Keratoconus disease classification with multimodel fusion and vision transformer: a pretrained model approach DOI Creative Commons

Shokufeh Yaraghi,

Toktam Khatibi

BMJ Open Ophthalmology, Journal Year: 2024, Volume and Issue: 9(1), P. e001589 - e001589

Published: April 1, 2024

Objective Our objective is to develop a novel keratoconus image classification system that leverages multiple pretrained models and transformer architecture achieve state-of-the-art performance in detecting keratoconus. Methods analysis Three were used extract features from the input images. These have been trained on large datasets demonstrated strong various computer vision tasks. The extracted three fused using feature fusion technique. This aimed combine strengths of each model capture more comprehensive representation then as transformer, powerful has shown excellent learnt classify images either indicative or not. proposed method was applied Shahroud Cohort Eye collection detection dataset. evaluated standard evaluation metrics such accuracy, precision, recall F1 score. Results research results achieved higher accuracy compared with individually. Conclusion findings this study suggest approach can significantly improve for detection. serve an effective decision support alongside physicians, aiding diagnosis potentially reducing need invasive procedures corneal transplantation severe cases.

Language: Английский

A comparison review of transfer learning and self-supervised learning: Definitions, applications, advantages and limitations DOI Creative Commons
Zehui Zhao, Laith Alzubaidi, Jinglan Zhang

et al.

Expert Systems with Applications, Journal Year: 2023, Volume and Issue: 242, P. 122807 - 122807

Published: Dec. 2, 2023

Deep learning has emerged as a powerful tool in various domains, revolutionising machine research. However, one persistent challenge is the scarcity of labelled training data, which hampers performance and generalisation deep models. To address this limitation, researchers have developed innovative methods to overcome data enhance model capabilities. Two prevalent techniques that gained significant attention are transfer self-supervised learning. Transfer leverages knowledge learned from pre-training on large-scale dataset, such ImageNet, applies it target task with limited data. This approach allows models benefit representations effectively new tasks, resulting improved generalisation. On other hand, focuses using pretext tasks do not require manual annotation, allowing them learn valuable large amounts unlabelled These can then be fine-tuned for downstream mitigating need extensive In recent years, found applications fields, including medical image processing, video recognition, natural language processing. approaches demonstrated remarkable achievements, enabling breakthroughs areas disease diagnosis, object understanding. while these offer numerous advantages, they also limitations. For example, may face domain mismatch issues between requires careful design ensure meaningful representations. review paper explores fields within past three years. It delves into advantages limitations each approach, assesses employing techniques, identifies potential directions future By providing comprehensive current methods, article offers guidance selecting best technique specific issue.

Language: Английский

Citations

109
Performance of ChatGPT in Diagnosis of Corneal Eye Diseases DOI
Mohammad Delsoz, Yeganeh Madadi, Hina Raja

et al.

Cornea, Journal Year: 2024, Volume and Issue: 43(5), P. 664 - 670

Published: Feb. 23, 2024

Purpose: The aim of this study was to assess the capabilities ChatGPT-4.0 and ChatGPT-3.5 for diagnosing corneal eye diseases based on case reports compare with human experts. Methods: We randomly selected 20 cases including infections, dystrophies, degenerations from a publicly accessible online database University Iowa. then input text each description into asked provisional diagnosis. finally evaluated responses correct diagnoses, compared them diagnoses made by 3 specialists (human experts), interobserver agreements. Results: diagnosis accuracy 85% (17 cases), whereas 60% (12 20). 100% (20 cases, P = 0.23, 0.0033), 90% (18 0.99, 0.6), respectively. agreement between 65% (13 80% (16 75% (15 However, cases). Conclusions: in patients various conditions markedly improved than promising potential clinical integration. A balanced approach that combines artificial intelligence–generated insights expertise holds key role unveiling its full care.

Language: Английский

Citations

37
Explainable AI-driven IoMT fusion: Unravelling techniques, opportunities, and challenges with Explainable AI in healthcare DOI
Niyaz Ahmad Wani, Ravinder Kumar,

­ Mamta

et al.

Information Fusion, Journal Year: 2024, Volume and Issue: 110, P. 102472 - 102472

Published: May 16, 2024

Language: Английский

Citations

30
Artificial intelligence and multimodal data fusion for smart healthcare: topic modeling and bibliometrics DOI Creative Commons
Xieling Chen, Haoran Xie, Xiaohui Tao

et al.

Artificial Intelligence Review, Journal Year: 2024, Volume and Issue: 57(4)

Published: March 15, 2024

Abstract Advancements in artificial intelligence (AI) have driven extensive research into developing diverse multimodal data analysis approaches for smart healthcare. There is a scarcity of large-scale literature this field based on quantitative approaches. This study performed bibliometric and topic modeling examination 683 articles from 2002 to 2022, focusing topics trends, journals, countries/regions, institutions, authors, scientific collaborations. Results showed that, firstly, the number has grown 1 220 with majority being published interdisciplinary journals that link healthcare medical information technology AI. Secondly, significant rise quantity can be attributed increasing contribution scholars non-English speaking countries/regions noteworthy contributions made by authors USA India. Thirdly, researchers show high interest issues, especially, cross-modality magnetic resonance imaging (MRI) brain tumor analysis, cancer prognosis through multi-dimensional AI-assisted diagnostics personalization healthcare, each experiencing increase interest. an emerging trend towards issues such as applying generative adversarial networks contrastive learning image fusion synthesis utilizing combined spatiotemporal resolution functional MRI electroencephalogram data-centric manner. valuable enhancing researchers’ practitioners’ understanding present focal points upcoming trajectories AI-powered analysis.

Language: Английский

Citations

24
Trustworthy deep learning framework for the detection of abnormalities in X-ray shoulder images DOI Creative Commons
Laith Alzubaidi, Asma Salhi, Mohammed A. Fadhel

et al.

PLoS ONE, Journal Year: 2024, Volume and Issue: 19(3), P. e0299545 - e0299545

Published: March 11, 2024

Musculoskeletal conditions affect an estimated 1.7 billion people worldwide, causing intense pain and disability. These lead to 30 million emergency room visits yearly, the numbers are only increasing. However, diagnosing musculoskeletal issues can be challenging, especially in emergencies where quick decisions necessary. Deep learning (DL) has shown promise various medical applications. previous methods had poor performance a lack of transparency detecting shoulder abnormalities on X-ray images due training data better representation features. This often resulted overfitting, generalisation, potential bias decision-making. To address these issues, new trustworthy DL framework been proposed detect (such as fractures, deformities, arthritis) using images. The consists two parts: same-domain transfer (TL) mitigate imageNet mismatch feature fusion reduce error rates improve trust final result. Same-domain TL involves pre-trained models large number labelled from body parts fine-tuning them target dataset Feature combines extracted features with seven train several ML classifiers. achieved excellent accuracy rate 99.2%, F1 Score Cohen’s kappa 98.5%. Furthermore, results was validated three visualisation tools, including gradient-based class activation heat map (Grad CAM), visualisation, locally interpretable model-independent explanations (LIME). outperformed orthopaedic surgeons invited classify test set, who obtained average 79.1%. proven effective robust, improving generalisation increasing results.

Language: Английский

Citations

17
Towards Risk-Free Trustworthy Artificial Intelligence: Significance and Requirements DOI Creative Commons
Laith Alzubaidi, Aiman Al-Sabaawi, Jinshuai Bai

et al.

International Journal of Intelligent Systems, Journal Year: 2023, Volume and Issue: 2023, P. 1 - 41

Published: Oct. 26, 2023

Given the tremendous potential and influence of artificial intelligence (AI) algorithmic decision-making (DM), these systems have found wide-ranging applications across diverse fields, including education, business, healthcare industries, government, justice sectors. While AI DM offer significant benefits, they also carry risk unfavourable outcomes for users society. As a result, ensuring safety, reliability, trustworthiness becomes crucial. This article aims to provide comprehensive review synergy between DM, focussing on importance trustworthiness. The addresses following four key questions, guiding readers towards deeper understanding this topic: (i) why do we need trustworthy AI? (ii) what are requirements In line with second question, that establish been explained, explainability, accountability, robustness, fairness, acceptance AI, privacy, accuracy, reproducibility, human agency, oversight. (iii) how can data? (iv) priorities in terms challenging applications? Regarding last six different discussed, environmental science, 5G-based IoT networks, robotics architecture, engineering construction, financial technology, healthcare. emphasises address before their deployment order achieve goal good. An example is provided demonstrates be employed eliminate bias resources management systems. insights recommendations presented paper will serve as valuable guide researchers seeking applications.

Language: Английский

Citations

39
MEFF – A model ensemble feature fusion approach for tackling adversarial attacks in medical imaging DOI Creative Commons
Laith Alzubaidi, Khamael Al-Dulaimi,

Huda Abdul-Hussain Obeed

et al.

Intelligent Systems with Applications, Journal Year: 2024, Volume and Issue: 22, P. 200355 - 200355

Published: March 16, 2024

Adversarial attacks pose a significant threat to deep learning models, specifically medical images, as they can mislead models into making inaccurate predictions by introducing subtle distortions the input data that are often imperceptible humans. Although adversarial training is common technique used mitigate these on it lacks flexibility address new attack methods and effectively improve feature representation. This paper introduces novel Model Ensemble Feature Fusion (MEFF) designed combat in image applications. The proposed model employs fusion combining features extracted from different DL then trains Machine Learning classifiers using fused features. It uses concatenation method merge features, forming more comprehensive representation enhancing model's ability classify classes accurately. Our experimental study has performed evaluation of MEFF, considering several challenging scenarios, including 2D 3D greyscale colour binary classification, multi-label classification. reported results demonstrate robustness MEFF against types across six distinct A key advantage its capability incorporate wide range without need train scratch. Therefore, contributes developing diverse robust defense strategy. More importantly, leveraging ensemble modeling, enhances resilience face attacks, paving way for improved reliability analysis.

Language: Английский

Citations

11
Deep Transfer Learning with Enhanced Feature Fusion for Detection of Abnormalities in X-ray Images DOI Open Access
Zaenab Alammar, Laith Alzubaidi, Jinglan Zhang

et al.

Cancers, Journal Year: 2023, Volume and Issue: 15(15), P. 4007 - 4007

Published: Aug. 7, 2023

Medical image classification poses significant challenges in real-world scenarios. One major obstacle is the scarcity of labelled training data, which hampers performance image-classification algorithms and generalisation. Gathering sufficient data often difficult time-consuming medical domain, but deep learning (DL) has shown remarkable performance, although it typically requires a large amount to achieve optimal results. Transfer (TL) played pivotal role reducing time, cost, need for number images. This paper presents novel TL approach that aims overcome limitations disadvantages are characteristic an ImageNet dataset, belongs different domain. Our proposed involves DL models on numerous images similar target dataset. These were then fine-tuned using small set annotated leverage knowledge gained from pre-training phase. We specifically focused X-ray imaging scenarios involve humerus wrist musculoskeletal radiographs (MURA) Both these tasks face regarding accurate classification. The trained with used extract features subsequently fused train several machine (ML) classifiers. combined diverse represent various relevant characteristics comprehensive way. Through extensive evaluation, our feature-fusion ML classifiers achieved For humerus, we accuracy 87.85%, F1-score 87.63%, Cohen's Kappa coefficient 75.69%. classification, 85.58%, 82.70%, 70.46%. results demonstrated outperformed those TL. employed visualisation techniques further validate findings, including gradient-based class activation heat map (Grad-CAM) locally interpretable model-independent explanations (LIME). tools provided additional evidence support superior compared Furthermore, exhibited greater robustness experiments Importantly, technique not limited specific tasks. They can be applied applications, thus extending their utility potential impact. To demonstrate concept reusability, computed tomography (CT) case was adopted. obtained method showed improvements.

Language: Английский

Citations

21
Performance of ChatGPT in Diagnosis of Corneal Eye Diseases DOI Creative Commons
Mohammad Delsoz, Yeganeh Madadi, Wuqaas M. Munir

et al.

medRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Aug. 28, 2023

ABSTRACT Introduction Assessing the capabilities of ChatGPT-4.0 and ChatGPT-3.5 for diagnosing corneal eye diseases based on case reports compare with human experts. Methods We randomly selected 20 cases including infections, dystrophies, degenerations, injuries from a publicly accessible online database University Iowa. then input text each description into ChatGPT3.5 asked provisional diagnosis. finally evaluated responses correct diagnoses compared three cornea specialists (Human experts) interobserver agreements. Results The diagnosis accuracy was 85% (17 out cases) while 60% (12 20). were 100% (20 cases), 90% (18 respectively. agreement between 65% (13 80% (16 75% (15 However, cases). Conclusions in patients various conditions markedly improved than promising potential clinical integration. Key summary points - aim this work to evaluate performance ChatGPT-4 providing different descriptions them specialists. significantly better specific cases. 85%, 80%, 75%,

Language: Английский

Citations

18
A new morphological classification of keratoconus using few-shot learning in candidates for intrastromal corneal ring implants DOI
Zhila Agharezaei, Mohammad Shirshekar,

Reza Firouzi

et al.

Biomedical Signal Processing and Control, Journal Year: 2025, Volume and Issue: 106, P. 107664 - 107664

Published: Feb. 17, 2025

Language: Английский

Citations

0