Skip to main content
Research project


Project overview

Acute Aortic Syndrome (AAS) refers to a group of life-threatening conditions that affect the aorta, the main artery that carries blood from the heart to the rest of the body. AAS encompasses three primary conditions: aortic dissection, intramural hematoma, and penetrating aortic ulcer. Prompt recognition and accurate diagnosis of AAS are crucial due to the high risk of catastrophic complications, including aortic rupture and sudden death.

Diagnosing AAS poses significant challenges due to its diverse clinical presentations which are not specific and may resemble other common conditions, for example, patients may present with severe chest or back pain, similar to a heart attack or musculoskeletal issues. AAS is a rare condition with only 5-30 patients per million, compared to heart attacks which occur in 4,000 patients per million (1).

The complexity of AAS can lead to diagnostic pitfalls, with potential misinterpretations or incomplete evaluations, with up to 40% of cases experiencing a delay in diagnosis (2). Clinicians must have a high index of suspicion and consider AAS in the differential diagnosis, particularly in patients with risk factors such as hypertension, connective tissue disorders, or a family history of aortic disease.

Artificial Intelligence (AI) and Machine Learning (ML) are promising technologies that can analyse large volumes of information and offer opportunities to enhance medical diagnoses and improve patient outcomes. Artificial Intelligence refers to the simulation of human intelligence in machines, enabling them to perform tasks that typically require human intelligence, such as reasoning, learning, and problem-solving. Machine Learning, a subset of AI, focuses on algorithms and statistical models that allow computers to learn from and make predictions or decisions based on data without explicit programming. In the case of Acute Aortic Syndrome (AAS), the complex nature of the condition and the challenges in accurate diagnosis make AI and ML valuable tools.

Using already available historical information on over 6000 patients presenting with AAS-like symptoms and also patients with confirmed AAS from two existing AAS studies, we aim to create an AI/ML decision support system to assist healthcare professionals in diagnosing AAS.

This system would provide real-time feedback and suggestions based on patient data aiding physicians in making informed rapid diagnostic decisions. Using an AI-system could augment healthcare professionals' diagnostic capabilities by detecting subtle patterns or features that might be missed by human observation alone. This will lead to improved accuracy in diagnosing AAS and reduce the likelihood of missed or incorrect diagnoses. It could be rolled out to all emergency departments as a screening tool for AAS and remove the ambiguity, clinical biases and potential variations in medical assessment based on the initial healthcare professional evaluating the patient. Instead, the first review is by a well-trained AI system to determine the personalised risk assessment, tailored treatment plans, and follow-up strategies ultimately optimising patient care and outcomes.


Other researchers

Dr Ganesh Vigneswaran MBBS, BSc, PhD, FRCR

NIHR Clinical Lecturer
to top