A team of researchers at Rice University and the Texas Heart Institute (THI) have received a prestigious National Science Foundation grant to develop machine learning algorithms to guide pacemakers to meet the needs of individual patients, in real-time, to provide patient-specific and episode-specific therapy.

Machine learning is the ability by which a machine can learn from datasets to carry out specific tasks. In healthcare, machine learning algorithms are already playing an important role in helping doctors improve patient care and ultimately overall patient quality of life. For example, machine learning algorithms have contributed to several advancements that have improved quality of life and standard of living.

In general, machine learning has allowed for safer transportation, more accurate weather predictions and in clicking great photos with your mobile phones. In the hospital, machine learning has contributed to increasing accuracy of already robust algorithms deployed in imaging tools and other diagnostic devices. These algorithms have helped physicians make quicker, more informed diagnosis. In the future, such alglorithms will pave way for a more personalized treatment.

Pacemakers could be among the next wave of devices to benefit from the integration of machine learning.

Under this grant, Dr. Mehdi Razavi and team are exploring the possibility of distributing a pacemaker’s effect across multiple places in the heart in a synchronized and precise fashion using machine learning algorithms. Algorithms for a novel, multisite pacing enabled therapy have yet to be explored.

“Our project could contribute significantly to overall improvement of the sensitivity of the current pacemaker algorithm,” according to Dr. Mehdi Razavi, Director of Electrophysiology Clinical Research & Innovations at THI. “Of course, the accuracy of any machine learning algorithm depends on the dataset from which it learns. So, in this collaboration with Rice we are collecting and managing a robust dataset of intracardiac electrogram data based on past pacemaker performance.”

The team of engineers at Rice and THI, led by Dr. Behnaam Aazhang and Dr. Yingyan Lin, will use this dataset to develop and validate the machine learning algorithm to classify different pathophysiologies.

“Because sophisticated machine learning algorithms can be rather power hungry and could affect the overall battery life of a pacemaker, we are primarily focused on deploying the algorithms on hardware in an energy efficient manner to reduce power consumption to improve the life of the device,” said Dr. Lin, Assistant Professor, Electrical and Computer Engineering.

 

Read: Customized pacemakers may learn what our hearts need

 

Special thank you to Payton Campbell for her assistance with this news story.

 

---

About Texas Heart Institute

The Texas Heart Institute (THI), founded by world-renowned cardiovascular surgeon Dr. Denton A. Cooley in 1962, is a nonprofit organization dedicated to reducing the devastating toll of cardiovascular disease through innovative and progressive programs in research, education and improved patient care. More information about THI (@Texas_Heart) is available at www.texasheart.org.