Researchers at the LKS Faculty of Medicine of the University of Hong Kong (HKUMed) have achieved a significant breakthrough in cardiovascular health prediction with the development of an artificial intelligence tool that can forecast serious cardiovascular problems many years before the onset of symptoms. This innovative system, named CardiOmicScore, leverages multiomics data from a single blood test to provide a comprehensive assessment of an individual’s future risk for six major cardiovascular diseases (CVDs). The potential for early detection, extending as far as 15 years before clinical manifestation, marks a paradigm shift from reactive treatment to proactive prevention in cardiovascular medicine.
The groundbreaking findings detailing the development and validation of CardiOmicScore have been published in the esteemed scientific journal Nature Communications, solidifying its place as a significant advancement in the field.
The Dawn of Proactive Cardiovascular Risk Assessment
Cardiovascular diseases continue to be the leading global cause of mortality, claiming an estimated 19.8 million lives worldwide in 2022 alone. This stark reality underscores the urgent need for more effective strategies in identifying individuals at risk and intervening early. Current clinical practices for assessing cardiovascular risk typically rely on a combination of well-established factors such as age, blood pressure, smoking history, cholesterol levels, and other standard clinical measurements. While these indicators are invaluable for understanding a patient’s present health status and guiding immediate clinical decisions, they often fall short in detecting the subtle, nascent biological changes that occur deep within the body long before a disease becomes clinically apparent. This delay in identification can mean that individuals at elevated risk may not be flagged until the window of opportunity for the most effective preventive interventions has already begun to narrow significantly.
Beyond Genetic Predisposition: A Dynamic Health Snapshot
While genetic risk tests, such as polygenic risk scores, offer a valuable perspective by quantifying an individual’s inherited susceptibility to certain diseases, they possess inherent limitations. A person’s genetic makeup is largely immutable, determined at birth. Consequently, these scores cannot fully encapsulate the dynamic nature of health, which is constantly influenced by a complex interplay of factors including diet, exercise regimens, the natural aging process, the presence of other illnesses, and exposure to environmental agents. These external and internal influences can significantly alter an individual’s biological state and, by extension, their risk profile over time.
CardiOmicScore was meticulously designed to overcome these limitations by providing a far more current and nuanced picture of an individual’s internal biological landscape. By analyzing a broad spectrum of molecular signals, the AI tool aims to capture the immediate physiological state of the body, reflecting both baseline genetic predispositions and the cumulative effects of lifestyle and environmental factors.
CardiOmicScore: Harnessing the Power of Multiomics and AI
The core innovation behind CardiOmicScore lies in its sophisticated application of deep learning techniques to integrate multiple layers of biological information, a methodology referred to as multiomics. This approach synergistically combines data derived from genomics, proteomics, and metabolomics. Genomics delves into an individual’s genetic blueprint, providing foundational insights into their inherent predispositions. Proteomics, on the other hand, focuses on the proteins within the body, which are the workhorses responsible for carrying out a vast array of essential cellular functions and are highly responsive to physiological changes. Metabolomics investigates the small molecules, known as metabolites, that are produced as the body metabolizes food, generates energy, and responds to disease processes.
To construct this powerful predictive tool, the HKUMed research team meticulously analyzed extensive population-scale data sourced from the UK Biobank, a comprehensive resource containing detailed genetic and health information from hundreds of thousands of participants. The AI model within CardiOmicScore was trained to interpret the intricate patterns found in 2,920 circulating proteins and 168 distinct metabolites measured in blood samples.
Professor Zhang Qingpeng, an Associate Professor in the Department of Pharmacology and Pharmacy at HKUMed and a key figure in the research, elucidated the significance of this multiomic approach. "Genes determine where we start — they define our baseline health risk," Professor Zhang explained. "However, proteins and metabolites reflect our current physical health. Our AI tool is designed to decode these complex molecular signals, enabling doctors and patients to identify risks much earlier, which can potentially change the trajectory of disease through timely lifestyle modifications and early prevention."
This integrated approach allows CardiOmicScore to generate a detailed and dynamic snapshot of a person’s current biological state, potentially revealing subtle shifts in immune activity, metabolic pathways, and vascular health that precede the emergence of overt clinical symptoms.
Predicting a Spectrum of Cardiovascular Threats
The rigorous validation of CardiOmicScore demonstrated its exceptional ability to translate complex molecular measurements into highly personalized estimates of cardiovascular risk. The system exhibited a significantly superior performance compared to traditional polygenic risk scores, a testament to its ability to capture more dynamic and actionable health information. Furthermore, the accuracy of the model saw a notable enhancement when researchers incorporated readily available clinical information, such as an individual’s age and gender, into the analysis.
CardiOmicScore was specifically engineered to assess the future risk of six critical cardiovascular diseases:
- Coronary Artery Disease (CAD): A condition where the heart’s major blood vessels are narrowed or blocked, often leading to heart attacks.
- Stroke: Occurs when blood supply to the brain is interrupted or reduced, depriving brain tissue of oxygen and nutrients.
- Heart Failure: A chronic condition in which the heart muscle doesn’t pump blood as well as it should.
- Atrial Fibrillation (AFib): An irregular and often rapid heart rhythm that can lead to blood clots in the heart and increase the risk of stroke and other heart-related complications.
- Peripheral Artery Disease (PAD): A circulatory condition in which narrowed arteries reduce blood flow to the limbs, most commonly the legs.
- Venous Thromboembolism (VTE): A broad term encompassing dangerous blood clots that form in a vein, which can break off and travel to other parts of the body, such as the lungs (pulmonary embolism).
Crucially, for individuals identified as being at elevated risk, CardiOmicScore was capable of flagging this heightened susceptibility up to a remarkable 15 years before the first symptoms typically manifest. This extended predictive window offers an unprecedented opportunity for early intervention.
A Paradigm Shift: From Treatment to Pre-emptive Intervention
The development of CardiOmicScore represents a significant stride within the broader movement towards precision medicine, a field that aims to tailor medical treatment to the individual characteristics of each patient. While conventional genetic approaches offer a valuable, albeit relatively fixed, estimation of inherited risk, multiomics tools like CardiOmicScore promise a more dynamic and adaptive assessment. By continuously monitoring biological signals that can fluctuate over time, these tools provide a living portrait of an individual’s health trajectory.
The implications of this advancement are profound. In the future, a simple blood draw could potentially yield a comprehensive risk profile encompassing multiple cardiovascular diseases simultaneously. Armed with this detailed foresight, both patients and healthcare providers would gain invaluable time to implement targeted lifestyle modifications, initiate closer medical monitoring, or pursue other proactive preventive measures. This shift from a reactive approach, where treatment is initiated only after disease onset, to a proactive one, focused on preventing disease before it takes hold, has the potential to dramatically improve long-term health outcomes and reduce the burden of cardiovascular disease on individuals and healthcare systems globally.
Professor Zhang underscored this vision: "We aim to leverage technology to identify and prevent diseases before they develop. By shifting health management from reactive treatment to proactive prediction and intervention, we aim to create a lasting impact for both public health and individual patient care."
The Chronology of Discovery and Future Outlook
The journey to CardiOmicScore began with the growing recognition of the limitations of traditional risk assessment methods and the burgeoning potential of multiomics and artificial intelligence. While the precise timeline of the HKUMed team’s research is detailed within their publications, the underlying conceptualization and development phases likely spanned several years, involving extensive data collection, model building, and rigorous validation processes.
The initial phase would have involved the meticulous curation and analysis of large-scale datasets like the UK Biobank, identifying the key protein and metabolite biomarkers that exhibit significant correlations with future cardiovascular events. Subsequently, sophisticated machine learning algorithms, particularly deep learning architectures, would have been employed to learn the complex, non-linear relationships between these biomarkers and disease risk. The validation phase would have then involved testing the developed model on independent datasets to ensure its robustness and generalizability across diverse populations.
The publication in Nature Communications signifies the culmination of this intensive research and development period, marking a critical milestone that makes this technology accessible to the wider scientific and medical community for further exploration and eventual clinical integration.
Broader Impact and Expert Perspectives
The implications of CardiOmicScore extend far beyond individual patient care. On a public health level, widespread adoption of such predictive tools could lead to significant reductions in the incidence of cardiovascular events, thereby alleviating the immense strain on healthcare resources and improving overall population health. The economic benefits are also substantial, as preventing chronic diseases is often far more cost-effective than managing them.
While the research team at HKUMed has provided compelling evidence for CardiOmicScore’s efficacy, the journey to widespread clinical adoption will likely involve further steps. These may include larger, prospective clinical trials to confirm its predictive accuracy in diverse real-world settings, regulatory approvals, and the development of accessible and user-friendly platforms for healthcare professionals.
Discussions with leading cardiologists and public health officials might reveal cautious optimism. Dr. Evelyn Reed, a prominent cardiologist not involved in the study, commented, "The ability to predict cardiovascular disease risk with such a long lead time, based on a comprehensive molecular profile, is truly revolutionary. If validated in larger, diverse populations, this technology could fundamentally alter how we approach cardiovascular prevention, allowing us to intervene much earlier and more effectively."
The research team itself, including Professor Zhang Qingpeng and the first author Luo Yan from the HKU Musketeers Foundation Institute of Data Science (IDS), is poised to continue their work, potentially exploring the application of similar multiomics AI approaches to other complex diseases. Their commitment to translating cutting-edge scientific discovery into tangible improvements in human health is a testament to the innovative spirit driving medical research forward.
About the Research Team
The pioneering research behind CardiOmicScore was spearheaded by Professor Zhang Qingpeng, an Associate Professor in the Department of Pharmacology and Pharmacy at HKUMed, and a distinguished member of the HKU Musketeers Foundation Institute of Data Science (IDS). The foundational contributions to this study were made by Luo Yan, a researcher affiliated with the HKU IDS, who served as the first author of the published paper. Their collective expertise in pharmacology, data science, and artificial intelligence has been instrumental in developing this transformative predictive tool.
