AI in Autoimmune Care: Transforming Diagnosis and Patient Perspectives
Artificial intelligence is transforming autoimmune disease care—from faster diagnosis and predictive monitoring to personalised treatment and patient engagement. In this interview, Dr. Anindita Santosa explores the evolving role of AI in rheumatology, offering insights into its potential, limitations, and the future of proactive, data-driven autoimmune management.
1. Why are autoimmune diseases a significant and growing health concern today?
Autoimmune diseases affect a large and increasing population. Approximately 8% of Americans live with an autoimmune disease, and the majority of patients are women. Over 100 distinct autoimmune conditions exist, and incidence rates are rising annually. This growing burden has severe impacts on quality of life and public health.
2. What challenges make autoimmune diseases difficult to diagnose accurately and early?
Diagnosis is often difficult because symptoms overlap with other illnesses and evolve gradually. There's usually no single definitive test. Patients often see multiple providers over several years — on average, about four physicians over 4.5 years — leading to delayed intervention and missed early treatment opportunities.
3. In what ways can artificial intelligence aid in the early detection or diagnosis of autoimmune diseases?
AI can analyze complex patterns in data that humans may overlook. Machine learning can scan EHRs and genetics to flag high-risk individuals, achieving up to ten-fold greater accuracy. AI is also being used to interpret medical images — such as detecting subtle joint damage on X-rays — earlier than standard methods.
4. How can AI support the long-term management and monitoring of patients with autoimmune diseases?
AI algorithms monitor symptoms, lab results, and wearable data to predict disease flares. At ACR 2024, one model predicted arthritis flares months in advance with ~89% accuracy. These tools enable proactive care and personalized follow-up, helping clinicians adjust treatment before symptoms worsen.
5. Could you share how you've incorporated AI tools in your own practice or research in rheumatology?
In 2023, we developed a chatbot that provided 24/7 self-care guidance without collecting personal data. Feedback was overwhelmingly positive. We’re now piloting an agentic AI chatbot to support longitudinal follow-up in patients with rheumatic diseases, making proactive care more scalable and accessible.
6. What potential does AI have in improving patient education and engagement for chronic autoimmune conditions?
AI can simplify communication through virtual assistants and chatbots that explain medical concepts and check in regularly. We’ve even used generative AI to convert treatment information into comic strips — making complex medication guidance more engaging, less intimidating, and easier for patients to understand and retain.
7. Which emerging AI-driven technologies do you find most promising for improving autoimmune disease care?
AI-driven blood tests can predict lupus flares 12 weeks in advance. Image analysis tools now flag inflammation or biopsy details to differentiate arthritis types. Wearables paired with AI are detecting subtle inflammation in IBD. These innovations support earlier intervention and more precise management.'
8. What are the limitations or risks of using AI in clinical decision-making for autoimmune diseases?
AI isn’t perfect. Many models are “black boxes” and lack explainability, which affects clinician trust. Bias in training data can skew results. Data privacy is a major concern too. AI must support, not replace, physician judgment — and transparency is essential to its responsible integration.
9. How do you foresee the relationship between AI tools and clinicians evolving? Will AI ever replace some roles, or enhance them?
AI will enhance, not replace, clinicians. It will automate data-heavy tasks, letting doctors focus on empathy and complex decisions. Future systems may summarize notes, flag risks, or suggest diagnoses quietly in the background — helping doctors be more present with patients, without losing the human touch.
10. In what ways can AI enable more personalized treatment plans for patients with autoimmune diseases?
AI helps tailor treatment by analyzing genes, biomarkers, and clinical data to predict drug responses. New models can forecast which patients may not respond to first-line biologics, allowing faster pivot to better alternatives. This reduces trial-and-error and brings precision medicine into daily practice.
11. What barriers exist to implementing AI solutions in routine autoimmune care (for example, data or system challenges)?
Key barriers include inconsistent or siloed medical data, integration issues with hospital IT systems, data privacy concerns, and clinician training gaps. Demonstrating clear cost-benefit will also be important for adoption. Overcoming these challenges requires coordinated efforts across tech, healthcare, and policy sectors.
12. How are patients reacting to the introduction of AI-based solutions in their care?
Most patients are open to AI when benefits are clearly explained. In surveys, the majority found AI symptom-checkers helpful and said they’d use them again. While some express concerns about privacy or missing human interaction, reassurance and transparency usually help build trust.
13. Are there any recent advancements in AI that you are particularly excited about for autoimmune disease diagnosis or treatment?
Recent breakthroughs include AI tools that identify at-risk individuals likely to progress to autoimmune disease, and AI-designed drug candidates for lupus. “Digital twin” models now simulate a patient’s disease and let clinicians test treatments virtually — opening new frontiers for proactive, personalised care.
14. Looking ahead 5 to 10 years, how do you envision the integration of AI transforming the landscape of autoimmune disease management?
In the next decade, AI will likely streamline diagnosis, personalise therapy choices, and monitor patients remotely through smart devices. Alerts will go to doctors before symptoms worsen. Done responsibly, this shift could make autoimmune care more efficient, predictive, and patient-centred.
References
- News-Medical. (2021). Autoimmune diseases affect approximately 8% of the population. News-Medical.net. Retrieved from https://www.news-medical.net/news/20211021/Autoimmune-diseases-affect-approximately-8-of-the-population.aspx
- National Health Council. (n.d.). About Chronic Diseases and Conditions. NationalHealthCouncil.org. Retrieved from https://nationalhealthcouncil.org/focus-areas/chronic-diseases-and-conditions/
- Hausmann, J. S. (2024). Machine Learning Predicts Arthritis Flares with High Accuracy. RheumNow. Retrieved from https://rheumnow.com/content/machine-learning-predicts-arthritis-flares-high-accuracy
- Santosa, A., & Team. (2023). Development of an AI-driven rheumatology chatbot for patient education and self-care. JMIR Formative Research, 7, e44600. https://formative.jmir.org/2023/1/e44600/
- Staff Writer. (2023). How AI is Changing the Doctor-Patient Relationship. Healthcare Asia Magazine. Retrieved from https://healthcareasiamagazine.com/healthcare/in-focus/how-ai-changing-doctor-patient-relationship
- Fierce Biotech. (2023). AI finds new lupus drug candidate in record time. FierceBiotech.com. Retrieved from https://www.fiercebiotech.com/research/ai-finds-new-lupus-drug-candidate-record-time
- EMJ. (2023). AI in Rheumatology: From Imaging to Biopsy Diagnostics. EMJReviews.com. Retrieved from https://www.emjreviews.com/rheumatology/article/ai-in-rheumatology-from-imaging-to-biopsy-diagnostics/
