[Editor's Note: A version of this interview originally appeared on In Vivo on 18 January.]
Medtronic plc’s integration of artificial intelligence (AI) into its implantable cardiac monitor (ICM) R&D program has been transformative, says Felicia Longobardi, the company’s cardiovascular diagnostics and services business unit director for western Europe.
Some 1.4 million Europeans experience ischemic strokes annually, and a quarter of them remain cryptogenic even after a comprehensive diagnostic workup. Up to 30% of these patients may have undetected paroxysmal atrial fibrillation (AF), a common cause of stroke. ICMs detect such arrhythmias, potentially preventing future strokes.
In this interview with In Vivo, Longobardi put into context the patient care advances delivered by Medtronic’s LINQ cardiovascular technology. She also stressed the wider benefits of AI in health care generally − primarily as an affordable way health systems can stay at the forefront of technological advancement and continue to provide patient access to them.
In Vivo: Medtronic’s history as an innovator of implantable cardiac monitor technologies goes back over 25 years. How has in-human ICM data and AI been applied to Medtronic’s R&D and product development as the technology has evolved?
Felicia Longobardi: Since 1998, Medtronic has pioneered the ICM market, consistently innovating based on evolving customer feedback and patient needs. The LINQ family of ICMs help clinicians identify or rule out heart rhythm disorders so patients can receive the right course of treatment to manage their condition.
The Reveal LINQ ICM and LINQ II ICM are inserted just under the skin on the chest and monitor the heart continuously for up to several years. ICMs are ideal for patients with heart rhythm abnormalities that may require long-term monitoring or ongoing management.
In 2014, we initiated a miniaturization revolution that led to the introduction of the world’s smallest ICM, the Reveal LINQ ICM. From there, we advanced to the LINQ II ICM, introducing remote monitoring, BlueSync technology, and now artificial intelligence algorithms that help reduce false alerts.
Incorporating real-world ICM data and AI into our R&D and product development has propelled us toward disease management and true patient-centred care that connects patients and physicians for comprehensive insights and recommendations.
The future is exciting as we continue to transform our data into clinical decisions, offering a bright prospect for both patients and healthcare providers.
What level of breakthrough does LINQ II represent for patients and users, with its CareLink monitoring options and AI capabilities?
The LINQ II is the world's most accurate ICM, tailored to both patient lifestyles and clinician workflows. Designed to detect both asymptomatic and infrequent symptoms like dizziness, palpitations, syncope, chest pain or having had a cardioembolic stroke with suspected atrial fibrillation, LINQ II is instrumental in helping diagnose patients experiencing symptoms so they can get the treatment or medication they need. It also helps patients requiring ongoing management for conditions like AF.
A ground-breaking feature of LINQ II is its remote programming capability, a first for ICMs. This allows direct clinic programming post-insertion, optimizing ongoing programming and reducing patient office visits. The device offers two monitoring options − MyCareLink Heart Mobile App or the traditional MyCareLink Relay Bedside Communicator − for diverse patient preferences.
Technologically, LINQ II has a longer battery life than its predecessor Reveal LINQ (up to 4.5 years), allowing for increased disease management capabilities in the future, and secure, wireless communication via Bluetooth without compromising longevity.
We introduced AccuRhythm AI in 2022, which is artificial intelligence algorithms applied to LINQ II ICMs via the cloud that significantly reduce the most common false alerts (AF and Pause) without compromising sensitivity to the alerts that matter. AccuRhythm AI has shown to reduce false alerts by more than 88%, saving clinicians approximately 200 hours of review time annually for every 100 LINQ II patients.
What market feedback do you get about AI-enabled technologies? What will the next iteration the ICM family of devices be able to achieve?
Our significant presence in western Europe has established a strong market footprint for our ICM technology. Customer satisfaction is paramount at Medtronic, and ensuring a positive and meaningful experience is central to all our engagements.
The launch of AccuRhythm AI is one example of a cloud-based technology improvement that has made a big impact for our customers, based on their feedback.
Looking ahead, our goal for the next iteration of the ICM family of devices is to further enhance diagnostic accuracy while retaining sensitivity and continue our expansion into disease management.
What are the AccuRhythm AI algorithms, how does Medtronic train them and how do they continue to evolve?
Medtronic’s ICMs are designed to be extremely sensitive in their ability to detect abnormal heart rhythms. Because of the sensitivity, they can also generate a large quantity of data, including false alerts, which clinicians and staff must review daily.
As staffing and resource shortages brought on by the COVID-19 pandemic continue to put a strain on clinics and hospitals around the world, reducing clinic burden has never been more important. Managing ICM data more seamlessly through AI technology reduces data burden for clinicians so they can spend their time reviewing the episodes that are clinically relevant.
AccuRhythm AI applies deep learning algorithms to LINQ II ICM data flowing into the CareLink Network. AccuRhythm AI algorithms further address the two most common sources of these ICM false alerts − AF and Pause (syncope) [as seen most recently in a white paper by Radtke and Hall, AccuRhythm AI AF & Pause Algorithms White Paper − April 2023 − Medtronic data on file].
The AI algorithms were rigorously trained and developed based on more than one million professionally adjudicated electrocardiograms (ECGs) for smarter, more accurate insight without data bias.
AccuRhythm AI is just one example of how we are taking bold actions to drive innovation, not just with the introduction of new medical devices, but with data mining and artificial intelligence that make existing technologies smarter and more efficient, while delivering more value to health care professionals and their patients.
How has Medtronic adjusted its ICM R&D and product development platforms to accommodate AI integration?
The integration of AI into our ICM R&D has been transformative. Prior to AI, memory and device longevity constrained our data capabilities. With AI now performed in the cloud, these constraints no longer exist. Our focus has shifted to specific data needs, algorithm training and maintaining industry-leading sensitivity.
This shift allows us to provide clinicians with targeted alerts, enhancing our ability to take meaningful action. It's an exciting journey, and I'm enthusiastic about the possibilities it brings.
How important was the NICE recommendation for the Reveal LINQ technology in September 2020? What value has the recommendation provided in the international sphere, aside from business advantage secured in the UK?
The NICE recommendation for Reveal LINQ technology in September 2020, particularly in their diagnostic guidance (DG41), was a significant milestone. Our investment in the Crystal-AF randomized controlled trial set us apart as the only ICM with sufficient evidence for NICE's endorsement.
The recommendation, specifically for detecting atrial fibrillation after cryptogenic stroke, not only validated our technology but also addressed a crucial need for patients. NICE's economic model estimated that using an ICM could prevent 52 strokes per 1,000 people with cryptogenic strokes. This breakthrough opens doors for underserved patients who had limited access to our technology before the recommendation.
The feedback from our European customers indicates it was a positive step forward. Beyond the business advantage in the UK, this recommendation is a significant win for patients, clinicians and healthcare systems in terms of the implementation of advanced cardiac monitoring technology.
Reveal LINQ
Reveal LINQ is recommended as an option to help to detect atrial fibrillation after cryptogenic stroke, including transient ischaemic attacks (TIA), only if non-invasive electrocardiogram (ECG) monitoring has been done and a cardiac arrhythmic cause of stroke is still suspected. NICE Diagnostics Guidance 41: Implantable cardiac monitors to detect atrial fibrillation after cryptogenic stroke.
NICE is due to review the guidance published back in 2020. What do you expect the outcome to be?
The extent of the NICE review scheduled this year will be dependent on the available new evidence for LINQ and the other implantable cardiac monitors that were reviewed in the original diagnostic appraisal assessment.
With this review, NICE may update the economic model with any new evidence available. One major update could be the improved battery life of LINQ II compared to Reveal LINQ. The original NICE economic model was sensitive to device longevity. A battery improvement of 18 months for LINQ II compared to Reveal LINQ would help drive down the Incremental Cost Effectiveness Ratio (ICER). This would make LINQ II even more cost-effective than Reveal LINQ in detecting atrial fibrillation after a cryptogenic stroke.
How has AI has transformed the abilities of innovators to serve the growing demand for cardiovascular innovations in an affordable way?
The AccuRhythm AI enhancement for LINQ II revolutionises cardiovascular care by offering overnight application for all patients, eliminating the need for device upgrades or new purchases. This innovation negates the hassles of explanting and inserting new devices, providing cost-effective access to the latest technology.
The incorporation of AI not only reduces false alerts but also enables enhancements that prioritize actionable data. This allows clinicians to deliver more focused care to a broader population. In essence, AI emerges as one of the most affordable ways for health systems to stay at the forefront of technological advancement, ensuring accessible and efficient cardiovascular innovations.
As business director of cardiovascular diagnostics and services, what preoccupies and concerns you and your Medtronic team about meeting the growing demand for cardiovascular care?
Today, almost half of patients admitted to hospitals with syncope are leaving without a diagnosis. Beyond compromising their quality of life, this poses a significant challenge for our health care systems. These undiagnosed patients often find themselves in a cycle of readmissions and referrals, creating a burden on both patients and the health care infrastructure.
Our LINQ family of ICMs have proven to be instrumental in providing a high diagnostic yield. While Europe has shown a higher utilization of this technology, there's still work to be done to ensure uniform access across the entire region. Every patient, regardless of location, deserves excellent care, and our mission is to empower health care professionals to deliver that.
My role extends beyond the business realm; it's about breaking down barriers that hinder the availability of this life-changing technology for all eligible patients. We're committed to educating health care professionals, fostering collaboration and improving existing workflows. Together, we can strive to ensure that every patient, regardless of their cardiovascular condition, receives the care they need.
