Detect Fatty Liver Early: Unlock the Full Potential of Cardiometabolic Health with AI-CVD™
AI-CVD™ revolutionizes the detection and management of fatty liver, a key indicator of metabolic syndrome and cardiovascular disease. By leveraging advanced AI algorithms, AI-CVD™ provides precise quantification of liver fat from routine CAC scans, enabling early intervention and personalized treatment plans. Take the first step towards comprehensive cardiometabolic care and improved patient outcomes.
Correlation with Metabolic Syndrome and Diabetes
Metabolic syndrome is a cluster of conditions, including obesity, hypertension, dyslipidemia, and insulin resistance, that collectively increase the risk of heart disease, stroke, and type 2 diabetes. Fatty liver is often considered the hepatic manifestation of metabolic syndrome and is strongly associated with these risk factors. Studies have shown that up to 75% of individuals with type 2 diabetes also have fatty liver. The presence of liver fat exacerbates insulin resistance and systemic inflammation, further elevating the risk of cardiometabolic events.
Enhancing Cardiometabolic Risk Assessment
Fatty liver, also known as hepatic steatosis, is a condition characterized by the accumulation of fat in the liver cells. Detected opportunistically in coronary artery calcium (CAC) scans, fatty liver is a significant marker for metabolic syndrome, diabetes, and an increased risk of cardiovascular disease (CVD).
AI-CVD™ leverages advanced imaging algorithms to quantify liver fat, providing a comprehensive assessment of an individual's cardiometabolic health.
AI-enabled opportunistic measurement of liver steatosis in coronary artery calcium scans predicts cardiovascular events and all-cause mortality: an AI-CVD study within the Multi-Ethnic Study of Atherosclerosis (MESA)
About one-third of adults in the United States have some degree of fatty liver, or hepatic steatosis, a condition in which hepatocytes have an excessive accumulation of fat, mainly in the form of triglycerides. The two most common forms of fatty liver diseases are non-alcoholic fatty liver disease (NAFLD) and alcohol-associated liver disease (ALD), but NAFLD is the more prevalent form in developed countries, often associated with obesity, insulin resistance, and metabolic syndrome.
Recently, the term MASLD (metabolic dysfunction associated steatotic liver disease) was introduced to emphasize the importance of metabolic dysfunction in NAFLD. MASLD is by far the most prevalent chronic liver disease globally. ALD is less prevalent, but associated with higher mortality. The case definition of MASLD requires liver steatosis and at least one of five cardiometabolic risk factors (obesity, prediabetes/diabetes, hypertension, high triglycerides, low HDL). The intertwining of MASLD, cardiovascular-kidney-metabolic (CKM) health, and cardiometabolic risk factors is a complex interaction that raises the question whether steatotic liver disease, is itself an independent predictor of cardiovascular disease (CVD).
This is an especially compelling area of investigation because the liver is the metabolic hub of the body and the processing site of triglycerides and lipoprotein particles—blood constituents that are well known drivers of cardiovascular risk.
Liver steatosis cannot be directly diagnosed using blood tests; however, a variety of imaging modalities are available for non-invasive detection. Although liver biopsy and magnetic resonance imaging (MRI) provide the most sensitive and accurate quantitation of liver fat, they are not suitable as an opportunistic screening tool for quantification of liver fat. Similarly, liver ultrasound elastography although more commonly available and less precise, is not suitable as an opportunistic screening test. Non-contrast computed tomography (CT) scans of the chest or abdomen provide for opportunistic detection of moderate-to-severe steatosis. Every year over 80 million abdominal and thoracic CT scans are performed in the US alone, including CT scans for coronary artery calcium (CAC) scoring and lung cancer screening, that can be used for opportunistic detection of liver steatosis. The resulting scans provide valuable data which can be extracted and used to detect moderate-to-severe steatosis at no additional X-ray or imaging cost to the patient and enable providers to gain important information. In CT imaging, liver attenuation measured in Hounsfield units (HU) is inversely related to liver fat content, with lower HU values indicating higher fat content. In particular, a mean density below 40 HU is correlated with at least 30% steatosis (moderate-to-severe) on histopathology. This method provides a reliable, reproducible indication of moderate-to-severe steatosis, making it a valuable tool in both clinical and research settings for evaluating metabolic health and its associated risks.
AI-CVD™ is a suite of AI software modules that enables opportunistic screening for multiple disorders in chest CT scans including non-contrast CAC, contrast-enhanced coronary CT angiography (CCTA) and low-dose lung cancer screening scans. We have previously published multiple studies on other outputs of AI-CVD™ from CAC scans in the Multi-Ethnic Study of Atherosclerosis (MESA), including bone mineral density, cardiac chambers volumetry, and left ventricular mass. We applied AI-CVD™ algorithms to measure hepatic steatosis in MESA CAC scans. Unlike previous CT-based AI-enabled evaluation of liver fat in which only a few spots of the liver are analyzed, our AI algorithms analyze the entire liver in the field of view.
We introduce and validate a novel metric, the liver attenuation index (LAI), which reports the percentage of voxels with attenuation below 40 HU (LAI) in non-contrast CAC scans. We evaluate the prognostic value of LAI for incident CVD events and all-cause mortality over 15 years in the Multi-Ethnic Study of Atherosclerosis (MESA). Additionally, we correlate LAI with the United States fatty liver index (US-FLI), a composite score designed for use in the US for all races/ethnicities, for indication of steatosis.
Figure 1: Examples of artificial intelligence-cardiovascular disease liver segmentations in a typical coronary artery calcium (CAC) scan along with the distribution of liver attenuation index (LAI) in Multi Ethnic Study of Atherosclerosis (MESA)-1.
Figure 6: 15-year cardiovascular disease (CVD) incidence by liver attenuation index (LAI) and Agatston coronary artery calcium (CAC) score.
Implications for Cardiovascular Risk Prediction
AI-enabled CT attenuation analysis of the entire liver visible in CAC scans provides opportunistic and actionable information for early detection of patients at elevated risk of CVD events and all-cause mortality. The clinical utility of incorporating LAI along with other opportunistic findings in CAC scans as part of the AI-CVD™ initiative to improve CVD risk prediction warrants investigation in other cohorts.
​
