-level: 80.8%), which was superior to junior endoscopists (accuracy: 71.8%, P < .0001) and similar to experts (accuracy: 79.7%, P = .732). The MCC model achieved an area under the receiver operating characteristic curve of 0.988 and balanced accuracy of 85.8% using external testing datasets. CONCLUSIONS These results enable this model to fit in the routine endoscopic workflow, and the contextual framework to be adopted for diagnosing other closely related diseases.
Collapse
Affiliation(s)
- Xiaobei Luo
- Guangdong Provincial Key Laboratory of Gastroenterology, Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou, Guangdong, China; Department of Gastroenterology, Zhuhai People's Hospital (Zhuhai Clinical Medical College of Jinan University), Zhuhai, Guangdong, China.
| | - Jiahao Wang
- Mechanobiology Institute, National University of Singapore, Singapore; Institute of Bioengineering and Bioimaging, Agency for Science, Technology and Research (A∗STAR), Singapore
| | - Chuanchuan Tan
- The First Hospital of Hunan University of Chinese Medicine, Hunan, China
| | - Qi Dou
- Department of Computer Science and Engineering, The Chinese University of Hong Kong, Hong Kong
| | - Zelong Han
- Guangdong Provincial Key Laboratory of Gastroenterology, Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou, Guangdong, China
| | - Zhenjiang Wang
- Department of Gastroenterology, Zhuhai People's Hospital (Zhuhai Clinical Medical College of Jinan University), Zhuhai, Guangdong, China
| | - Farah Tasnim
- Institute of Bioengineering and Bioimaging, Agency for Science, Technology and Research (A∗STAR), Singapore
| | - Xiyu Wang
- Guangdong Provincial Key Laboratory of Gastroenterology, Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou, Guangdong, China
| | - Qiang Zhan
- Department of Gastroenterology, The Affiliated Wuxi People's Hospital of Nanjing Medical University, Wuxi People's Hospital, Wuxi Medical Center, Nanjing Medical University, Wuxi, Jiangsu, China
| | - Xiang Li
- Digestive Department of The Second Affiliated Hospital, School of Medicine, The Chinese University of Hong Kong, Shenzhen & Longgang District People's Hospital of Shenzhen, Shenzhen, China
| | - Qunyan Zhou
- Department of Gastroenterology, The Affiliated Wuxi People's Hospital of Nanjing Medical University, Wuxi People's Hospital, Wuxi Medical Center, Nanjing Medical University, Wuxi, Jiangsu, China
| | - Jianbin Cheng
- Department of Gastroenterology, Zhuhai People's Hospital (Zhuhai Clinical Medical College of Jinan University), Zhuhai, Guangdong, China
| | - Fabiao Liao
- Digestive Department of The Second Affiliated Hospital, School of Medicine, The Chinese University of Hong Kong, Shenzhen & Longgang District People's Hospital of Shenzhen, Shenzhen, China
| | - Hon Chi Yip
- Division of Upper Gastrointestinal and Metabolic Surgery, Department of Surgery, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong
| | - Jiayi Jiang
- Guangdong Provincial Key Laboratory of Gastroenterology, Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou, Guangdong, China
| | - Robby T Tan
- Department of Electrical and Computer Engineering, National University of Singapore, Singapore
| | - Side Liu
- Guangdong Provincial Key Laboratory of Gastroenterology, Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou, Guangdong, China; Department of Gastroenterology, Zhuhai People's Hospital (Zhuhai Clinical Medical College of Jinan University), Zhuhai, Guangdong, China.
| | - Hanry Yu
- Guangdong Provincial Key Laboratory of Gastroenterology, Department of Gastroenterology, Nanfang Hospital, Southern Medical University, Guangzhou, Guangdong, China; Mechanobiology Institute, National University of Singapore, Singapore; Institute of Bioengineering and Bioimaging, Agency for Science, Technology and Research (A∗STAR), Singapore; CAMP, Singapore-MIT Alliance for Research and Technology, Singapore; Department of Physiology, The Institute for Digital Medicine (WisDM), Yong Loo Lin School of Medicine, Singapore.
| |
Collapse
|
2
|
K P AG, D RR, N MS, P LB. Gastrointestinal tract disease detection via deep learning based structural and statistical features optimized hexa-classification model. Technol Health Care 2024:THC240603. [PMID: 39031411 DOI: 10.3233/thc-240603] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 07/22/2024]
Abstract
BACKGROUND Gastrointestinal tract (GIT) diseases impact the entire digestive system, spanning from the mouth to the anus. Wireless Capsule Endoscopy (WCE) stands out as an effective analytic instrument for Gastrointestinal tract diseases. Nevertheless, accurately identifying various lesion features, such as irregular sizes, shapes, colors, and textures, remains challenging in this field. OBJECTIVE Several computer vision algorithms have been introduced to tackle these challenges, but many relied on handcrafted features, resulting in inaccuracies in various instances. METHODS In this work, a novel Deep SS-Hexa model is proposed which is a combination two different deep learning structures for extracting two different features from the WCE images to detect various GIT ailment. The gathered images are denoised by weighted median filter to remove the noisy distortions and augment the images for enhancing the training data. The structural and statistical (SS) feature extraction process is sectioned into two phases for the analysis of distinct regions of gastrointestinal. In the first stage, statistical features of the image are retrieved using MobileNet with the support of SiLU activation function to retrieve the relevant features. In the second phase, the segmented intestine images are transformed into structural features to learn the local information. These SS features are parallelly fused for selecting the best relevant features with walrus optimization algorithm. Finally, Deep belief network (DBN) is used classified the GIT diseases into hexa classes namely normal, ulcer, pylorus, cecum, esophagitis and polyps on the basis of the selected features. RESULTS The proposed Deep SS-Hexa model attains an overall average accuracy of 99.16% in GIT disease detection based on KVASIR and KID datasets. The proposed Deep SS-Hexa model achieves high level of accuracy with minimal computational cost in the recognition of GIT illness. CONCLUSIONS The proposed Deep SS-Hexa Model progresses the overall accuracy range of 0.04%, 0.80% better than GastroVision, Genetic algorithm based on KVASIR dataset and 0.60%, 1.21% better than Modified U-Net, WCENet based on KID dataset respectively.
Collapse
Affiliation(s)
- Ajitha Gladis K P
- Department of Information Technology, CSI Institute of Technology, Thovalai, India
| | - Roja Ramani D
- Department of Computer Science and Engineering, New Horizon College of Engineering, Bengaluru, India
| | - Mohana Suganthi N
- Department of Computer Science and Engineering, Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, Chennai, India
| | - Linu Babu P
- Department of Electronics and Communication Engineering, IES College of Engineering, Thrissur, India
| |
Collapse
|
3
|
Mir H, Sadeghi V, Vard A, Dehnavi AM. Identification of Circular Patterns in Capsule Endoscopy Bubble Frames. JOURNAL OF MEDICAL SIGNALS & SENSORS 2024; 14:15. [PMID: 39100744 PMCID: PMC11296570 DOI: 10.4103/jmss.jmss_50_23] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/22/2023] [Revised: 01/15/2024] [Accepted: 01/30/2024] [Indexed: 08/06/2024]
Abstract
Background A significant number of frames captured by the wireless capsule endoscopy are involved with varying amounts of bubbles. Whereas different studies have considered bubbles as nonuseful agents due to the fact that they reduce the visualization quality of the small intestine mucosa, this research aims to develop a practical way of assessing the rheological capability of the circular bubbles as a suggestion for future clinical diagnostic purposes. Methods From the Kvasir-capsule endoscopy dataset, frames with varying levels of bubble engagements were chosen in two categories based on bubble size. Border reflections are present on the edges of round-shaped bubbles in their boundaries, and in the frequency domain, high-frequency bands correspond to these edges in the spatial domain. The first step is about high-pass filtering of border reflections using wavelet transform (WT) and Differential of Gaussian, and the second step is related to applying the Fast Circlet Transform (FCT) and the Hough transform as circle detection tools on extracted borders and evaluating the distribution and abundance of all bubbles with the variety of radii. Results Border's extraction using WT as a preprocessing approach makes it easier for circle detection tool for better concentration on high-frequency circular patterns. Consequently, applying FCT with predefined parameters can specify the variety and range of radius and the abundance for all bubbles in an image. The overall discrimination factor (ODF) of 15.01, and 7.1 showing distinct bubble distributions in the gastrointestinal (GI) tract. The discrimination in ODF from datasets 1-2 suggests a relationship between the rheological properties of bubbles and their coverage area plus their abundance, highlighting the WT and FCT performance in determining bubbles' distributions for diagnostic objectives. Conclusion The implementation of an object-oriented attitude in gastrointestinal analysis makes it intelligible for gastroenterologists to approximate the constituent features of intra-intestinal fluids. this can't be evaluated until the bubbles are considered as non-useful agents. The obtained results from the datasets proved that the difference between the calculated ODF can be used as an indicator for the quality estimation of intraintestinal fluids' rheological features like viscosity, which helps gastroenterologists evaluate the quality of patient digestion.
Collapse
Affiliation(s)
- Hossein Mir
- Department of Bio-Electrics and Biomedical Engineering, School of Advanced Technologies in Medicine, Isfahan University of Medical Sciences, Isfahan, Iran
| | - Vahid Sadeghi
- Department of Bio-Electrics and Biomedical Engineering, School of Advanced Technologies in Medicine, Isfahan University of Medical Sciences, Isfahan, Iran
| | - Alireza Vard
- Department of Bio-Electrics and Biomedical Engineering, School of Advanced Technologies in Medicine, Isfahan University of Medical Sciences, Isfahan, Iran
- Medical Image and Signal Processing Research Center, School of Advanced Technologies in Medicine, Isfahan University of Medical Sciences, Isfahan, Iran
| | - Alireza Mehri Dehnavi
- Department of Bio-Electrics and Biomedical Engineering, School of Advanced Technologies in Medicine, Isfahan University of Medical Sciences, Isfahan, Iran
- Medical Image and Signal Processing Research Center, School of Advanced Technologies in Medicine, Isfahan University of Medical Sciences, Isfahan, Iran
| |
Collapse
|
4
|
Qu A, Wu Q, Wang J, Yu L, Li J, Liu J. TNCB: Tri-Net With Cross-Balanced Pseudo Supervision for Class Imbalanced Medical Image Classification. IEEE J Biomed Health Inform 2024; 28:2187-2198. [PMID: 38329849 DOI: 10.1109/jbhi.2024.3362243] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/10/2024]
Abstract
In clinical settings, the implementation of deep neural networks is impeded by the prevalent problems of label scarcity and class imbalance in medical images. To mitigate the need for labeled data, semi-supervised learning (SSL) has gained traction. However, existing SSL schemes exhibit certain limitations. 1) They commonly fail to address the class imbalance problem. Training with imbalanced data makes the model's prediction biased towards majority classes, consequently introducing prediction bias. 2) They usually suffer from training bias arising from unreasonable training strategies, such as strong coupling between the generation and utilization of pseudo labels. To address these problems, we propose a novel SSL framework called Tri-Net with Cross-Balanced pseudo supervision (TNCB). Specifically, two student networks focusing on different learning tasks and a teacher network equipped with an adaptive balancer are designed. This design enables the teacher model to pay more focus on minority classes, thereby reducing prediction bias. Additionally, we propose a virtual optimization strategy to further enhance the teacher model's resistance to class imbalance. Finally, to fully exploit valuable knowledge from unlabeled images, we employ cross-balanced pseudo supervision, where an adaptive cross loss function is introduced to reduce training bias. Extensive evaluation on four datasets with different diseases, image modalities, and imbalance ratios consistently demonstrate the superior performance of TNCB over state-of-the-art SSL methods. These results indicate the effectiveness and robustness of TNCB in addressing imbalanced medical image classification challenges.
Collapse
|
5
|
Yokote A, Umeno J, Kawasaki K, Fujioka S, Fuyuno Y, Matsuno Y, Yoshida Y, Imazu N, Miyazono S, Moriyama T, Kitazono T, Torisu T. Small bowel capsule endoscopy examination and open access database with artificial intelligence: The SEE-artificial intelligence project. DEN OPEN 2024; 4:e258. [PMID: 37359150 PMCID: PMC10288072 DOI: 10.1002/deo2.258] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 03/08/2023] [Revised: 05/31/2023] [Accepted: 06/05/2023] [Indexed: 06/28/2023]
Abstract
OBJECTIVES Artificial intelligence (AI) may be practical for image classification of small bowel capsule endoscopy (CE). However, creating a functional AI model is challenging. We attempted to create a dataset and an object detection CE AI model to explore modeling problems to assist in reading small bowel CE. METHODS We extracted 18,481 images from 523 small bowel CE procedures performed at Kyushu University Hospital from September 2014 to June 2021. We annotated 12,320 images with 23,033 disease lesions, combined them with 6161 normal images as the dataset, and examined the characteristics. Based on the dataset, we created an object detection AI model using YOLO v5 and we tested validation. RESULTS We annotated the dataset with 12 types of annotations, and multiple annotation types were observed in the same image. We test validated our AI model with 1396 images, and sensitivity for all 12 types of annotations was about 91%, with 1375 true positives, 659 false positives, and 120 false negatives detected. The highest sensitivity for individual annotations was 97%, and the highest area under the receiver operating characteristic curve was 0.98, but the quality of detection varied depending on the specific annotation. CONCLUSIONS Object detection AI model in small bowel CE using YOLO v5 may provide effective and easy-to-understand reading assistance. In this SEE-AI project, we open our dataset, the weights of the AI model, and a demonstration to experience our AI. We look forward to further improving the AI model in the future.
Collapse
Affiliation(s)
- Akihito Yokote
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Junji Umeno
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Keisuke Kawasaki
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Shin Fujioka
- Department of Endoscopic Diagnostics and Therapeutics Kyushu University Hospital Fukuoka Japan
| | - Yuta Fuyuno
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Yuichi Matsuno
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Yuichiro Yoshida
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Noriyuki Imazu
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Satoshi Miyazono
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Tomohiko Moriyama
- International Medical Department Kyushu University Hospital Fukuoka Japan
| | - Takanari Kitazono
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| | - Takehiro Torisu
- Department of Medicine and Clinical Science Graduate School of Medical Science Kyushu University Fukuoka Japan
| |
Collapse
|
6
|
Jiang B, Dorosan M, Leong JWH, Ong MEH, Lam SSW, Ang TL. Development and validation of a deep learning system for detection of small bowel pathologies in capsule endoscopy: a pilot study in a Singapore institution. Singapore Med J 2024; 65:133-140. [PMID: 38527297 PMCID: PMC11060635 DOI: 10.4103/singaporemedj.smj-2023-187] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/28/2023] [Accepted: 12/10/2023] [Indexed: 03/27/2024]
Abstract
INTRODUCTION Deep learning models can assess the quality of images and discriminate among abnormalities in small bowel capsule endoscopy (CE), reducing fatigue and the time needed for diagnosis. They serve as a decision support system, partially automating the diagnosis process by providing probability predictions for abnormalities. METHODS We demonstrated the use of deep learning models in CE image analysis, specifically by piloting a bowel preparation model (BPM) and an abnormality detection model (ADM) to determine frame-level view quality and the presence of abnormal findings, respectively. We used convolutional neural network-based models pretrained on large-scale open-domain data to extract spatial features of CE images that were then used in a dense feed-forward neural network classifier. We then combined the open-source Kvasir-Capsule dataset (n = 43) and locally collected CE data (n = 29). RESULTS Model performance was compared using averaged five-fold and two-fold cross-validation for BPMs and ADMs, respectively. The best BPM model based on a pre-trained ResNet50 architecture had an area under the receiver operating characteristic and precision-recall curves of 0.969±0.008 and 0.843±0.041, respectively. The best ADM model, also based on ResNet50, had top-1 and top-2 accuracies of 84.03±0.051 and 94.78±0.028, respectively. The models could process approximately 200-250 images per second and showed good discrimination on time-critical abnormalities such as bleeding. CONCLUSION Our pilot models showed the potential to improve time to diagnosis in CE workflows. To our knowledge, our approach is unique to the Singapore context. The value of our work can be further evaluated in a pragmatic manner that is sensitive to existing clinician workflow and resource constraints.
Collapse
Affiliation(s)
- Bochao Jiang
- Department of Gastroenterology and Hepatology, Changi General Hospital, Singapore
| | - Michael Dorosan
- Health Services Research Centre, Singapore Health Services Pte Ltd, Singapore
| | - Justin Wen Hao Leong
- Department of Gastroenterology and Hepatology, Changi General Hospital, Singapore
| | - Marcus Eng Hock Ong
- Health Services and Systems Research, Duke-NUS Medical School, Singapore
- Department of Emergency Medicine, Singapore General Hospital, Singapore
| | - Sean Shao Wei Lam
- Health Services Research Centre, Singapore Health Services Pte Ltd, Singapore
| | - Tiing Leong Ang
- Department of Gastroenterology and Hepatology, Changi General Hospital, Singapore
| |
Collapse
|
7
|
Bordbar M, Helfroush MS, Danyali H, Ejtehadi F. Wireless capsule endoscopy multiclass classification using three-dimensional deep convolutional neural network model. Biomed Eng Online 2023; 22:124. [PMID: 38098015 PMCID: PMC10722702 DOI: 10.1186/s12938-023-01186-9] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/10/2023] [Accepted: 11/29/2023] [Indexed: 12/17/2023] Open
Abstract
BACKGROUND Wireless capsule endoscopy (WCE) is a patient-friendly and non-invasive technology that scans the whole of the gastrointestinal tract, including difficult-to-access regions like the small bowel. Major drawback of this technology is that the visual inspection of a large number of video frames produced during each examination makes the physician diagnosis process tedious and prone to error. Several computer-aided diagnosis (CAD) systems, such as deep network models, have been developed for the automatic recognition of abnormalities in WCE frames. Nevertheless, most of these studies have only focused on spatial information within individual WCE frames, missing the crucial temporal data within consecutive frames. METHODS In this article, an automatic multiclass classification system based on a three-dimensional deep convolutional neural network (3D-CNN) is proposed, which utilizes the spatiotemporal information to facilitate the WCE diagnosis process. The 3D-CNN model fed with a series of sequential WCE frames in contrast to the two-dimensional (2D) model, which exploits frames as independent ones. Moreover, the proposed 3D deep model is compared with some pre-trained networks. The proposed models are trained and evaluated with 29 subject WCE videos (14,691 frames before augmentation). The performance advantages of 3D-CNN over 2D-CNN and pre-trained networks are verified in terms of sensitivity, specificity, and accuracy. RESULTS 3D-CNN outperforms the 2D technique in all evaluation metrics (sensitivity: 98.92 vs. 98.05, specificity: 99.50 vs. 86.94, accuracy: 99.20 vs. 92.60). In conclusion, a novel 3D-CNN model for lesion detection in WCE frames is proposed in this study. CONCLUSION The results indicate the performance of 3D-CNN over 2D-CNN and some well-known pre-trained classifier networks. The proposed 3D-CNN model uses the rich temporal information in adjacent frames as well as spatial data to develop an accurate and efficient model.
Collapse
Affiliation(s)
- Mehrdokht Bordbar
- Department of Electrical Engineering, Shiraz University of Technology, Shiraz, Iran
| | | | - Habibollah Danyali
- Department of Electrical Engineering, Shiraz University of Technology, Shiraz, Iran
| | - Fardad Ejtehadi
- Department of Internal Medicine, Gastroenterohepatology Research Center, School of Medicine, Shiraz University of Medical Sciences, Shiraz, Iran
| |
Collapse
|
8
|
Chen Z, Li W, Xing X, Yuan Y. Medical federated learning with joint graph purification for noisy label learning. Med Image Anal 2023; 90:102976. [PMID: 37806019 DOI: 10.1016/j.media.2023.102976] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 10/06/2022] [Revised: 02/08/2023] [Accepted: 09/18/2023] [Indexed: 10/10/2023]
Abstract
In terms of increasing privacy issues, Federated Learning (FL) has received extensive attention in medical imaging. Through collaborative training, FL can produce superior diagnostic models with global knowledge, while preserving private data locally. In practice, medical diagnosis suffers from intra-/inter-observer variability, thus label noise is inevitable in dataset preparation. Different from existing studies on centralized datasets, the label noise problem in FL scenarios confronts more challenges, due to data inaccessibility and even noise heterogeneity. In this work, we propose a federated framework with joint Graph Purification (FedGP) to address the label noise in FL through server and clients collaboration. Specifically, to overcome the impact of label noise on local training, we first devise a noisy graph purification on the client side to generate reliable pseudo labels by progressively expanding the purified graph with topological knowledge. Then, we further propose a graph-guided negative ensemble loss to exploit the topology of the client-side purified graph with robust complementary supervision against label noise. Moreover, to address the FL label noise with data silos, we propose a global centroid aggregation on the server side to produce a robust classifier with global knowledge, which can be optimized collaboratively in the FL framework. Extensive experiments are conducted on endoscopic and pathological images with the comparison under the homogeneous, heterogeneous, and real-world label noise for medical FL. Among these diverse noisy FL settings, our FedGP framework significantly outperforms denoising and noisy FL state-of-the-arts by a large margin. The source code is available at https://github.com/CUHK-AIM-Group/FedGP.
Collapse
Affiliation(s)
- Zhen Chen
- Centre for Artificial Intelligence and Robotics (CAIR), Hong Kong Institute of Science & Innovation, Chinese Academy of Sciences, Hong Kong Special Administrative Region of China
| | - Wuyang Li
- Department of Electrical Engineering, City University of Hong Kong, Hong Kong Special Administrative Region of China
| | - Xiaohan Xing
- Department of Electrical Engineering, City University of Hong Kong, Hong Kong Special Administrative Region of China; Department of Radiation Oncology, Stanford University, CA, USA
| | - Yixuan Yuan
- Department of Electronic Engineering, Chinese University of Hong Kong, Hong Kong Special Administrative Region of China.
| |
Collapse
|
9
|
Zhu S, Gao J, Liu L, Yin M, Lin J, Xu C, Xu C, Zhu J. Public Imaging Datasets of Gastrointestinal Endoscopy for Artificial Intelligence: a Review. J Digit Imaging 2023; 36:2578-2601. [PMID: 37735308 PMCID: PMC10584770 DOI: 10.1007/s10278-023-00844-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/27/2023] [Revised: 05/03/2023] [Accepted: 05/03/2023] [Indexed: 09/23/2023] Open
Abstract
With the advances in endoscopic technologies and artificial intelligence, a large number of endoscopic imaging datasets have been made public to researchers around the world. This study aims to review and introduce these datasets. An extensive literature search was conducted to identify appropriate datasets in PubMed, and other targeted searches were conducted in GitHub, Kaggle, and Simula to identify datasets directly. We provided a brief introduction to each dataset and evaluated the characteristics of the datasets included. Moreover, two national datasets in progress were discussed. A total of 40 datasets of endoscopic images were included, of which 34 were accessible for use. Basic and detailed information on each dataset was reported. Of all the datasets, 16 focus on polyps, and 6 focus on small bowel lesions. Most datasets (n = 16) were constructed by colonoscopy only, followed by normal gastrointestinal endoscopy and capsule endoscopy (n = 9). This review may facilitate the usage of public dataset resources in endoscopic research.
Collapse
Affiliation(s)
- Shiqi Zhu
- Department of Gastroenterology, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Suzhou , Jiangsu, 215000, China
- Suzhou Clinical Center of Digestive Diseases, Suzhou, 215000, China
| | - Jingwen Gao
- Department of Gastroenterology, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Suzhou , Jiangsu, 215000, China
- Suzhou Clinical Center of Digestive Diseases, Suzhou, 215000, China
| | - Lu Liu
- Department of Gastroenterology, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Suzhou , Jiangsu, 215000, China
- Suzhou Clinical Center of Digestive Diseases, Suzhou, 215000, China
| | - Minyue Yin
- Department of Gastroenterology, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Suzhou , Jiangsu, 215000, China
- Suzhou Clinical Center of Digestive Diseases, Suzhou, 215000, China
| | - Jiaxi Lin
- Department of Gastroenterology, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Suzhou , Jiangsu, 215000, China
- Suzhou Clinical Center of Digestive Diseases, Suzhou, 215000, China
| | - Chang Xu
- Department of Gastroenterology, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Suzhou , Jiangsu, 215000, China
- Suzhou Clinical Center of Digestive Diseases, Suzhou, 215000, China
| | - Chunfang Xu
- Department of Gastroenterology, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Suzhou , Jiangsu, 215000, China.
- Suzhou Clinical Center of Digestive Diseases, Suzhou, 215000, China.
| | - Jinzhou Zhu
- Department of Gastroenterology, The First Affiliated Hospital of Soochow University, 188 Shizi Street, Suzhou , Jiangsu, 215000, China.
- Suzhou Clinical Center of Digestive Diseases, Suzhou, 215000, China.
| |
Collapse
|
10
|
Quindós A, Laiz P, Vitrià J, Seguí S. Self-supervised out-of-distribution detection in wireless capsule endoscopy images. Artif Intell Med 2023; 143:102606. [PMID: 37673575 DOI: 10.1016/j.artmed.2023.102606] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/22/2022] [Revised: 06/02/2023] [Accepted: 06/05/2023] [Indexed: 09/08/2023]
Abstract
While deep learning has displayed excellent performance in a broad spectrum of application areas, neural networks still struggle to recognize what they have not seen, i.e., out-of-distribution (OOD) inputs. In the medical field, building robust models that are able to detect OOD images is highly critical, as these rare images could show diseases or anomalies that should be detected. In this study, we use wireless capsule endoscopy (WCE) images to present a novel patch-based self-supervised approach comprising three stages. First, we train a triplet network to learn vector representations of WCE image patches. Second, we cluster the patch embeddings to group patches in terms of visual similarity. Third, we use the cluster assignments as pseudolabels to train a patch classifier and use the Out-of-Distribution Detector for Neural Networks (ODIN) for OOD detection. The system has been tested on the Kvasir-capsule, a publicly released WCE dataset. Empirical results show an OOD detection improvement compared to baseline methods. Our method can detect unseen pathologies and anomalies such as lymphangiectasia, foreign bodies and blood with AUROC>0.6. This work presents an effective solution for OOD detection models without needing labeled images.
Collapse
Affiliation(s)
- Arnau Quindós
- Departament de Matemàtiques i Informàtica, Universitat de Barcelona (UB), Barcelona, Spain
| | - Pablo Laiz
- Departament de Matemàtiques i Informàtica, Universitat de Barcelona (UB), Barcelona, Spain
| | - Jordi Vitrià
- Departament de Matemàtiques i Informàtica, Universitat de Barcelona (UB), Barcelona, Spain
| | - Santi Seguí
- Departament de Matemàtiques i Informàtica, Universitat de Barcelona (UB), Barcelona, Spain.
| |
Collapse
|
11
|
Huang Z, Wu J, Wang T, Li Z, Ioannou A. Class-Specific Distribution Alignment for semi-supervised medical image classification. Comput Biol Med 2023; 164:107280. [PMID: 37517324 DOI: 10.1016/j.compbiomed.2023.107280] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/18/2022] [Revised: 07/11/2023] [Accepted: 07/16/2023] [Indexed: 08/01/2023]
Abstract
Despite the success of deep neural networks in medical image classification, the problem remains challenging as data annotation is time-consuming, and the class distribution is imbalanced due to the relative scarcity of diseases. To address this problem, we propose Class-Specific Distribution Alignment (CSDA), a semi-supervised learning framework based on self-training that is suitable to learn from highly imbalanced datasets. Specifically, we first provide a new perspective to distribution alignment by considering the process as a change of basis in the vector space spanned by marginal predictions, and then derive CSDA to capture class-dependent marginal predictions on both labeled and unlabeled data, in order to avoid the bias towards majority classes. Furthermore, we propose a Variable Condition Queue (VCQ) module to maintain a proportionately balanced number of unlabeled samples for each class. Experiments on three public datasets HAM10000, CheXpert and Kvasir show that our method provides competitive performance on semi-supervised skin disease, thoracic disease, and endoscopic image classification tasks.
Collapse
Affiliation(s)
- Zhongzheng Huang
- Fujian Provincial Key Laboratory of Information Processing and Intelligent Control, College of Computer and Control Engineering, Minjiang University, Fuzhou, China; College of Computer and Data Science, Fuzhou University, Fuzhou, China
| | - Jiawei Wu
- Fujian Provincial Key Laboratory of Information Processing and Intelligent Control, College of Computer and Control Engineering, Minjiang University, Fuzhou, China; College of Mechanical and Electrical Engineering, Fujian Agriculture and Forestry University, Fuzhou, China
| | - Tao Wang
- Fujian Provincial Key Laboratory of Information Processing and Intelligent Control, College of Computer and Control Engineering, Minjiang University, Fuzhou, China; International Digital Economy College, Minjiang University, Fuzhou, China.
| | - Zuoyong Li
- Fujian Provincial Key Laboratory of Information Processing and Intelligent Control, College of Computer and Control Engineering, Minjiang University, Fuzhou, China.
| | - Anastasia Ioannou
- International Digital Economy College, Minjiang University, Fuzhou, China; Department of Computer Science and Engineering, European University Cyprus, Nicosia, Cyprus
| |
Collapse
|
12
|
Laiz P, Vitrià J, Gilabert P, Wenzek H, Malagelada C, Watson AJM, Seguí S. Anatomical landmarks localization for capsule endoscopy studies. Comput Med Imaging Graph 2023; 108:102243. [PMID: 37267757 DOI: 10.1016/j.compmedimag.2023.102243] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/14/2022] [Revised: 02/25/2023] [Accepted: 05/05/2023] [Indexed: 06/04/2023]
Abstract
Wireless Capsule Endoscopy is a medical procedure that uses a small, wireless camera to capture images of the inside of the digestive tract. The identification of the entrance and exit of the small bowel and of the large intestine is one of the first tasks that need to be accomplished to read a video. This paper addresses the design of a clinical decision support tool to detect these anatomical landmarks. We have developed a system based on deep learning that combines images, timestamps, and motion data to achieve state-of-the-art results. Our method does not only classify the images as being inside or outside the studied organs, but it is also able to identify the entrance and exit frames. The experiments performed with three different datasets (one public and two private) show that our system is able to approximate the landmarks while achieving high accuracy on the classification problem (inside/outside of the organ). When comparing the entrance and exit of the studied organs, the distance between predicted and real landmarks is reduced from 1.5 to 10 times with respect to previous state-of-the-art methods.
Collapse
Affiliation(s)
- Pablo Laiz
- Department of Mathematics and Computer Science, Universitat de Barcelona, Barcelona, Spain.
| | - Jordi Vitrià
- Department of Mathematics and Computer Science, Universitat de Barcelona, Barcelona, Spain
| | - Pere Gilabert
- Department of Mathematics and Computer Science, Universitat de Barcelona, Barcelona, Spain
| | | | | | | | - Santi Seguí
- Department of Mathematics and Computer Science, Universitat de Barcelona, Barcelona, Spain
| |
Collapse
|
13
|
Vats A, Pedersen M, Mohammed A, Hovde Ø. Evaluating clinical diversity and plausibility of synthetic capsule endoscopic images. Sci Rep 2023; 13:10857. [PMID: 37407635 PMCID: PMC10322862 DOI: 10.1038/s41598-023-36883-x] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/08/2022] [Accepted: 06/12/2023] [Indexed: 07/07/2023] Open
Abstract
Wireless Capsule Endoscopy (WCE) is being increasingly used as an alternative imaging modality for complete and non-invasive screening of the gastrointestinal tract. Although this is advantageous in reducing unnecessary hospital admissions, it also demands that a WCE diagnostic protocol be in place so larger populations can be effectively screened. This calls for training and education protocols attuned specifically to this modality. Like training in other modalities such as traditional endoscopy, CT, MRI, etc., a WCE training protocol would require an atlas comprising of a large corpora of images that show vivid descriptions of pathologies, ideally observed over a period of time. Since such comprehensive atlases are presently lacking in WCE, in this work, we propose a deep learning method for utilizing already available studies across different institutions for the creation of a realistic WCE atlas using StyleGAN. We identify clinically relevant attributes in WCE such that synthetic images can be generated with selected attributes on cue. Beyond this, we also simulate several disease progression scenarios. The generated images are evaluated for realism and plausibility through three subjective online experiments with the participation of eight gastroenterology experts from three geographical locations and a variety of years of experience. The results from the experiments indicate that the images are highly realistic and the disease scenarios plausible. The images comprising the atlas are available publicly for use in training applications as well as supplementing real datasets for deep learning.
Collapse
Affiliation(s)
- Anuja Vats
- Department of Computer Science, NTNU, 2819, Gjøvik, Norway.
| | | | - Ahmed Mohammed
- Department of Computer Science, NTNU, 2819, Gjøvik, Norway
- SINTEF Digital, Smart Sensor Systems, Oslo, Norway
| | - Øistein Hovde
- Department of Computer Science, NTNU, 2819, Gjøvik, Norway
- Innlandet Hospital Trust, 2819, Gjøvik, Norway
| |
Collapse
|
14
|
Bidokh E, Hassanpour H. Enhancing Wireless Capsule Endoscopy images from intense illumination specular reflections using the homomorphic filter. Biomed Signal Process Control 2023. [DOI: 10.1016/j.bspc.2023.104723] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 03/08/2023]
|
15
|
Joseph J, George SN, Raja K. Parameter-Free Matrix Decomposition for Specular Reflections Removal in Endoscopic Images. IEEE JOURNAL OF TRANSLATIONAL ENGINEERING IN HEALTH AND MEDICINE 2023; 11:360-374. [PMID: 37435543 PMCID: PMC10332471 DOI: 10.1109/jtehm.2023.3283444] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Figures] [Subscribe] [Scholar Register] [Received: 12/27/2022] [Revised: 05/11/2023] [Accepted: 05/24/2023] [Indexed: 07/13/2023]
Abstract
Objective: Endoscopy is a medical diagnostic procedure used to see inside the human body with the help of a camera-attached system called the endoscope. Endoscopic images and videos suffer from specular reflections (or highlight) and can have an adverse impact on the diagnostic quality of images. These scattered white regions severely affect the visual appearance of images for both endoscopists and the computer-aided diagnosis of diseases. Methods & Results: We introduce a new parameter-free matrix decomposition technique to remove the specular reflections. The proposed method decomposes the original image into a highlight-free pseudo-low-rank component and a highlight component. Along with the highlight removal, the approach also removes the boundary artifacts present around the highlight regions, unlike the previous works based on family of Robust Principal Component Analysis (RPCA). The approach is evaluated on three publicly available endoscopy datasets: Kvasir Polyp, Kvasir Normal-Pylorus and Kvasir Capsule datasets. Our evaluation is benchmarked against 4 different state-of-the-art approaches using three different well-used metrics such as Structural Similarity Index Measure (SSIM), Percentage of highlights remaining and Coefficient of Variation (CoV). Conclusions: The results show significant improvements over the compared methods on all three metrics. The approach is further validated for statistical significance where it emerges better than other state-of-the-art approaches.Clinical and Translational Impact Statement-The mathematical concepts of low rank and rank decomposition in matrix algebra are translated to remove specularities in the endoscopic images The result shows the impact of the proposed method in removing specular reflections from endoscopic images indicating improved diagnosis efficiency for both endoscopists and computer-aided diagnosis systems.
Collapse
Affiliation(s)
- Jithin Joseph
- Department of Electronics and Communication EngineeringNational Institute of Technology at CalicutKozhikode673601India
- Department of Computer ScienceNorwegian University of Science and Technology7034TrondheimNorway
| | - Sudhish N. George
- Department of Electronics and Communication EngineeringNational Institute of Technology at CalicutKozhikode673601India
| | - Kiran Raja
- Department of Computer ScienceNorwegian University of Science and Technology7034TrondheimNorway
| |
Collapse
|
16
|
Vats A, Pedersen M, Mohammed A. Concept-based reasoning in medical imaging. Int J Comput Assist Radiol Surg 2023:10.1007/s11548-023-02920-3. [PMID: 37231202 PMCID: PMC10329620 DOI: 10.1007/s11548-023-02920-3] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/09/2023] [Accepted: 04/13/2023] [Indexed: 05/27/2023]
Abstract
PURPOSE As concept-based reasoning for improving model interpretability becomes promising, the question of how to define good concepts becomes more pertinent. In domains like medical, it is not always feasible to access instances clearly representing good concepts. In this work, we propose an approach to use organically mined concepts from unlabeled data to explain classifier predictions. METHODS A Concept Mapping Module (CMM) is central to this approach. Given a capsule endoscopy image predicted as abnormal, the CMM's main task is to identify which concept explains the abnormality. It consists of two parts, namely a convolutional encoder and a similarity block. The encoder maps the incoming image into the latent vector, while the similarity block retrieves the closest aligning concept as explanation. RESULTS Abnormal images can be explained in terms of five pathology-related concepts retrieved from the latent space given by inflammation (mild and severe), vascularity, ulcer and polyp. Other non-pathological concepts found include anatomy, debris, intestinal fluid and capsule modality. CONCLUSIONS This method outlines an approach through which concept-based explanations can be generated. Exploiting the latent space of styleGAN to look for variations and using task-relevant variations for defining concepts is a powerful way through which an initial concept dictionary can be created which can subsequently be iteratively refined with much less time and resource.
Collapse
Affiliation(s)
- Anuja Vats
- Department of Computer Science, NTNU, 2815, Gjøvik, Norway.
| | | | - Ahmed Mohammed
- Department of Computer Science, NTNU, 2815, Gjøvik, Norway
- SINTEF Digital, 0373, Oslo, Norway
| |
Collapse
|
17
|
Horovistiz A, Oliveira M, Araújo H. Computer vision-based solutions to overcome the limitations of wireless capsule endoscopy. J Med Eng Technol 2023; 47:242-261. [PMID: 38231042 DOI: 10.1080/03091902.2024.2302025] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/09/2022] [Accepted: 12/28/2023] [Indexed: 01/18/2024]
Abstract
Endoscopic investigation plays a critical role in the diagnosis of gastrointestinal (GI) diseases. Since 2001, Wireless Capsule Endoscopy (WCE) has been available for small bowel exploration and is in continuous development. Over the last decade, WCE has achieved impressive improvements in areas such as miniaturisation, image quality and battery life. As a result, WCE is currently a very useful alternative to wired enteroscopy in the investigation of various small bowel abnormalities and has the potential to become the leading screening technique for the entire gastrointestinal tract. However, commercial solutions still have several limitations, namely incomplete examination and limited diagnostic capacity. These deficiencies are related to technical issues, such as image quality, motion estimation and power consumption management. Computational methods, based on image processing and analysis, can help to overcome these challenges and reduce both the time required by reviewers and human interpretation errors. Research groups have proposed a series of methods including algorithms for locating the capsule or lesion, assessing intestinal motility and improving image quality.In this work, we provide a critical review of computational vision-based methods for WCE image analysis aimed at overcoming the technological challenges of capsules. This article also reviews several representative public datasets used to evaluate the performance of WCE techniques and methods. Finally, some promising solutions of computational methods based on the analysis of multiple-camera endoscopic images are presented.
Collapse
Affiliation(s)
- Ana Horovistiz
- Institute of Systems and Robotics, University of Coimbra, Coimbra, Portugal
| | - Marina Oliveira
- Institute of Systems and Robotics, University of Coimbra, Coimbra, Portugal
- Department of Electrical and Computer Engineering (DEEC), Faculty of Sciences and Technology, University of Coimbra, Coimbra, Portugal
| | - Helder Araújo
- Institute of Systems and Robotics, University of Coimbra, Coimbra, Portugal
- Department of Electrical and Computer Engineering (DEEC), Faculty of Sciences and Technology, University of Coimbra, Coimbra, Portugal
| |
Collapse
|
18
|
A multi-centre polyp detection and segmentation dataset for generalisability assessment. Sci Data 2023; 10:75. [PMID: 36746950 PMCID: PMC9902556 DOI: 10.1038/s41597-023-01981-y] [Citation(s) in RCA: 8] [Impact Index Per Article: 8.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/20/2022] [Accepted: 01/23/2023] [Indexed: 02/08/2023] Open
Abstract
Polyps in the colon are widely known cancer precursors identified by colonoscopy. Whilst most polyps are benign, the polyp's number, size and surface structure are linked to the risk of colon cancer. Several methods have been developed to automate polyp detection and segmentation. However, the main issue is that they are not tested rigorously on a large multicentre purpose-built dataset, one reason being the lack of a comprehensive public dataset. As a result, the developed methods may not generalise to different population datasets. To this extent, we have curated a dataset from six unique centres incorporating more than 300 patients. The dataset includes both single frame and sequence data with 3762 annotated polyp labels with precise delineation of polyp boundaries verified by six senior gastroenterologists. To our knowledge, this is the most comprehensive detection and pixel-level segmentation dataset (referred to as PolypGen) curated by a team of computational scientists and expert gastroenterologists. The paper provides insight into data construction and annotation strategies, quality assurance, and technical validation.
Collapse
|
19
|
Cuevas-Rodriguez EO, Galvan-Tejada CE, Maeda-Gutiérrez V, Moreno-Chávez G, Galván-Tejada JI, Gamboa-Rosales H, Luna-García H, Moreno-Baez A, Celaya-Padilla JM. Comparative study of convolutional neural network architectures for gastrointestinal lesions classification. PeerJ 2023; 11:e14806. [PMID: 36945355 PMCID: PMC10024900 DOI: 10.7717/peerj.14806] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/31/2022] [Accepted: 01/05/2023] [Indexed: 03/18/2023] Open
Abstract
The gastrointestinal (GI) tract can be affected by different diseases or lesions such as esophagitis, ulcers, hemorrhoids, and polyps, among others. Some of them can be precursors of cancer such as polyps. Endoscopy is the standard procedure for the detection of these lesions. The main drawback of this procedure is that the diagnosis depends on the expertise of the doctor. This means that some important findings may be missed. In recent years, this problem has been addressed by deep learning (DL) techniques. Endoscopic studies use digital images. The most widely used DL technique for image processing is the convolutional neural network (CNN) due to its high accuracy for modeling complex phenomena. There are different CNNs that are characterized by their architecture. In this article, four architectures are compared: AlexNet, DenseNet-201, Inception-v3, and ResNet-101. To determine which architecture best classifies GI tract lesions, a set of metrics; accuracy, precision, sensitivity, specificity, F1-score, and area under the curve (AUC) were used. These architectures were trained and tested on the HyperKvasir dataset. From this dataset, a total of 6,792 images corresponding to 10 findings were used. A transfer learning approach and a data augmentation technique were applied. The best performing architecture was DenseNet-201, whose results were: 97.11% of accuracy, 96.3% sensitivity, 99.67% specificity, and 95% AUC.
Collapse
|
20
|
Vats A, Mohammed A, Pedersen M. From labels to priors in capsule endoscopy: a prior guided approach for improving generalization with few labels. Sci Rep 2022; 12:15708. [PMID: 36127404 PMCID: PMC9489743 DOI: 10.1038/s41598-022-19675-7] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/13/2022] [Accepted: 09/01/2022] [Indexed: 01/15/2023] Open
Abstract
The lack of generalizability of deep learning approaches for the automated diagnosis of pathologies in Wireless Capsule Endoscopy (WCE) has prevented any significant advantages from trickling down to real clinical practices. As a result, disease management using WCE continues to depend on exhaustive manual investigations by medical experts. This explains its limited use despite several advantages. Prior works have considered using higher quality and quantity of labels as a way of tackling the lack of generalization, however this is hardly scalable considering pathology diversity not to mention that labeling large datasets encumbers the medical staff additionally. We propose using freely available domain knowledge as priors to learn more robust and generalizable representations. We experimentally show that domain priors can benefit representations by acting in proxy of labels, thereby significantly reducing the labeling requirement while still enabling fully unsupervised yet pathology-aware learning. We use the contrastive objective along with prior-guided views during pretraining, where the view choices inspire sensitivity to pathological information. Extensive experiments on three datasets show that our method performs better than (or closes gap with) the state-of-the-art in the domain, establishing a new benchmark in pathology classification and cross-dataset generalization, as well as scaling to unseen pathology categories.
Collapse
Affiliation(s)
- Anuja Vats
- grid.5947.f0000 0001 1516 2393Department of Computer Science, NTNU, 2819 Gjøvik, Norway
| | - Ahmed Mohammed
- grid.5947.f0000 0001 1516 2393Department of Computer Science, NTNU, 2819 Gjøvik, Norway ,grid.4319.f0000 0004 0448 3150SINTEF Digital, Smart Sensor Systems, Oslo, Norway
| | - Marius Pedersen
- grid.5947.f0000 0001 1516 2393Department of Computer Science, NTNU, 2819 Gjøvik, Norway
| |
Collapse
|
21
|
A Robust Deep Model for Classification of Peptic Ulcer and Other Digestive Tract Disorders Using Endoscopic Images. Biomedicines 2022; 10:biomedicines10092195. [PMID: 36140296 PMCID: PMC9496137 DOI: 10.3390/biomedicines10092195] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/14/2022] [Revised: 08/23/2022] [Accepted: 08/24/2022] [Indexed: 11/17/2022] Open
Abstract
Accurate patient disease classification and detection through deep-learning (DL) models are increasingly contributing to the area of biomedical imaging. The most frequent gastrointestinal (GI) tract ailments are peptic ulcers and stomach cancer. Conventional endoscopy is a painful and hectic procedure for the patient while Wireless Capsule Endoscopy (WCE) is a useful technology for diagnosing GI problems and doing painless gut imaging. However, there is still a challenge to investigate thousands of images captured during the WCE procedure accurately and efficiently because existing deep models are not scored with significant accuracy on WCE image analysis. So, to prevent emergency conditions among patients, we need an efficient and accurate DL model for real-time analysis. In this study, we propose a reliable and efficient approach for classifying GI tract abnormalities using WCE images by applying a deep Convolutional Neural Network (CNN). For this purpose, we propose a custom CNN architecture named GI Disease-Detection Network (GIDD-Net) that is designed from scratch with relatively few parameters to detect GI tract disorders more accurately and efficiently at a low computational cost. Moreover, our model successfully distinguishes GI disorders by visualizing class activation patterns in the stomach bowls as a heat map. The Kvasir-Capsule image dataset has a significant class imbalance problem, we exploited a synthetic oversampling technique BORDERLINE SMOTE (BL-SMOTE) to evenly distribute the image among the classes to prevent the problem of class imbalance. The proposed model is evaluated against various metrics and achieved the following values for evaluation metrics: 98.9%, 99.8%, 98.9%, 98.9%, 98.8%, and 0.0474 for accuracy, AUC, F1-score, precision, recall, and loss, respectively. From the simulation results, it is noted that the proposed model outperforms other state-of-the-art models in all the evaluation metrics.
Collapse
|
22
|
Srivastava A, Tomar NK, Bagci U, Jha D. Video Capsule Endoscopy Classification using Focal Modulation Guided Convolutional Neural Network. PROCEEDINGS. IEEE INTERNATIONAL SYMPOSIUM ON COMPUTER-BASED MEDICAL SYSTEMS 2022; 2022:323-328. [PMID: 36777397 PMCID: PMC9914988 DOI: 10.1109/cbms55023.2022.00064] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/05/2022]
Abstract
Video capsule endoscopy is a hot topic in computer vision and medicine. Deep learning can have a positive impact on the future of video capsule endoscopy technology. It can improve the anomaly detection rate, reduce physicians' time for screening, and aid in real-world clinical analysis. Computer-Aided diagnosis (CADx) classification system for video capsule endoscopy has shown a great promise for further improvement. For example, detection of cancerous polyp and bleeding can lead to swift medical response and improve the survival rate of the patients. To this end, an automated CADx system must have high throughput and decent accuracy. In this study, we propose FocalConvNet, a focal modulation network integrated with lightweight convolutional layers for the classification of small bowel anatomical landmarks and luminal findings. FocalConvNet leverages focal modulation to attain global context and allows global-local spatial interactions throughout the forward pass. Moreover, the convolutional block with its intrinsic inductive/learning bias and capacity to extract hierarchical features allows our FocalConvNet to achieve favourable results with high throughput. We compare our FocalConvNet with other state-of-the-art (SOTA) on Kvasir-Capsule, a large-scale VCE dataset with 44,228 frames with 13 classes of different anomalies. We achieved the weighted F1-score, recall and Matthews correlation coefficient (MCC) of 0.6734, 0.6373 and 0.2974, respectively, outperforming SOTA methodologies. Further, we obtained the highest throughput of 148.02 images/second rate to establish the potential of FocalConvNet in a real-time clinical environment. The code of the proposed FocalConvNet is available at https://github.com/NoviceMAn-prog/FocalConvNet.
Collapse
Affiliation(s)
| | - Nikhil Kumar Tomar
- School of Computer Science and Informatics, Indira Gandhi National Open University
| | - Ulas Bagci
- Machine and Hybrid Intelligence Lab, Department of Radiology, Northwestern University, USA
| | - Debesh Jha
- Machine and Hybrid Intelligence Lab, Department of Radiology, Northwestern University, USA
| |
Collapse
|
23
|
Khadka R, Jha D, Hicks S, Thambawita V, Riegler MA, Ali S, Halvorsen P. Meta-learning with implicit gradients in a few-shot setting for medical image segmentation. Comput Biol Med 2022; 143:105227. [PMID: 35124439 DOI: 10.1016/j.compbiomed.2022.105227] [Citation(s) in RCA: 12] [Impact Index Per Article: 6.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/30/2021] [Revised: 01/05/2022] [Accepted: 01/05/2022] [Indexed: 12/26/2022]
Abstract
Widely used traditional supervised deep learning methods require a large number of training samples but often fail to generalize on unseen datasets. Therefore, a more general application of any trained model is quite limited for medical imaging for clinical practice. Using separately trained models for each unique lesion category or a unique patient population will require sufficiently large curated datasets, which is not practical to use in a real-world clinical set-up. Few-shot learning approaches can not only minimize the need for an enormous number of reliable ground truth labels that are labour-intensive and expensive, but can also be used to model on a dataset coming from a new population. To this end, we propose to exploit an optimization-based implicit model agnostic meta-learning (iMAML) algorithm under few-shot settings for medical image segmentation. Our approach can leverage the learned weights from diverse but small training samples to perform analysis on unseen datasets with high accuracy. We show that, unlike classical few-shot learning approaches, our method improves generalization capability. To our knowledge, this is the first work that exploits iMAML for medical image segmentation and explores the strength of the model on scenarios such as meta-training on unique and mixed instances of lesion datasets. Our quantitative results on publicly available skin and polyp datasets show that the proposed method outperforms the naive supervised baseline model and two recent few-shot segmentation approaches by large margins. In addition, our iMAML approach shows an improvement of 2%-4% in dice score compared to its counterpart MAML for most experiments.
Collapse
Affiliation(s)
- Rabindra Khadka
- SimulaMet, Oslo, Norway; Oslo Metropolitan University, Oslo, Norway
| | - Debesh Jha
- SimulaMet, Oslo, Norway; UiT the Arctic University of Norway, Tromsø, Norway.
| | - Steven Hicks
- SimulaMet, Oslo, Norway; Oslo Metropolitan University, Oslo, Norway
| | | | - Michael A Riegler
- SimulaMet, Oslo, Norway; UiT the Arctic University of Norway, Tromsø, Norway
| | - Sharib Ali
- Department of Engineering Science, Institute of Biomedical Engineering, University of Oxford, Oxford, UK; NIHR Oxford Biomedical Research Centre, University of Oxford, Oxford, UK.
| | - Pål Halvorsen
- SimulaMet, Oslo, Norway; Oslo Metropolitan University, Oslo, Norway
| |
Collapse
|
24
|
Muruganantham P, Balakrishnan SM. Attention Aware Deep Learning Model for Wireless Capsule Endoscopy Lesion Classification and Localization. J Med Biol Eng 2022. [DOI: 10.1007/s40846-022-00686-8] [Citation(s) in RCA: 6] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/18/2023]
|
25
|
Dray X, Toth E, de Lange T, Koulaouzidis A. Artificial intelligence, capsule endoscopy, databases, and the Sword of Damocles. Endosc Int Open 2021; 9:E1754-E1755. [PMID: 34790540 PMCID: PMC8589560 DOI: 10.1055/a-1521-4882] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Track Full Text] [Download PDF] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Indexed: 11/28/2022] Open
Affiliation(s)
- Xavier Dray
- Sorbonne University, Centre for Digestive Endoscopy, Hôpital Saint Antoine, APHP, Paris, France
| | - Ervin Toth
- Skane University Hospitals, Endoscopy Unit, Department of Gastroenterology, Malmo, Sweden
| | - Thomas de Lange
- Sahlgrenska University Hospital-Molndal, Medical Department, Gothenburg, Sweden
| | - Anastasio Koulaouzidis
- Pomeranian Medical University in Szczecin, Department of Social Medicine & Public Health, Faculty of Health Sciences, Zchodniopomorskie, Poland
| |
Collapse
|
26
|
Amiri Z, Hassanpour H, Beghdadi A. A Computer-Aided Method for Digestive System Abnormality Detection in WCE Images. JOURNAL OF HEALTHCARE ENGINEERING 2021; 2021:7863113. [PMID: 34707798 PMCID: PMC8545542 DOI: 10.1155/2021/7863113] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Download PDF] [Figures] [Subscribe] [Scholar Register] [Received: 08/31/2021] [Revised: 09/25/2021] [Accepted: 10/06/2021] [Indexed: 12/01/2022]
Abstract
Wireless capsule endoscopy (WCE) is a powerful tool for the diagnosis of gastrointestinal diseases. The output of this tool is in video with a length of about eight hours, containing about 8000 frames. It is a difficult task for a physician to review all of the video frames. In this paper, a new abnormality detection system for WCE images is proposed. The proposed system has four main steps: (1) preprocessing, (2) region of interest (ROI) extraction, (3) feature extraction, and (4) classification. In ROI extraction, at first, distinct areas are highlighted and nondistinct areas are faded by using the joint normal distribution; then, distinct areas are extracted as an ROI segment by considering a threshold. The main idea is to extract abnormal areas in each frame. Therefore, it can be used to extract various lesions in WCE images. In the feature extraction step, three different types of features (color, texture, and shape) are employed. Finally, the features are classified using the support vector machine. The proposed system was tested on the Kvasir-Capsule dataset. The proposed system can detect multiple lesions from WCE frames with high accuracy.
Collapse
Affiliation(s)
- Zahra Amiri
- Image Processing and Data Mining Lab, Shahrood University of Technology, Shahrood, Iran
| | - Hamid Hassanpour
- Image Processing and Data Mining Lab, Shahrood University of Technology, Shahrood, Iran
| | - Azeddine Beghdadi
- Department of Computer Science and Engineering, University Sorbonne Paris Nord, Villetaneuse, France
| |
Collapse
|
27
|
Li K, Fathan MI, Patel K, Zhang T, Zhong C, Bansal A, Rastogi A, Wang JS, Wang G. Colonoscopy polyp detection and classification: Dataset creation and comparative evaluations. PLoS One 2021; 16:e0255809. [PMID: 34403452 PMCID: PMC8370621 DOI: 10.1371/journal.pone.0255809] [Citation(s) in RCA: 25] [Impact Index Per Article: 8.3] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/05/2021] [Accepted: 07/25/2021] [Indexed: 12/12/2022] Open
Abstract
Colorectal cancer (CRC) is one of the most common types of cancer with a high mortality rate. Colonoscopy is the preferred procedure for CRC screening and has proven to be effective in reducing CRC mortality. Thus, a reliable computer-aided polyp detection and classification system can significantly increase the effectiveness of colonoscopy. In this paper, we create an endoscopic dataset collected from various sources and annotate the ground truth of polyp location and classification results with the help of experienced gastroenterologists. The dataset can serve as a benchmark platform to train and evaluate the machine learning models for polyp classification. We have also compared the performance of eight state-of-the-art deep learning-based object detection models. The results demonstrate that deep CNN models are promising in CRC screening. This work can serve as a baseline for future research in polyp detection and classification.
Collapse
Affiliation(s)
- Kaidong Li
- Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, United States of America
| | - Mohammad I. Fathan
- Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, United States of America
| | - Krushi Patel
- Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, United States of America
| | - Tianxiao Zhang
- Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, United States of America
| | - Cuncong Zhong
- Department of Electrical Engineering and Computer Science, The University of Kansas, Lawrence, KS, United States of America
| | - Ajay Bansal
- Gastroenterology, Hepatology and Motility, The University of Kansas Medical Center, Kansas City, KS, United States of America
| | - Amit Rastogi
- Gastroenterology, Hepatology and Motility, The University of Kansas Medical Center, Kansas City, KS, United States of America
| | - Jean S. Wang
- Department of Medicine, Washington University School of Medicine, Saint Louis, MO, United States of America
| | - Guanghui Wang
- Department of Computer Science, Ryerson University, Toronto, ON, Canada
| |
Collapse
|
28
|
Jha D, Ali S, Tomar NK, Johansen HD, Johansen D, Rittscher J, Riegler MA, Halvorsen P. Real-Time Polyp Detection, Localization and Segmentation in Colonoscopy Using Deep Learning. IEEE ACCESS : PRACTICAL INNOVATIONS, OPEN SOLUTIONS 2021; 9:40496-40510. [PMID: 33747684 PMCID: PMC7968127 DOI: 10.1109/access.2021.3063716] [Citation(s) in RCA: 69] [Impact Index Per Article: 23.0] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Subscribe] [Scholar Register] [Received: 02/02/2021] [Accepted: 02/15/2021] [Indexed: 05/16/2023]
Abstract
Computer-aided detection, localisation, and segmentation methods can help improve colonoscopy procedures. Even though many methods have been built to tackle automatic detection and segmentation of polyps, benchmarking of state-of-the-art methods still remains an open problem. This is due to the increasing number of researched computer vision methods that can be applied to polyp datasets. Benchmarking of novel methods can provide a direction to the development of automated polyp detection and segmentation tasks. Furthermore, it ensures that the produced results in the community are reproducible and provide a fair comparison of developed methods. In this paper, we benchmark several recent state-of-the-art methods using Kvasir-SEG, an open-access dataset of colonoscopy images for polyp detection, localisation, and segmentation evaluating both method accuracy and speed. Whilst, most methods in literature have competitive performance over accuracy, we show that the proposed ColonSegNet achieved a better trade-off between an average precision of 0.8000 and mean IoU of 0.8100, and the fastest speed of 180 frames per second for the detection and localisation task. Likewise, the proposed ColonSegNet achieved a competitive dice coefficient of 0.8206 and the best average speed of 182.38 frames per second for the segmentation task. Our comprehensive comparison with various state-of-the-art methods reveals the importance of benchmarking the deep learning methods for automated real-time polyp identification and delineations that can potentially transform current clinical practices and minimise miss-detection rates.
Collapse
Affiliation(s)
- Debesh Jha
- SimulaMet0167OsloNorway
- Department of Engineering ScienceBig Data Institute, University of OxfordOxfordOX3 7XFU.K.
| | - Sharib Ali
- Department of Engineering ScienceBig Data Institute, University of OxfordOxfordOX3 7XFU.K.
- Oxford NIHR Biomedical Research CentreOxfordOX4 2PGvU.K.
| | | | - Håvard D. Johansen
- Department of Computer ScienceUiT–The Arctic University of Norway9037TromsøNorway
| | - Dag Johansen
- Department of Computer ScienceUiT–The Arctic University of Norway9037TromsøNorway
| | - Jens Rittscher
- Department of Engineering ScienceBig Data Institute, University of OxfordOxfordOX3 7XFU.K.
- Oxford NIHR Biomedical Research CentreOxfordOX4 2PGvU.K.
| | | | - Pål Halvorsen
- SimulaMet0167OsloNorway
- Department of Computer ScienceOslo Metropolitan University0167OsloNorway
| |
Collapse
|