Engineers and doctors have developed a wearable ultrasound machine that can assess the structure and function of the human heart. The portable device, which is about the size of a postage stamp, can be worn for up to 24 hours and works even during a tough workout.
The goal is to make ultrasound accessible to a larger population, said Xing Xu, a professor of nanoengineering at UCSD who leads the project. Currently, echocardiograms — ultrasound examinations of the heart — require highly trained technicians and bulky equipment.
“Technology enables anyone to use ultrasound on the go,” Xu said.
Thanks to customized AI algorithms, the device can measure the amount of blood pumped by the heart. This is important because the heart not pumping enough blood is the root cause of most cardiovascular diseases. Problems with heart function often only appear when the body is in motion.
The work is described in the January 25 issue of the magazine nature.
Cardiac imaging is an essential clinical tool for assessing long-term heart health, detecting problems as they arise, and caring for critically ill patients. A new wearable, non-invasive heart monitor provides humans with automatic, real-time insights into heart pumping activity that are hard to capture, even when a person is exercising.
The wearable heart monitoring system uses ultrasound to continuously capture images of the heart’s four chambers at different angles, analyzing a clinically relevant subset of the images in real time using specially designed AI technology. The project builds on the team’s previous developments in wearable deep-tissue imaging technologies.
The increased risk of heart disease calls for more advanced and comprehensive monitoring measures. By providing patients and clinicians with more granular detail, continuous, real-time imaging and monitoring of the heart is poised to fundamentally improve and reshape the model of cardiac diagnostics.”
Sheng Xu, Professor of Nanoengineering, University of California San Diego
In comparison, current non-invasive methods have limited sampling capabilities and limited data availability. The wearable technology developed by the Xu team enables safe, non-invasive, and high-quality cardiac imaging, resulting in images with high spatial resolution, resolution, and temporal contrast. “It also reduces patient discomfort and overcomes some limitations of non-invasive technologies such as CT and PET, which can expose patients to radiation,” said Hao Huang, a doctoral student in Xu’s group at UC San Diego.
The unique design of the sensor makes it ideal for moving objects. “The device can be attached to the chest with minimal restrictions on people’s movement, even providing continuous recording of heart activities before, during, and after exercise,” said Xiaoxiang Gao, a postdoctoral researcher in Xu’s group at UC San Diego.
The importance of imaging the heart
Heart disease is the leading cause of death among the elderly, and it is also becoming more prevalent among the young due to lifestyle factors. Signs of heart disease are transient and unpredictable, which makes them difficult to detect. This has created a demand for more advanced, comprehensive, non-invasive and cost-effective monitoring technologies such as long-term cardiac imaging, which this wearable device facilitates.
Cardiac imaging is one of the most powerful tools for screening and diagnosing heart problems before they become problems. “The heart is subject to all kinds of different diseases,” said Hongjie Hu, a postdoctoral researcher in Xu’s lab at UC San Diego. “Cardiac imaging will reveal the real story underneath. Whether it is the vigorous but natural contraction of the heart’s chambers causing fluctuating volumes, or a cardiac morphological problem occurred as an emergency, real-time image monitoring on the heart shows the whole picture with vivid detail and visual impact.”
How it works in detail
The new system collects information through a wearable patch that is as soft as human skin, designed for optimal adherence. The patch measures 3/4″ (L) x 2.2″ (W) x 0.09″ (T), which is approximately the size of a postage stamp. It sends and receives ultrasound waves that are used to generate a continuous stream of images of the heart’s structure in real time. The ultrasound patch is soft and stretchy, and adheres well to human skin, even during exercise.
The system can examine the left ventricle of the heart in separate, two-plane views using ultrasound, resulting in more clinically useful images than was previously available. As a use case, the team demonstrated imaging of the heart during exercise, which is not possible with the rigid, cumbersome equipment used in clinical settings.
The performance of the heart is characterized by three factors: stroke volume (the volume of blood the heart pumps out in each beat), ejection fraction (the percentage of blood pumped from the left ventricle of the heart in each beat) and cardiac output (the volume of blood pumped out by the heart in each beat). The volume of blood pumped by the heart every minute).
Xu’s team has developed an algorithm to facilitate continuous automatic processing with the help of artificial intelligence.
“The deep learning model automatically cuts out the shape of the left ventricle from the continuous image recording, extracts its size frame by frame, and generates waveforms to measure stroke volume, cardiac output, and ejection fraction,” said Mohan Li, a master’s student. Xu’s group at the University of California, San Diego.
“Specifically, the AI component includes a deep learning image segmentation model, a heart size calculation algorithm, and a data computation algorithm,” said Ruixiang Qi, a master’s student in Xu’s group at UC San Diego. “We use this machine learning model to calculate the heart size based on the shape and area of left ventricular segmentation. The deep learning model by photogrammetry is the first to be used in wearable ultrasound devices. It enables the device to accurately and efficiently provide continuous waveforms of key cardiac indicators in cases of different physical conditions, including rest and after exercise, which has not been achieved before.”
Thus, this technology can generate curves for these three indicators continuously and non-invasively, as the AI component processes the continuous flow of images to generate numbers and curves.
To create the platform, the team faced some technical challenges that required careful decisions. To produce the wearable device itself, the researchers used a 1-3 piezoelectric composite bonded to an Ag-epoxy support as the material for the ultrasound machine’s transducers, which reduces risks and improves efficiency compared to previous methods. When choosing a transmission configuration for a group of converters, they achieved superior results with their broadband composite transmission. They were also selected from among nine popular machine learning-based image segmentation models, and landed on the FCN-32, which achieved the highest possible accuracy.
In the current iteration, the patch is connected through cables to a computer, which can download data automatically while the patch is in progress. The team developed a wireless debugging circuit, which will be covered in an upcoming post.
Xu plans to commercialize this technology through Softsonics, a company that grew out of the University of California, San Diego and that he co-founded with engineer Xu Xiang. He also encourages others in his scientific community to follow his lead and work on areas of this research that require further exploration.
To follow up on these results, Xu recommends the following four immediate steps:
- Imaging in B mode, which allows for more diagnostic capabilities involving different organs
- The soft imager design, which allows the researchers to fabricate large transducer probes that cover multiple sites simultaneously
- Minimize the rear system that operates the soft camera
- Working on a generic machine learning model that fits more topics