In order to bring you the best possible user experience, this site uses Javascript. If you are seeing this message, it is likely that the Javascript option in your browser is disabled. For optimal viewing of this site, please ensure that Javascript is enabled for your browser.
Did you know that your browser is out of date? To get the best experience using our website we recommend that you upgrade to a newer version. Learn more.

Machine learning overtakes humans in predicting death or heart attack

Machine algorithm uses 85 variables to calculate risk in individuals Formulas used by humans include a handful of variables and have modest accuracy

Cross-Modality and Multi-Modality Imaging Topics
Cardiac Computed Tomography
Nuclear Imaging
Risk Factors and Prevention


Lisbon, Portugal – 12 May 2019: Machine learning is overtaking humans in predicting death or heart attack. That’s the main message of a study presented today at ICNC 2019 (1).

The International Conference on Nuclear Cardiology and Cardiac CT (ICNC) is co-organised by the American Society of Nuclear Cardiology (ASNC), the European Association of Cardiovascular Imaging (EACVI) of the European Society of Cardiology (ESC), and the European Association of Nuclear Medicine (EANM).

By repeatedly analysing 85 variables in 950 patients with known six-year outcomes, an algorithm “learned” how imaging data interacts. It then identified patterns correlating the variables to death and heart attack with more than 90% accuracy.

Machine learning, the modern bedrock of artificial intelligence (AI), is used every day. Google’s search engine, face recognition on smartphones, self-driving cars, Netflix and Spotify recommendation systems all use machine learning algorithms to adapt to the individual user.

Study author Dr Luis Eduardo Juarez-Orozco, of the Turku PET Centre, Finland, said: “These advances are far beyond what has been done in medicine, where we need to be cautious about how we evaluate risk and outcomes. We have the data but we are not using it to its full potential yet.”

Doctors use risk scores to make treatment decisions. But these scores are based on just a handful of variables and often have modest accuracy in individual patients. Through repetition and adjustment, machine learning can exploit large amounts of data and identify complex patterns that may not be evident to humans.

Dr Juarez-Orozco explained: “Humans have a very hard time thinking further than three dimensions (a cube) or four dimensions (a cube through time). The moment we jump into the fifth dimension we’re lost. Our study shows that very high dimensional patterns are more useful than single dimensional patterns to predict outcomes in individuals and for that we need machine learning.”

The study enrolled 950 patients with chest pain who underwent the centre’s usual protocol to look for coronary artery disease. A coronary computed tomography angiography (CCTA) scan yielded 58 pieces of data on presence of coronary plaque, vessel narrowing, and calcification. Those with scans suggestive of disease underwent a positron emission tomography (PET) scan which produced 17 variables on blood flow. Ten clinical variables were obtained from medical records including sex, age, smoking and diabetes.

During an average six-year follow-up there were 24 heart attacks and 49 deaths from any cause. The 85 variables were entered into a machine learning algorithm called LogitBoost, which analysed them over and over again until it found the best structure to predict who had a heart attack or died.

Dr Juarez-Orozco said: “The algorithm progressively learns from the data and after numerous rounds of analyses, it figures out the high dimensional patterns that should be used to efficiently identify patients who have the event. The result is a score of individual risk.”

The predictive performance using the ten clinical variables alone (similar to current clinical practice) was modest, with an area under the curve (AUC) of 0.65 (where 1.0 is a perfect test and 0.5 is a random result). When PET data were added, AUC increased to 0.69. The predictive performance increased significantly (p=0.005) when CCTA data were added to clinical and PET data, giving an AUC 0.82 and more than 90% accuracy.

Dr Juarez-Orozco said: “Doctors already collect a lot of information about patients – for example those with chest pain. We found that machine learning can integrate these data and accurately predict individual risk. This should allow us to personalise treatment and ultimately lead to better outcomes for patients.”

ENDS

Notes to editor

Authors: ESC Press Office
Tel: +33 (0)4 8987 2499
Mobile: +33 (0) 7 8531 2036

Email: press@escardio.org

Follow us on Twitter @ESCardioNews 

The hashtag for ICNC 2019 is #ICNC2019.

Sources of funding: This work was supported by The Academy of Finland Centre of Excellence on Cardiovascular and Metabolic Disease, Helsinki, Finland and the Finnish Foundation for Cardiovascular Research.

Disclosures: None.

References and notes

(1) The abstract ‘Refining the long-term prognostic value of hybrid PET/CT through machine learning’ will be presented during the Young Investigator Awards session on Sunday 12 May at 09:00 to 10:00 WEST in room Faro.

About the International Conference on Nuclear Cardiology and Cardiac CT (ICNC)

The International Conference on Nuclear Cardiology and Cardiac CT (ICNC) is held every two years and is co-organised by the European Association of Cardiovascular Imaging (EACVI) of the European Society of Cardiology (ESC), the American Society of Nuclear Cardiology (ASNC), and the European Association of Nuclear Medicine (EANM).

About the American Society of Nuclear Cardiology (ASNC)

ASNC, located in Fairfax (Virginia, USA) is committed to excellence in imaging. Representing over 4,500 nuclear cardiologists, radiologists, technologists, scientists and academics all over the world. ASNC is the leader in the development of established standards of care and guidelines in nuclear cardiovascular imaging. ASNC continuously supports the advancement of the profession through education, research, and advocacy.

About the European Association of Cardiovascular Imaging (EACVI)

The European Association of Cardiovascular Imaging (EACVI) - a branch of the ESC - is the world leading network of Cardiovascular Imaging (CVI) experts, gathering four imaging modalities under one entity (Echocardiography, Cardiovascular Magnetic Resonance, Nuclear Cardiology and Cardiac Computed Tomography). Its aim is to promote excellence in clinical diagnosis, research, technical development, and education in cardiovascular imaging. The EACVI welcomes over 11,000 professionals including cardiologists, sonographers, nurses, basic scientists and allied professionals.

About the European Society of Cardiology

The European Society of Cardiology brings together health care professionals from more than 150 countries, working to advance cardiovascular medicine and help people lead longer, healthier lives.

 

About the European Association of Nuclear Medicine (EANM)

The European Association of Nuclear Medicine (EANM) is the largest organisation dedicated to nuclear medicine and multimodality imaging in Europe, with a strong focus on interdisciplinary. This is especially reflected in the educational offers of the European School of Multimodality Imaging and Therapy (ESMIT) and the programme of the EANM Annual Congress.

Information for journalists attending ICNC 2019

ICNC 2019 will be held 12 to 14 May at the Lisbon Congress Centre (CCL) in Lisbon, Portugal. Explore the scientific programme.

  • To register on-site please bring avalid press card or appropriate letter of assignment with proof of three recent published articles (cardiology or health-related, or referring to a previous ESC Event).
  • Press registration is not available to industry or its public relations representatives, event management, marketing or communications representatives.