Modern Methods and Prospects for Using Artificial Intelligence in Disease Diagnostics: A Narrative Review
Open Access
- 10 July 2024
- journal article
- Published by Futurity Research Publishing in Futurity Medicine
- Vol. 3 (4)
Abstract
Aims: Assess AI methodologies in early disease detection, identify application areas like imaging and prognosis, determine key findings and clinical impacts, explore AI's prospects for accuracy and personalization, and analyze implementation challenges. Methodology: This narrative review examines AI's role in medical diagnostics by analyzing peer-reviewed articles from 2019-2024 from PubMed. A comprehensive search strategy identified 338 publications, which were filtered for relevance, resulting in 10 key studies. Focus areas include AI techniques like machine learning and their applications and challenges in disease diagnosis. Results: Results showed that in 2024, notable AI studies included a large experimental study with 27,558 samples, a comparative study with 1,653 samples, and a cross-sectional study with 20 samples. Previous years saw a retrospective cross-sectional study with 90,000 samples, an observational study with 1,052 samples, and a retrospective study with 219 samples. AI techniques featured EfficientNet-B2, CNNs, VGG-16, and ResNet variants, with transfer learning models like VER-Net and methods such as COVID-DSNet. Advancements highlighted EfficientNet-B2's 97.57% accuracy in malaria detection, VER-Net's superior lung cancer detection, and AI’s effectiveness in diagnosing retinal diseases, heart conditions, diabetic nephropathy, and COVID-19. Challenges included computational demands and dataset needs, with recommendations for model optimization and clinical integration. Scientific Novelty: This review is about the integration of advanced AI techniques in disease diagnostics, showing new algorithms and machine learning models that improve accuracy, speed, and personalized treatment strategies in medical practice. Conclusion: This study has shown that there has been a significant progress in AI-based disease diagnostics, with examples of high performing models such as EfficientNet-B2 and VER-Net. Despite challenges like computational requirement and interpretability of the model, AI has the potential to revolutionize diagnosis.Keywords
This publication has 57 references indexed in Scilit:
- Evaluation of artificial intelligence techniques in disease diagnosis and predictionDiscover Artificial Intelligence, 2023
- Artificial Intelligence and Sustainable DecisionsEuropean Business Organization Law Review, 2022
- RETRACTED ARTICLE: Artificial intelligence and machine learning in precision and genomic medicineMedical Oncology, 2022
- RETRACTED ARTICLE: Artificial intelligence in disease diagnosis: a systematic literature review, synthesizing framework and future research agendaJournal of Ambient Intelligence and Humanized Computing, 2022
- AI-based diagnosis of COVID-19 patients using X-ray scans with stochastic ensemble of CNNsPhysical and Engineering Sciences in Medicine, 2021
- Artificial intelligence in disease diagnostics: A critical review and classification on the current state of research guiding future directionHealth and Technology, 2021
- Artificial intelligence (AI) for medical imaging to combat coronavirus disease (COVID-19): a detailed review with direction for future researchArtificial Intelligence Review, 2021
- Privacy, Security, and Ethical Issues2021
- Artificial Intelligence and Machine Learning in Global Healthcare2020
- Solving the Black Box Problem: A Normative Framework for Explainable Artificial IntelligencePhilosophy & Technology, 2019