India
November 1, 2021
Intel®
November 1, 2021Cardiology
Artificial Intelligence-driven Multiple Protein Blood Test for Heart Disease
Prevencio, Inc. introduced a next-generation exact blood test, HART KD, to recognize Kawasaki disease. The company cooperated with Seattle Children’s Research Institute and leveraged its Artificial Intelligence (AI) driven HART platform. A unique key is using an AI-driven algorithm, or weighting of each protein, to enhance the accuracy of an unmatched ROC-AUC of 92% compared to nonspecific, single protein tests. This approach addresses the vast need for a diagnostic blood test to distinguish Kawasaki disease from other infectious and inflammatory conditions.(1)
Prevencio, Inc. introduced a next-generation exact blood test, HART KD, to recognize Kawasaki disease. The company cooperated with Seattle Children’s Research Institute and leveraged its Artificial Intelligence (AI) driven HART platform.
A unique key is using an AI-driven algorithm, or weighting of each protein, to enhance the accuracy of an unmatched ROC-AUC of 92% compared to nonspecific, single protein tests. This approach addresses the vast need for a diagnostic blood test to distinguish Kawasaki disease from other infectious and inflammatory conditions.(1)
"Selfies"could be used to detect heart disease
Selfies will be the cheapest way to detect heart disease; a research study supports artificial intelligence utilization for facial recognition to detect heart disease in the general population. It showed it is reasonable to employ a deep learning computer algorithm to identify coronary artery disease by analyzing four photographs of a person’s face. The Brain and Cognition Institute in the Department of Automation at Tsinghua University, Beijing, enrolled 5,796 patients from eight hospitals in China. The patients underwent imaging procedures to investigate their blood vessels, such as coronary angiography. They found that the algorithm’s performance compares to existing methods of predicting heart disease risk. In the validation group of patients, the algorithm correctly detected heart disease in 80% of cases.(2)
Selfies will be the cheapest way to detect heart disease; a research study supports artificial intelligence utilization for facial recognition to detect heart disease in the general population. It showed it is reasonable to employ a deep learning computer algorithm to identify coronary artery disease by analyzing four photographs of a person’s face.The Brain and Cognition Institute in the Department of Automation at Tsinghua University, Beijing, enrolled 5,796 patients from eight hospitals in China.
The patients underwent imaging procedures to investigate their blood vessels, such as coronary angiography. They found that the algorithm’s performance compares to existing methods of predicting heart disease risk. In the validation group of patients, the algorithm correctly detected heart disease in 80% of cases.(2)
Reducing the risk of blood clots with artificial intelligence
ARTORG, Center for Biomedical Engineering Research at the University of Bern, has successfully identified a mechanism responsible for clot formation in patients with artificial heart valves. The research team demonstrated that the shape of the prosthetic valve’s flow-regulating flaps leads to considerable turbulence in blood flow. They coupled complex mathematical models of hydrodynamic stability with sophisticated computer simulations and artificial intelligence. The group further studied how to improve heart valves and discovered slight alterations in its flaps that exhibited non-turbulent blood flow. This laminar flow resembles more to a healthy heart. This AI-designed mechanic-improvement of heart valves aims for their recipients to no longer need anticoagulant therapy. These results could lead to an anticoagulant-free life, in turn decreasing the comorbidities associated with these therapies.(3)
ARTORG, Center for Biomedical Engineering Research at the University of Bern, has successfully identified a mechanism responsible for clot formation in patients with artificial heart valves. The research team demonstrated that the shape of the prosthetic valve’s flow-regulating flaps leads to considerable turbulence in blood flow. They coupled complex mathematical models of hydrodynamic stability with sophisticated computer simulations and artificial intelligence.
The group further studied how to improve heart valves and discovered slight alterations in its flaps that exhibited non-turbulent blood flow. This laminar flow resembles more to a healthy heart. This AI-designed mechanic-improvement of heart valves aims for their recipients to no longer need anticoagulant therapy. These results could lead to an anticoagulant-free life, in turn decreasing the comorbidities associated with these therapies.(3)
Artificial Intelligence May Accelerate Heart Failure Diagnosis
Artificial intelligence-enhanced electrocardiogram (ECG) may properly detect heart failure in patients assessed for dyspnea in the ER. The group tested the accuracy of AI-enabled ECG to recognize left ventricular systolic dysfunction (LVSD) in ER patients with shortness of breath compared with results of biomarker blood tests (natriuretic peptides). Researchers used data on thousands of patients as training for the computer to distinguish between the ECGs of recently diagnosed LVSD from those without it. Standard ECG recordings can be analyzed in around ten seconds with the AI software application to identify the heart pathology. Researchers used the AI tool to the ECGs of 1,606 patients who had gotten an ECG and blood testing in the emergency, eventually followed by definitive testing utilizing an echocardiogram. The researchers found that AI-enhanced ECG was superior to regular blood tests. (4)
Artificial intelligence-enhanced electrocardiogram (ECG) may properly detect heart failure in patients assessed for dyspnea in the ER. The group tested the accuracy of AI-enabled ECG to recognize left ventricular systolic dysfunction (LVSD) in ER patients with shortness of breath compared with results of biomarker blood tests (natriuretic peptides). Researchers used data on thousands of patients as training for the computer to distinguish between the ECGs of recently diagnosed LVSD from those without it.
Standard ECG recordings can be analyzed in around ten seconds with the AI software application to identify the heart pathology. Researchers used the AI tool to the ECGs of 1,606 patients who had gotten an ECG and blood testing in the emergency, eventually followed by definitive testing utilizing an echocardiogram. The researchers found that AI-enhanced ECG was superior to regular blood tests. (4)
Machine Learning Predicts Long-Term Risk of Heart Attack
Machine learning tools could predict patients long-term risk of heart attacks and cardiac deaths better than typical methods employed by cardiologists. Coronary artery calcium (CAC) scoring with non-contrast computed tomography is regularly used for cardiovascular risk stratification. The team observed that machine learning scores, combined clinical data, and quantitative CT analysis after CAC scoring could considerably improve risk assessment. Over fifteen years, 76 of the 1,912 subjects studied presented an event of infarction or cardiac death. The data revealed that the subjects predicted machine learning scores correlate accurately with events actual distribution. High machine learning-estimated risk was significantly associated with a higher incidence of cardiac events. Moreover, the atherosclerotic cardiovascular risk disease score, recognized as the standard clinical assessment, overestimates the risk of fatal events in the higher-risk categories. In contrast, the algorithm demonstrated more reliable performance in this category.(5)
Machine learning tools could predict patients long-term risk of heart attacks and cardiac deaths better than typical methods employed by cardiologists. Coronary artery calcium (CAC) scoring with non-contrast computed tomography is regularly used for cardiovascular risk stratification. The team observed that machine learning scores, combined clinical data, and quantitative CT analysis after CAC scoring could considerably improve risk assessment.
Over fifteen years, 76 of the 1,912 subjects studied presented an event of infarction or cardiac death. The data revealed that the subjects predicted machine learning scores correlate accurately with events actual distribution. High machine learning-estimated risk was significantly associated with a higher incidence of cardiac events. Moreover, the atherosclerotic cardiovascular risk disease score, recognized as the standard clinical assessment, overestimates the risk of fatal events in the higher-risk categories. In contrast, the algorithm demonstrated more reliable performance in this category.(5)
AI Enabled Wireless Cardiac Ultrasound Will be Used by Astronauts in Space
In May 2021, the Israel Space Agency selected UltraSight, a digital health company, to conduct the cardiac ultrasound study aboard the Rakia space mission. The study will use a combination of real-time AI guidance software and high-performance, wireless pocket-sized ultrasound to record high-quality images of an astronaut’s heart in microgravity. This study will help scientists gain insights into the potential impact of reduced gravity forces on the heart. The main goal of this study is to demonstrate how portable cardiac ultrasound is and how AI algorithms can develop successful scannings with very limited training.(6)
In May 2021, the Israel Space Agency selected UltraSight, a digital health company, to conduct the cardiac ultrasound study aboard the Rakia space mission. The study will use a combination of real-time AI guidance software and high-performance, wireless pocket-sized ultrasound to record high-quality images of an astronaut’s heart in microgravity.
This study will help scientists gain insights into the potential impact of reduced gravity forces on the heart. The main goal of this study is to demonstrate how portable cardiac ultrasound is and how AI algorithms can develop successful scannings with very limited training.(6)
AI may detect heart disease with the help of gut microbiome bacteria
Scientists have designed a machine learning algorithm using gut microbiome data sets to identify cardiovascular disease. Researchers from the department of physiology and pharmacology at the University of Toledo used artificial intelligence to screen stool samples. The study assessed a machine learning model’s diagnostic efficacy utilizing five algorithms that included unplanned forest plots and neural networks. They documented that the machine learning model identified 39 bacterial taxonomies between participants with and without CVD. Using these differential microbiome characteristics, the AI explored the taxonomic groups’ differences between sick and healthy patients. The researchers reported that the accuracy increased for the top 500 high-variance features of operational taxonomic units for machine learning models.(7)
Scientists have designed a machine learning algorithm using gut microbiome data sets to identify cardiovascular disease. Researchers from the department of physiology and pharmacology at the University of Toledo used artificial intelligence to screen stool samples. The study assessed a machine learning model’s diagnostic efficacy utilizing five algorithms that included unplanned forest plots and neural networks.
They documented that the machine learning model identified 39 bacterial taxonomies between participants with and without CVD. Using these differential microbiome characteristics, the AI explored the taxonomic groups’ differences between sick and healthy patients. The researchers reported that the accuracy increased for the top 500 high-variance features of operational taxonomic units for machine learning models.(7)
AI and Virtual Realty in Cardiology
The recently FDA-cleared Holoscope-i system, developed by the Israeli start-up company RealView Imaging, is advancing the field of interactive live holography. The AI technology projects high-quality 3D/4D holograms of a patient’s heart similar to 3D printing and optically indistinguishable from actual objects. RealView Imaging has developed the newest generation of reality technology, producing a super realistic volumetric representation of the imaging data through AI. Based on a Nobel-prize-winning holographic technology, this new technology provides all 3D visual depth features by projecting all the image points simultaneously at multiple depth locations, like a 3D printer of light points.(8)
The recently FDA-cleared Holoscope-i system, developed by the Israeli start-up company RealView Imaging, is advancing the field of interactive live holography. The AI technology projects high-quality 3D/4D holograms of a patient’s heart similar to 3D printing and optically indistinguishable from actual objects.
RealView Imaging has developed the newest generation of reality technology, producing a super realistic volumetric representation of the imaging data through AI. Based on a Nobel-prize-winning holographic technology, this new technology provides all 3D visual depth features by projecting all the image points simultaneously at multiple depth locations, like a 3D printer of light points.(8)
References