January 2025 digest

AI in AD and Allergy Research: Enhancing Collaboration with AI Experts

Dear Colleagues, Dear Friends,

Advances in AI over the past two years have significantly transformed the medical landscape, challenging vigorously our professional environment. A thorough review by Lisik et al. [link] provides an insightful overview of AI’s applications in AD and allergy research—a field that lags behind specialties like radiology or pathology where AI’s integration into imaging has advanced more rapidly. The following three items are highlighted among others:

• Characterizing Disease Endotypes/Phenotypes
  AI has the potential to classify subtypes of diseases by clustering diverse data inputs, such as clinical assessments, genetic profiles, and environmental factors. This could address more scientifically the usual question is AD one or several diseases? However, access to well-curated and AI-compatible datasets remains a significant barrier.
• Developing Predictive Models
  Machine learning models trained on historical data can predict future disease trajectories, e.g. such models may provide improved predictions of AD severity based on early-life data and identified risk factors. Integrating data from multiple sources (e.g., genomics, proteomics, metabolomics) will enable AI to deliver more precise predictions regarding the factors influencing AD progression.
• Identifying actionable risk factors and enhancing clinical decision support
  Identifying specific risk factors with strong associations to disease outcomes may lead to personalized interventions and targeted treatment strategies. AI support to healthcare providers by offering real-time predictions and tailored recommendations based on patient data may improve the precision and efficiency of clinical decisions.

The authors underline that despite these promising applications, most published studies rely on simplistic datasets and underutilize advanced algorithms. Additionally, a lack of comprehensive reporting impedes evidence synthesis and critical appraisal. Addressing these challenges requires international collaboration of AD researchers with AD specialists to ensure the availability of high-quality data and access to state-of-the-art algorithms.

Best wishes,

Alain TAÏEB
President ISAD

---

#StayInformed: