ARTIFICIAL intelligence (AI) has demonstrated high accuracy in identifying the causes of seasonal allergic rhinitis (SAR), offering a potential step forward in personalised allergy care.
SAR, commonly triggered by pollen exposure, remains challenging to diagnose accurately, particularly in temperate regions where patients are often sensitised to multiple overlapping allergens. Precise identification of the causative allergen is essential to guide allergen-specific immunotherapy (AIT), the only disease-modifying treatment available. However, traditional diagnostic approaches rely heavily on clinician interpretation, which can be complex and variable.
In the @IT-2020 project, researchers developed a modular Clinical Decision Support System (CDSS) enhanced by machine learning to improve the etiologic diagnosis of SAR. The system integrated three progressive diagnostic layers: clinical history with skin prick testing, molecular immunoglobulin E testing, and an electronic clinical and environmental diary. These components were used to train AI models against expert-defined “gold standard” diagnoses.
Seasonal Allergic Rhinitis AI Diagnosis Shows High Accuracy
The resulting models demonstrated excellent diagnostic performance, achieving an area under the receiver operating characteristic curve above 95%. Notably, the system maintained high accuracy across different patient populations, including external validation in a separate cohort from a different geographical region.
The AI-driven CDSS also showed strong adaptability. Its performance varied appropriately with patient complexity, while interpretability analyses confirmed that both clinical features and sensitisation patterns contributed meaningfully to diagnostic decisions. Importantly, the system reproduced expert AIT prescriptions with considerable reliability, suggesting real-world clinical relevance.
AI Diagnosis Outperforms Clinicians
In a direct comparison, the AI models outperformed 24 clinicians when diagnosing a subset of patients, highlighting the potential of augmented decision-making in allergy practice. Furthermore, reducing the monitoring period to 45 days did not significantly compromise accuracy, suggesting the approach could be practical in routine care settings.
Despite these promising findings, the study was a proof-of-concept and relied on relatively small cohorts. Further independent validation and prospective trials are needed to determine whether AI-guided diagnosis can improve long-term disease control and patient outcomes.
If confirmed, this approach could transform SAR management by enabling more precise, efficient, and standardised diagnosis, ultimately supporting personalised treatment strategies and improving quality of life for patients with allergic disease.
Reference
Matricardi PM et al. Etiologic diagnosis of seasonal allergic rhinitis supported by artificial intelligence: the @IT-2020 project. J Allergy Clin Immunol. 2026;DOI:10.1016/j.jaci.2026.03.011.
Featured image: da on Adobe Stock
