SandboxAQ developed AI-enhanced magnetocardiography (MCG) technology that offers a non-invasive, radiation-free cardiac assessment and aims to help clinicians make a faster, more accurate diagnosis.
Mayo Clinic is partnering with startup SandboxAQ to study a new medical device that uses quantum sensing technology and advanced AI algorithms for rapid diagnosis in cardiac care, such as potentially detecting a heart attack.SandboxAQ was spun out of Alphabet in 2022 and is working on solutions at the nexus of AI and quantum technology, with a particular focus on cybersecurity, navigation, sensing technology, which is where its medical devices fall, and also using large quantitative models to aid drug discovery and materials science."The A and the Q stand for AI and quantum but what we're really good at is delivering these deep tech to solve impactful problems in different industries," said Kit Yee Au-Yeung, Ph.D., general manager of the sensing division at SandboxAQ, in an exclusive interview with Fierce Healthcare.She points out that the startup's work in healthcare is different from other efforts to use large language models (LLMs) or generative AI. "We're focused on a physics-informed application of AI to extract useful information," she said.A biomedical engineer by background, Au-Yeung has worked in the medical devices field for 20 years and she was part of the founding team of iRhythm Technologies. She joined Google X, the moonshot factory, in 2018 before joining SandboxAQ. Her 16-person team is made up of physicists and engineers working to build a diagnostic device to impact cardiac care. Au-Yeung's team developed AI-enhanced magnetocardiography (MCG) technology that offers a non-invasive, radiation-free cardiac assessment and aims to help clinicians make a faster, more accurate diagnosis. To guide its work, SandboxAQ put together a clinical advisory board, which includes Toby Cosgrove, M.D., cardiac surgeon and former president and CEO of the Cleveland Clinic for 13 years. Faster diagnosis can improve patient care and reduce the economic burden of medical care, according to a study from the American Heart Association.Chest pain represents nearly 5% of all the reasons prompting patients to seek care in the ED, amounting to an estimated 6.5 million encounters in the United States in 2017. And, it's a condition frequently hospitalized. It's estimated that as much as $3 billion of national health care expenditures could be avoided annually by narrowing the variation in ED hospitalizations for chest pain.The non-contact cardiac imaging device uses magnetic sensors and advanced AI algorithms can produce a rapid visualization and assessment of the magnetic signals of the heart. The Mayo Clinic is working with SandboxAQ on a clinical research study to examine the relationship between MCG and the findings seen in angiography. The study, which will involve about 150 patients, will take place throughout 2024 and 2025 at the Mayo Clinic in Rochester, Minnesota, the organizations said."We partner with prestigious institutions like the Mayo Clinic because we believe in their research rigor. We're good at the AI and engineering and the physics and Mayo Clinic brings an expertise in clinical practice and understanding the gaps in clinical service. We really want to take this technology out of the research lab and into a clinicians' work," Au-Yeung said.The startup's novel MCG imaging system, called CardiAQ, is still in research and development as an investigation device but Au-Yeung sees the potential for the technology to overcome many of the existing tools used by cardiologists to make heart health decisions.SandboxAQ completed a feasibility study with the University of California at San Francisco (UCSF) Medical Center, and have ongoing studies at Mount Sinai Medical Center in New York City to test MCG’s ability to improve cardiac diagnosis.The startup eventually wants to pursue regulatory clearance through the FDA for its medical device. "My vision is that we know enough about our technology and can be ready for something like a pivotal study in 2025 and that study would last between nine to 12 months and then that will go into an FDA submission and clearance," she said."Magnetocardiography as a concept dates back to 1960s, so we're not the first group to do that. But how we're different is that we use a new approach to do that," Au-Yeung said.Widespread adoption of MCG technology has been a challenge. Historically, MCG systems required large dedicated rooms and expensive maintenance and came with a large price tag, according to the company."Previous generations of MCG to get this data requires really big machinery, couple million dollars; you will have to construct a room to shield out all the magnetic noise. around the area in order to be able to capture heart signal. My team's approach is different. We want the device to go to the patient as opposed to asking the patient to go to the device. We want it to be a bedside tool," she said."We leverage high-performance sensors, like quantum sensors, so that we can capture this tiny signal and we use AI to do noise rejection, to learn the environment, so that we can separate out what's coming from the heart and what's coming out from the environment. That's what enables us to build something that can be used anywhere. That's the goal," she added. Au-Yeung and her team see the potential to improve how clinicians approach cardiac diagnostics and patient care. The standard diagnostic tool currently used to monitor heart activity and detect cardiac issues is the electrocardiogram (EKG), a 150-year-old technology that has vastly improved cardiac care.But, EKG tech has limitations. When measured by EKG, the electrical signals of the heart may be distorted as they pass through the body. This can complicate diagnosis and can lead to inaccurate and potentially dangerous conclusions, according to SandboxAQ executives. Biomarker tests also are often used."These tools are good, but they are suboptimal because of their sensitivity and specificity performance. A lot of patients fall into a gray zone, even when you have these results. A patient's physician will decide they don't have enough information so they decide to keep the patient for observation and have to admit them to the hospital to do additional testing and that becomes very costly," Au-Yeung said."The first application that we are going after is heart attack triage in the emergency department. Magnetocardiography has been proposed to be a frontline tool, added to the toolbox of physicians to make this decision much faster and more accurately," she said, adding, "This is just our first application. There are other indications that we will be researching in the future."CardiAQ detects the magnetic fields of the heart using room-temperature sensors and AI software in a process. The device does not need any room shielding which significantly opens up the use cases from ER, to bedside and even in ambulances."Cardiovascular disease is the number one killer across most of the world. As a doctor, having the most accurate and timely information about a patient’s cardiac state is critical for making life-saving treatment decisions,” Cosgrove said in a statement. “SandboxAQ's CardiAQ device shows how AI and this next generation of sensing technologies have the potential to revolutionize cardiac diagnosis and address the unmet clinical need for more effective patient care. This is the future of medicine.” Novel methods to easily and accurately diagnose disease are urgently needed, Au-Yeung said. "CardiAQ captures detailed data from the heart, which could lead to faster, more accurate diagnoses and fill in the gap between low-cost EKG or biomarker tests and high-cost CT scans or invasive angiography.”Some of the benefits of the CardiAQ device include non-invasive, radiation-free cardiac assessment, the potential for earlier detection of cardiac issues, improved accessibility of advanced cardiac diagnostics and AI-driven analysis for more accurate and rapid diagnosis, according to the company. “The current methods for heart attack diagnosis can miss many cardiac conditions. Blood tests are costly, take time to process, and may give false negatives and positives. SandboxAQ’s goal is to revolutionize cardiac diagnoses with contactless, high-performance sensors and sophisticated AI that has the potential to benefit patients, save lives, and impact cardiac care globally with on-demand, real-time analysis," said Jeffrey Bander, M.D., Chief of Cardiology at Mount Sinai West.The company is part of a wave of AI-based startups making inroads in healthcare, which is poised to be a lucrative market for AI. The AI-enhanced diagnostic market is set to grow from $16.3 billion in 2023 to $71.2 billion in 2027. In 2023, SandboxAQ raised $500 million to fuel its product development in AI and quantum computing.