New AI Software can Identify Cardiac Devices More Accurately from X-Ray Images

Researchers have developed new AI software with the ability to spot cardiac rhythm devices in X-rays more precisely and rapidly when compared to existing techniques.

The research group hopes this software could accelerate the diagnosis and treatment of patients with defective devices in an emergency setting.

Developed by scientists at Imperial College London, the software has been able to recognize the make and model of different cardiac rhythm devices, for example, defibrillators and pacemakers, within seconds. The research, reported in JACC: Clinical Electrophysiology, was conducted at Hammersmith Hospital, part of Imperial College Healthcare NHS Trust.

Pacemakers and defibrillators have improved the lives of millions of patients from around the world. However, in some rare cases, these devices can fail and patients can deteriorate as a result. In these situations, clinicians must quickly identify the type of device a patient has so they can provide treatment such as changing the device’s settings or replacing the leads. Unfortunately, current methods are slow and out-dated and there is a real need to find new and improved ways of identifying devices during emergency settings. Our new software could be a solution as it can identify devices accurately and instantly. This could help clinicians make the best decisions for treating patients.

Dr James Howard, Study Lead Author and Clinical Research Fellow, Imperial College London.

Over one million people worldwide go through implantation of a cardiac rhythm device every year, with more than 50,000 being implanted per year in the United Kingdom. These devices are implanted beneath the skin of patients to either assist the heart’s electrical system to work properly or measure heart rhythm. Pacemakers heal slow heart rhythms by “pacing” the heart to beat more rapidly, whereas the defibrillators deal with fast heart rhythms by giving electric shocks to reset the heartbeat back to a normal rhythm.

However, in some atypical conditions, these devices can lose their potential to control the heartbeat, either due to the malfunction of the device or dislocation of the wires connecting it to the heart. When this occurs, patients may suffer tremors, inappropriate electric shocks, or loss of consciousness.

In such cases, clinicians have to identify the model of a device to analyze why it has stopped working. Only if they have access to the records where implantation was done, or if the patient can tell them, staff can identify the pacemakers; otherwise, they must use a flowchart algorithm to determine it by a process of elimination. The flowchart includes a range of shapes and circuit board components of different pacemakers developed to aid clinicians in identifying the make and model of a patient’s pacemaker. In addition to being time-consuming, these flow charts are now out of date and are thus inaccurate. This can cause delays in providing care to patients, who are usually in serious conditions.

In the new research, scientists trained the software program known as a neural network to recognize over 1600 different cardiac devices from patients.

To employ the neural network, the clinician uploads the X-ray image showing the device into a computer and the software deciphers the image to offer a result on the make and model of the device in seconds.

The group used the program to find out whether it could recognize the devices from radiographic images of over 1500 patients at Hammersmith Hospital between 1998 and 2018. Next, they compared the results with five cardiologists who used the traditional flowchart algorithm to determine the devices.

The group discovered that the software was more efficient than existing techniques. It was 99% accurate in recognizing the make and model of a device, whereas the flow chart was just 72% accurate. The researchers believe the software could significantly accelerate the treatment to patients with heart rhythm device issues.

The scientists will seek to conduct a further trial to corroborate the results in a larger group of patients and finding out ways to develop a more portable device that can be used in hospital wards.

The work was funded by NIHR Imperial Biomedical Research Centre, the Medical Research Council, the Wellcome Trust, and the British Heart Foundation.