Respiratory Diseases
Artificial Intelligence
Positive skin or blood allergy tests are more likely to indicate asthma than negative results. Despite this, allergy testing is rarely used in medical practice to diagnose or manage asthma. This is partly because almost 25% of people with positive allergy tests do not have any symptoms.
A new approach, called 'component-resolved diagnostics', measures allergy antibodies to over 100 specific parts of pollen, dust mites, or foods that the immune system reacts to ('allergen components'). This method can produce highly detailed information, but there remains a challenge in meaningfully translating these complex results for doctors and patients.
Dr Sara Fontanella at Imperial College London has carried out research revealing that interconnections across many of these molecules are effective predictors of asthma, capable of distinguishing between mild and severe disease. The team found far more complex molecule systems or arrangements in patients with asthma than those without.
In this project, Dr Fontanella will use artificial intelligence tools to help doctors and patients identify clinically important allergy profiles. This will support more accurate asthma diagnoses and help clinicians to more effectively predict children at-risk of developing the condition in the future.