Modern AI-Enabled Pacemakers
A cardiac pacemaker is a medical equipment that generates impulses within the heart in case of an abnormal heart rhythm and coordinates the electrical signaling between the heart’s chambers. Since the development of the first cardiac pacemaker in 1958, this therapeutic device has undergone several modifications. Every year, thousands of surgeries are performed in cardiac patients for pacemaker placement or revision. Although a pacemaker is commonly indicated in patients over 60 years old, it is placed in infants and children if indicated.
There are two types of pacemakers- permanent and temporary. While temporary pacemakers are used in situations where arrhythmias are anticipated to be temporary, permanent pacemakers are often recommended for patients with bradyarrhythmia or tachyarrhythmia. Advancements in implantable pacemakers have resulted in several modifications, such as reducing the size of the device, remote monitoring using the internet, and a considerable increase in battery longevity.
France based Implicity is a startup that can monitor implanted cardiac-connected devices remotely with artificial intelligence. Their product connects with pacemakers and implanted defibrillators to facilitate heartbeat at a regular pace and also treat in cases of emergency such as heart attack. All devices being connected also enables to transmit the data back to the server and the physician. The technology at Implicity examines the patients’ data from the implanted pacemaker and from the patient’s medical record to identify high-risk patients that would require modifications in medication or change in the treatment plan.
A research team at Bios leverages machine learning to identify patterns of neural biomarkers and then synchronize them based on other organ functions to develop an algorithm. Towards this, they record raw neural data through neural interfaces. They compare with the recordings of physiological signals such as heart rate, blood pressure, glucose levels, body temperature, and physical activity levels.
Usually, patients with pacemakers have certain contraindications while undergoing medical and dental treatments. One such contraindication is undergoing MRI imaging. But the recent FDA-approved MRI pacemakers have revolutionized the way cardiac patients undergo imaging examinations without adverse effects, and Biotronik is one such example.
Shallow Blue is a startup that deals with pacemaker predictive analytics. Their system can predict the onset of cardiac arrest in a person with a pacemaker or ICD so that appropriate intervention can be administered in advance.
Furthermore, scientists at Rice University and the Texas Heart Institute (THI) are working on machine learning algorithms that could improve the current pacemaker algorithm’s overall sensitivity. It aims to identify an algorithm that can distribute the pacemaker’s effect across several areas in the heart in a synchronized and accurate manner with the help of machine learning algorithms and validate the machine learning algorithm to classify different pathophysiology.
Another application of implantable pacemaker is the detection of an electrical storm (ES) that is a dangerous arhythmic condition with increased mortality and morbidity. There is ongoing research to construct and validate machine learning models to predict ES based on an intracardiac device’s remote monitoring possibility (ICD). A recent study by Professor Saeed Shakibfar from the University of Copenhagen showed that in patients with ICD, the occurrence of ES could be predicted one day in advance based on the data stored in ICD summaries with good diagnostic accuracies.
Professor Truong and the team at the Linder Research Centre, Ohio, investigated the feasibility of predicting pacemaker risk following transcatheter aortic valve replacement using ECG data and comparing the traditional logistic regression with machine learning approaches. It was found that machine learning methodology is significantly more robust than conventional logistic regression in predicting permanent pacemaker implantation risk following transcatheter aortic valve replacement with good diagnostic accuracy.
Bioelectronic devices are revolutionizing the pacemaker industry primarily by replicating neural signals. These control the activity of abnormal neural circuits that possibly cause the disease. But there is a need for a better understanding of neural biomarkers to receive signals and deliver appropriate treatment autonomously.
There is a need to develop a more efficient implantable device for improved interpretation of neural data. But it may not be as easy as it sounds because the human heart does not function alone in conjunction with other organs. Impulses from different organs and activities heavily affect the functioning of the heart. What’s more, machine learning can also utilize medical data to support patients’ personalized clinical decisions after running validated models.