The Problem
Background
Atrial fibrillation, or AFib, is an irregular heartbeat that can cause painful episodes, significantly impacting daily life, and leading to severe complications such as stroke, heart failure, and reduced overall well-being. Today, AFib already affects approximately 33 and a half million people worldwide. With our aging population, the number of cases is expected to double by 2050, placing even greater pressure on global healthcare systems. Many existing AFib solutions remain hard to interpret and difficult to use effectively, leaving both patients and clinicians uncertain about how to take meaningful, timely actions.
Problem Statement
We hope to create a product that can predict and monitor AFib episodes to help people catch episodes early, and bring them better health outcomes and more peace of mind.
Why It Matters
The market opportunity is equally compelling. The global AFib devices market is projected to expand from $4.6 billion in 2024 to over $13 billion by 2034, driven by breakthroughs in wearable technology and AI integration. Our solution meets an urgent clinical need and positions us at the forefront of a rapidly growing market, promising significant health benefits and strong commercial potential.
- According to a recent study conducted at UCSF, an estimated 10.5 million Americans have AFib: 3x more than previously thought
- An estimated 13% of people with AFib are undiagnosed because they have asymptomatic episodes. A wearable that monitors for AFib episodes could get asymptomatic people to the doctor earlier, before things become more serious
- In 2005, AFib hospitalizations in the United States incurred $6 billion in healthcare costs
Current Approaches and Limitations
Advancements in AFib detection and management have become increasingly important. From wearable technology to AI-driven models, new solutions are emerging to improve early diagnosis and intervention.
Consumer smartwatches, like those from Apple and Samsung, now integrate AFib detection using photoplethysmography, providing at-risk individuals with continuous heart monitoring.
Beyond wearables, Cardiac Monitoring and Electrophysiology Devices play a critical role. Implantable Cardiac Monitors offer long-term AFib tracking for sporadic episodes, while Holter Monitors provide short-term ECG monitoring in clinical settings. On the treatment side, Catheter Ablation Devices are a leading intervention, driving the electrophysiology market to over $3.2 billion in 2023. With higher success rates than drug therapies, these minimally invasive procedures are increasingly preferred.
AI is also transforming AFib detection. The Mayo Clinic’s convolutional neural network predicts new-onset AFib with an AUC of 0.87, while Philips’ Cardiologs AI model forecasts short-term risk using 24-hour Holter data. These advancements highlight AI’s potential in early intervention.
Our Approach
Our minimum viable product, an AFib Classification Dashboard, centers on two core elements: timestamped AFib data and an explainable, data-driven classification approach.
By recording the exact times of AFib episodes, our dashboard provides users with a clear picture of when events occur. This information is invaluable for personalizing lifestyle adjustments. When individuals can correlate an AFib episode with specific moments in their day—be it after exercise, during periods of high stress, or following a poor night’s sleep—they gain the insight needed to modify their behavior. This kind of targeted feedback can drive meaningful changes, reducing the frequency or severity of future episodes.
Rather than operating as a “black box,” our model highlights which parts of the ECG data most influence the classification of an episode. This transparency is crucial. For clinicians, it means a clearer understanding of the decision-making process behind each diagnosis, which enhances trust in the tool and fosters better integration into clinical practice. For patients, it demystifies the process. They’re not just told that an episode occurred—they can see the specific factors that contributed to that classification. This level of insight empowers users to take proactive steps in managing their heart health.
By combining precise, timestamped data with an explainable classification method, we enable timely interventions that can improve patient outcomes and reduce complications. This personalized approach not only benefits individuals but also supports clinicians in optimizing care.