1
|
Lo CM, Wang CC, Hung PH. Interactive content-based image retrieval with deep learning for CT abdominal organ recognition. Phys Med Biol 2024; 69:045004. [PMID: 38232396 DOI: 10.1088/1361-6560/ad1f86] [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: 10/11/2023] [Accepted: 01/17/2024] [Indexed: 01/19/2024]
Abstract
Objective.Recognizing the most relevant seven organs in an abdominal computed tomography (CT) slice requires sophisticated knowledge. This study proposed automatically extracting relevant features and applying them in a content-based image retrieval (CBIR) system to provide similar evidence for clinical use.Approach.A total of 2827 abdominal CT slices, including 638 liver, 450 stomach, 229 pancreas, 442 spleen, 362 right kidney, 424 left kidney and 282 gallbladder tissues, were collected to evaluate the proposed CBIR in the present study. Upon fine-tuning, high-level features used to automatically interpret the differences among the seven organs were extracted via deep learning architectures, including DenseNet, Vision Transformer (ViT), and Swin Transformer v2 (SwinViT). Three images with different annotations were employed in the classification and query.Main results.The resulting performances included the classification accuracy (94%-99%) and retrieval result (0.98-0.99). Considering global features and multiple resolutions, SwinViT performed better than ViT. ViT also benefited from a better receptive field to outperform DenseNet. Additionally, the use of hole images can obtain almost perfect results regardless of which deep learning architectures are used.Significance.The experiment showed that using pretrained deep learning architectures and fine-tuning with enough data can achieve successful recognition of seven abdominal organs. The CBIR system can provide more convincing evidence for recognizing abdominal organs via similarity measurements, which could lead to additional possibilities in clinical practice.
Collapse
Affiliation(s)
- Chung-Ming Lo
- Graduate Institute of Library, Information and Archival Studies, National Chengchi University, Taipei, Taiwan
| | - Chi-Cheng Wang
- Department of Radiology, Mackay Memorial Hospital, Taipei, Taiwan
| | - Peng-Hsiang Hung
- Department of Radiology, Mackay Memorial Hospital, Taipei, Taiwan
| |
Collapse
|
2
|
Lo CM, Jiang JK, Lin CC. Detecting microsatellite instability in colorectal cancer using Transformer-based colonoscopy image classification and retrieval. PLoS One 2024; 19:e0292277. [PMID: 38271352 PMCID: PMC10810505 DOI: 10.1371/journal.pone.0292277] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 03/25/2023] [Accepted: 09/15/2023] [Indexed: 01/27/2024] Open
Abstract
Colorectal cancer (CRC) is a major global health concern, with microsatellite instability-high (MSI-H) being a defining characteristic of hereditary nonpolyposis colorectal cancer syndrome and affecting 15% of sporadic CRCs. Tumors with MSI-H have unique features and better prognosis compared to MSI-L and microsatellite stable (MSS) tumors. This study proposed establishing a MSI prediction model using more available and low-cost colonoscopy images instead of histopathology. The experiment utilized a database of 427 MSI-H and 1590 MSS colonoscopy images and vision Transformer (ViT) with different feature training approaches to establish the MSI prediction model. The accuracy of combining pre-trained ViT features was 84% with an area under the receiver operating characteristic curve of 0.86, which was better than that of DenseNet201 (80%, 0.80) in the experiment with support vector machine. The content-based image retrieval (CBIR) approach showed that ViT features can obtain a mean average precision of 0.81 compared to 0.79 of DenseNet201. ViT reduced the issues that occur in convolutional neural networks, including limited receptive field and gradient disappearance, and may be better at interpreting diagnostic information around tumors and surrounding tissues. By using CBIR, the presentation of similar images with the same MSI status would provide more convincing deep learning suggestions for clinical use.
Collapse
Affiliation(s)
- Chung-Ming Lo
- Graduate Institute of Library, Information and Archival Studies, National Chengchi University, Taipei, Taiwan
| | - Jeng-Kai Jiang
- Department of Surgery, Division of Colon and Rectal Surgery, Taipei Veterans General Hospital, Taipei, Taiwan
- Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Chun-Chi Lin
- Department of Surgery, Division of Colon and Rectal Surgery, Taipei Veterans General Hospital, Taipei, Taiwan
- Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| |
Collapse
|
3
|
Lo CM, Hung PH. Predictive stroke risk model with vision transformer-based Doppler features. Med Phys 2024; 51:126-138. [PMID: 38043124 DOI: 10.1002/mp.16861] [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: 04/08/2023] [Revised: 11/17/2023] [Accepted: 11/17/2023] [Indexed: 12/05/2023] Open
Abstract
BACKGROUND Acute stroke is the leading cause of death and disability globally, with an estimated 16 million cases each year. The progression of carotid stenosis reduces blood flow to the intracranial vasculature, causing stroke. Early recognition of ischemic stroke is crucial for disease treatment and management. PURPOSE A computer-aided diagnosis (CAD) system was proposed in this study to rapidly evaluate ischemic stroke in carotid color Doppler (CCD). METHODS Based on the ground truth from the clinical examination report, the vision transformer (ViT) features extracted from all CCD images (513 stroke and 458 normal images) were combined in machine learning classifiers to generate the likelihood of ischemic stroke for each image. The pretrained weights from ImageNet reduced the time-consuming training process. The accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve were calculated to evaluate the stroke prediction model. The chi-square test, DeLong test, and Bonferroni correction for multiple comparisons were applied to deal with the type-I error. Only p values equal to or less than 0.00125 were considered to be statistically significant. RESULTS The proposed CAD system achieved an accuracy of 89%, a sensitivity of 94%, a specificity of 84%, and an area under the receiver operating characteristic curve of 0.95, outperforming the convolutional neural networks AlexNet (82%, p < 0.001), Inception-v3 (78%, p < 0.001), ResNet101 (84%, p < 0.001), and DenseNet201 (85%, p < 0.01). The computational time in model training was only 30 s, which would be efficient and practical in clinical use. CONCLUSIONS The experiment shows the promising use of CCD images in stroke estimation. Using the pretrained ViT architecture, the image features can be automatically and efficiently generated without human intervention. The proposed CAD system provides a rapid and reliable suggestion for diagnosing ischemic stroke.
Collapse
Affiliation(s)
- Chung-Ming Lo
- Graduate Institute of Library, Information and Archival Studies, National Chengchi University, Taipei, Taiwan
| | - Peng-Hsiang Hung
- Department of Radiology, Mackay Memorial Hospital, Taipei, Taiwan
| |
Collapse
|
4
|
Lo CM, Yang YW, Lin JK, Lin TC, Chen WS, Yang SH, Chang SC, Wang HS, Lan YT, Lin HH, Huang SC, Cheng HH, Jiang JK, Lin CC. Modeling the survival of colorectal cancer patients based on colonoscopic features in a feature ensemble vision transformer. Comput Med Imaging Graph 2023; 107:102242. [PMID: 37172354 DOI: 10.1016/j.compmedimag.2023.102242] [Citation(s) in RCA: 3] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/21/2023] [Revised: 05/05/2023] [Accepted: 05/07/2023] [Indexed: 05/14/2023]
Abstract
The prognosis of patients with colorectal cancer (CRC) mostly relies on the classic tumor node metastasis (TNM) staging classification. A more accurate and convenient prediction model would provide a better prognosis and assist in treatment. From May 2014 to December 2017, patients who underwent an operation for CRC were enrolled. The proposed feature ensemble vision transformer (FEViT) used ensemble classifiers to benefit the combinations of relevant colonoscopy features from the pretrained vision transformer and clinical features, including sex, age, family history of CRC, and tumor location, to establish the prognostic model. A total of 1729 colonoscopy images were enrolled in the current retrospective study. For the prediction of patient survival, FEViT achieved an accuracy of 94 % with an area under the receiver operating characteristic curve of 0.93, which was better than the TNM staging classification (90 %, 0.83) in the experiment. FEViT reduced the limited receptive field and gradient disappearance in the conventional convolutional neural network and was a relatively effective and efficient procedure. The promising accuracy of FEViT in modeling survival makes the prognosis of CRC patients more predictable and practical.
Collapse
Affiliation(s)
- Chung-Ming Lo
- Graduate Institute of Library, Information and Archival Studies, National Chengchi University, Taipei, Taiwan
| | - Yi-Wen Yang
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Jen-Kou Lin
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Tzu-Chen Lin
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Wei-Shone Chen
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Shung-Haur Yang
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan; Department of Surgery, National Yang Ming Chiao Tung University Hospital, Yilan, Taiwan
| | - Shih-Ching Chang
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Huann-Sheng Wang
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Yuan-Tzu Lan
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Hung-Hsin Lin
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Sheng-Chieh Huang
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Hou-Hsuan Cheng
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Jeng-Kai Jiang
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan
| | - Chun-Chi Lin
- Division of Colon and Rectal Surgery, Department of Surgery, Taipei Veterans General Hospital, Taipei, Taiwan; Department of Surgery, School of Medicine, National Yang Ming Chiao Tung University, Taipei, Taiwan.
| |
Collapse
|
5
|
Lo CM, Lai KL. Deep learning-based assessment of knee septic arthritis using transformer features in sonographic modalities. COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE 2023; 237:107575. [PMID: 37148635 DOI: 10.1016/j.cmpb.2023.107575] [Citation(s) in RCA: 5] [Impact Index Per Article: 5.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 12/23/2022] [Revised: 02/21/2023] [Accepted: 05/02/2023] [Indexed: 05/08/2023]
Abstract
PURPOSE Septic arthritis is an infectious disease. Conventionally, the diagnosis of septic arthritis can only be based on the identification of causal pathogens taken from synovial fluid, synovium or blood samples. However, the cultures require several days for the isolation of pathogens. A rapid assessment performed through computer-aided diagnosis (CAD) would bring timely treatment. METHODS A total of 214 non-septic arthritis and 64 septic arthritis images generated by gray-scale (GS) and Power Doppler (PD) ultrasound modalities were collected for the experiment. A deep learning-based vision transformer (ViT) with pre-trained parameters were used for image feature extraction. The extracted features were then combined in machine learning classifiers with ten-fold cross validation in order to evaluate the abilities of septic arthritis classification. RESULTS Using a support vector machine, GS and PD features can achieve an accuracy rate of 86% and 91%, with the area under the receiver operating characteristic curves (AUCs) being 0.90 and 0.92, respectively. The best accuracy (92%) and best AUC (0.92) was obtained by combining both feature sets. CONCLUSIONS This is the first CAD system based on a deep learning approach for the diagnosis of septic arthritis as seen on knee ultrasound images. Using pre-trained ViT, both the accuracy and computation costs improved more than they had through convolutional neural networks. Additionally, automatically combining GS and PD generates a higher accuracy to better assist the physician's observations, thus providing a timely evaluation of septic arthritis.
Collapse
Affiliation(s)
- Chung-Ming Lo
- Graduate Institute of Library, Information and Archival Studies, National Chengchi University, Taipei, Taiwan
| | - Kuo-Lung Lai
- Division of Allergy, Immunology and Rheumatology, Department of Internal Medicine, Taichung Veterans General Hospital, Taichung, Taiwan; Department of Post-Baccalaureate Medicine, College of Medicine, National Chung Hsing University, Taichung, Taiwan.
| |
Collapse
|
6
|
Miceli G, Basso MG, Rizzo G, Pintus C, Cocciola E, Pennacchio AR, Tuttolomondo A. Artificial Intelligence in Acute Ischemic Stroke Subtypes According to Toast Classification: A Comprehensive Narrative Review. Biomedicines 2023; 11:biomedicines11041138. [PMID: 37189756 DOI: 10.3390/biomedicines11041138] [Citation(s) in RCA: 5] [Impact Index Per Article: 5.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/27/2023] [Revised: 03/29/2023] [Accepted: 04/06/2023] [Indexed: 05/17/2023] Open
Abstract
The correct recognition of the etiology of ischemic stroke (IS) allows tempestive interventions in therapy with the aim of treating the cause and preventing a new cerebral ischemic event. Nevertheless, the identification of the cause is often challenging and is based on clinical features and data obtained by imaging techniques and other diagnostic exams. TOAST classification system describes the different etiologies of ischemic stroke and includes five subtypes: LAAS (large-artery atherosclerosis), CEI (cardio embolism), SVD (small vessel disease), ODE (stroke of other determined etiology), and UDE (stroke of undetermined etiology). AI models, providing computational methodologies for quantitative and objective evaluations, seem to increase the sensitivity of main IS causes, such as tomographic diagnosis of carotid stenosis, electrocardiographic recognition of atrial fibrillation, and identification of small vessel disease in magnetic resonance images. The aim of this review is to provide overall knowledge about the most effective AI models used in the differential diagnosis of ischemic stroke etiology according to the TOAST classification. According to our results, AI has proven to be a useful tool for identifying predictive factors capable of subtyping acute stroke patients in large heterogeneous populations and, in particular, clarifying the etiology of UDE IS especially detecting cardioembolic sources.
Collapse
Affiliation(s)
- Giuseppe Miceli
- Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties (ProMISE), Università Degli Studi di Palermo, Piazza Delle Cliniche 2, 90127 Palermo, Italy
- Internal Medicine and Stroke Care Ward, University Hospital, Policlinico "P. Giaccone", 90141 Palermo, Italy
| | - Maria Grazia Basso
- Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties (ProMISE), Università Degli Studi di Palermo, Piazza Delle Cliniche 2, 90127 Palermo, Italy
- Internal Medicine and Stroke Care Ward, University Hospital, Policlinico "P. Giaccone", 90141 Palermo, Italy
| | - Giuliana Rizzo
- Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties (ProMISE), Università Degli Studi di Palermo, Piazza Delle Cliniche 2, 90127 Palermo, Italy
- Internal Medicine and Stroke Care Ward, University Hospital, Policlinico "P. Giaccone", 90141 Palermo, Italy
| | - Chiara Pintus
- Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties (ProMISE), Università Degli Studi di Palermo, Piazza Delle Cliniche 2, 90127 Palermo, Italy
- Internal Medicine and Stroke Care Ward, University Hospital, Policlinico "P. Giaccone", 90141 Palermo, Italy
| | - Elena Cocciola
- Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties (ProMISE), Università Degli Studi di Palermo, Piazza Delle Cliniche 2, 90127 Palermo, Italy
- Internal Medicine and Stroke Care Ward, University Hospital, Policlinico "P. Giaccone", 90141 Palermo, Italy
| | - Andrea Roberta Pennacchio
- Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties (ProMISE), Università Degli Studi di Palermo, Piazza Delle Cliniche 2, 90127 Palermo, Italy
- Internal Medicine and Stroke Care Ward, University Hospital, Policlinico "P. Giaccone", 90141 Palermo, Italy
| | - Antonino Tuttolomondo
- Department of Health Promotion, Mother and Child Care, Internal Medicine and Medical Specialties (ProMISE), Università Degli Studi di Palermo, Piazza Delle Cliniche 2, 90127 Palermo, Italy
- Internal Medicine and Stroke Care Ward, University Hospital, Policlinico "P. Giaccone", 90141 Palermo, Italy
| |
Collapse
|