Posted in | Medical Robotics

Study Uses Machine-Learning Algorithm to Predict Sudden Death in Heart Failure Patients

According to a work presented on May 13th at ICNC 2019, artificial intelligence (AI) has exhibited the potential to choose heart failure patients for costly treatments to avoid lethal arrhythmias.

It is the first study to predict unexpected death in heart failure patients using a machine-learning algorithm.

The International Conference on Nuclear Cardiology and Cardiac CT (ICNC) is organized together with 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).

About 1%–2% of adults in developed countries have heart failure, which is a clinical syndrome characterized by ankle swelling, fatigue, and breathlessness. A high proportion of deaths in these patients, particularly those with minor symptoms, happen all of a sudden because of ventricular arrhythmias.

Cardiac resynchronization therapy with a pacemaker and defibrillator (CRT-D) or implantable cardioverter defibrillators (ICDs) are suggested for some patients to correct possibly lethal arrhythmias and lower the risk of unexpected death. However, these treatments are costly and may not be effective in all patients.

Our model calculated the probability of a sudden arrhythmic event with an area under the curve (AUC) of 0.74, where 1.0 is perfect prediction and 0.5 is a random result. This could be used to identify very low-risk patients for whom an ICD or CRT-D is not required and very high-risk patients who should receive a device. Optimizing risk evaluation in this way will improve the cost-effectiveness of treatment.

Professor Kenichi Nakajima, Study Author, Kanazawa University Hospital, Japan

The study comprised of 529 heart failure patients with established two-year results for unexpected arrhythmic events (for example, appropriate shock from an ICD, arrhythmic death, and sudden cardiac death) and death due to heart failure.

Machine learning—a kind of AI employed by the Google search engine and face recognition on smartphones—was used to identify how eight variables used to forecast prognosis of heart failure patients were linked and to develop a formula linking them to two-year results.

The eight factors included gender, age, heart pumping function (left ventricular ejection fraction), heart failure severity (New York Heart Association functional class), whether heart failure was due to limited blood supply (ischemia), B-type natriuretic peptide level in the blood, a nuclear imaging parameter, and kidney function (estimated glomerular filtration rate).

In the course of the two-year follow-up, there were 141 events (27%) including 37 sudden arrhythmic events (7%) and 104 deaths caused by heart failure (20%). The AUC for predicting all incidents was 0.87, whereas for arrhythmic events, it was 0.74, and for heart failure death, it was 0.91.

This is a preliminary study and we can improve the prediction of arrhythmic events by adding variables and continuing to train the machine learning algorithm.

Professor Kenichi Nakajima, Study Author, Kanazawa University Hospital, Japan

Heart-to-mediastinum ratio (HMR) of 123Iodine-metaiodobenzylguanidine (MIBG) uptake was the imaging parameter. MIBG is a radioisotope analog of norepinephrine and is employed to evaluate the action of cardiac sympathetic nerves. Earlier researches have demonstrated that HMR predicts cardiac death in heart failure patients. The HMR is measured by injecting MIBG into a vein and subsequently using imaging to evaluate uptake in the heart and upper mediastinum (center of the thoracic cavity).

Professor Nakajima remarked that although MIBG imaging is accepted in the United States and Japan for clinical practice, and in Europe for clinical research, it is not often used outside Japan owing to its cost. A standard MIBG tracer costs €350 in Japan in comparison with €1,900–3,400 in the United States.

While the costs of the scan may be high, it would be value for money if unnecessary device implantations were avoided.

Professor Kenichi Nakajima, Study Author, Kanazawa University Hospital, Japan

Tell Us What You Think

Do you have a review, update or anything you would like to add to this news story?

Leave your feedback