AI Nexus for Early Disease Detection and Risk Prediction: Revolutionizing Healthcare through Intelligence
DOI:
https://doi.org/10.47941/jts.3791Keywords:
Artificial Intelligence, Early Disease Detection, Clinical Evidence, Medical Imaging, Risk Prediction, Federated LearningAbstract
Purpose: This paper provides an evidence base and clinical context for the use of Artificial Intelligence (AI) for early disease detection and disease risk prediction.
Methodology: For this evidence-based analysis, a literature-based analytical approach was used. For the analysis of the available publications, a search was conducted in PubMed/MEDLINE, IEEE Xplore, ACM Digital Library, and CrossRef. The search included only English-language, peer-reviewed publications between January 2021 and August 2025. A structured catalog of four key aspects was used to organize the found evidence: 1) a description of the clinical tasks that were analyzed using models; 2) model families; 3) a description of the used validation; 4) outcomes of the models (measures, etc.).
Findings: The analysis provides strong evidence that, in non-inferior or better terms than the current clinical screening workflows, AI can be used to overcome some of the limitations of the manual analysis of a patient’s data, such as subjectivity, delay, and information overload. However, new challenges also arise in interpreting the decisions produced by the AI system (as the computer arrived at them), given the possible presence of bias in the development datasets, and in the infrastructure needed to implement the different applications in clinical settings where they will be used.
Unique Contribution to Theory, Policy and Practice: Health systems must prioritize the external validation of their AI models. Health systems and organizations must develop de-identified datasets that reflect the demographics of the patients and settings where the models will be used. Data sharing within and between organizations must occur using standardized protocols. Clinically used AI must be explainable, and health systems and organizations must require their developers to make their clinical AI tools explainable. Health IT is evolving rapidly, and corresponding regulatory frameworks must keep pace without putting patients at risk.
Downloads
References
[1] E. Topol, Deep Medicine: How Artificial Intelligence Can Make Healthcare Human Again, Basic Books, 2019.
[2] K. Dembrower et al., “Artificial intelligence for breast cancer detection in screening mammography in Sweden: a prospective, population-based, paired-reader, non-inferiority study,” Lancet Digital Health, vol. 5, no. 10, pp. e703–e711, 2023.
[3] R. M. Wolf et al., “Autonomous artificial intelligence increases screening and follow-up for diabetic retinopathy in youth: the ACCESS randomized controlled trial,” Nature Communications, vol. 15, art. 44676, 2024.
[4] A. E. W. Johnson et al., “MIMIC-IV, a freely accessible electronic health record dataset,” Scientific Data, vol. 10, art. 1, 2023.
[5] G. A. Kaissis et al., “Secure, privacy-preserving and federated machine learning in medical imaging,” Nature Machine Intelligence, vol. 2, no. 6, pp. 305–311, 2020.
[6] H. Chen, O. Rudovic, and R. Picard, “Explainable medical imaging AI needs human-centered design,” npj Digital Medicine, vol. 5, art. 97, 2022.
[7] World Health Organization, “The importance of early diagnosis in improving cancer outcomes,” WHO Report, 2019.
[8] P. Jha, “Reducing premature mortality in low-income and middle-income countries,” New England Journal of Medicine, vol. 365, no. 10, pp. 869–874, 2011.
[9] J. W. Wang et al., “Deep learning in medical diagnostics: Current status and future potential,” Nature Medicine, vol. 27, pp. 89–101, 2021.
[10] B. L. White et al., “Predictive analytics for chronic disease prevention,” Health Informatics Journal, vol. 35, pp. 67–81, 2021.
[11] G. McKee, “Delayed diagnosis in chronic diseases: A review of contributing factors,” Journal of Chronic Disease Management, vol. 15, no. 3, pp. 67–78, 2019.
[12] J. S. Rankin et al., “The invention and impact of X-ray imaging in medical diagnostics,” Radiology Journal, vol. 23, pp. 89–102, 2020.
[13] M. W. Brejnebøl et al., “Interobserver Agreement and Performance of Concurrent AI Assistance for Radiographic Evaluation of Knee Osteoarthritis,” Radiology, vol. 312, no. 1, e233341, 2024.
[14] P. W. Chan et al., “NLP in oncology: Applications for clinical decision support,” Journal of Clinical Informatics, vol. 34, pp. 112–125, 2022.
[15] E. H. Shortliffe, “The MYCIN project: A case study in rule-based medical diagnosis,” Artificial Intelligence in Medicine, vol. 6, pp. 45–58, 1977.
[16] S. Pati et al., “Federated learning enables big data for rare cancer boundary detection,” Nature Communications, vol. 13, art. 7346, 2022.
[17] M. W. Brejnebøl et al., “Interobserver Agreement and Performance of Concurrent AI Assistance,” Radiology, vol. 312, no. 1, e233341, 2024.
[18] M. A. Perez et al., “Natural language processing in clinical text analysis,” Journal of Biomedical Informatics, vol. 89, pp. 75–89, 2020.
[19] C. F. Liu et al., “CNN-based lung cancer detection: Performance evaluation,” IEEE Access, vol. 10, pp. 1155–1170, 2022.
[20] A. Linardos et al., “Federated learning for multi-centre imaging diagnostics: A cardiovascular MRI case,” Scientific Reports, vol. 12, art. 3897, 2022.
[21] J. Yang, Z. Obermeyer, and S. Mullainathan, “The limits of ‘fair’ medical imaging AI,” Nature Medicine, vol. 30, pp. 2452–2462, 2024.
[22] J. X. Zhang et al., “RETFound: foundation model for generalist medical AI in retinal imaging,” Nature, vol. 619, pp. 306–313, 2023.
[23] T. Yun et al., “Unsupervised representation learning on high-dimensional clinical data,” Nature Genetics, 2024.
[24] Z. Qin et al., “Automated detection of pulmonary tuberculosis on chest radiographs: a prospective, multicentre study,” Lancet Digital Health, vol. 6, no. 4, pp. e279–e289, 2024.
[25] J. P. Rodriguez et al., “Bias in AI models: Implications for healthcare,” Health AI Review, vol. 10, no. 3, pp. 89–102, 2022.
[26] M. E. S. Sadanandan et al., “Impact of AI-enabled imaging solutions on radiology workflows: a living meta-analysis,” npj Digital Medicine, vol. 7, art. 147, 2024.
[27] K. Lång et al., “Artificial intelligence-supported screen reading versus standard double reading in the MASAI trial,” Lancet Oncology, vol. 24, no. 8, pp. 936–944, 2023.
[28] R. H. Fernandez et al., “NLP applications in clinical diagnostics: From research to practice,” IEEE Journal of Health Informatics, vol. 12, pp. 45–62, 2021.
[29] B. He et al., “Blinded, randomized trial of sonographer versus AI cardiac function assessment,” Nature, vol. 613, pp. 113–120, 2023.
[30] K. L. Tang, S. K. Choi, and S. H. Park, “Artificial intelligence in medical imaging: clinical applications and challenges,” Radiology, vol. 303, no. 3, pp. 430–444, 2022.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 Sadhasivam Mohanadas
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under a Creative Commons Attribution (CC-BY) 4.0 License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
