About My Health Careers Internship MedBlogs Contact us
AI can Identify Risk of Genetic Disease-Causing High Cholesterol

AI can Identify Risk of Genetic Disease Causing High Cholesterol

Font : A-A+

  • New AI-based technology can detect patients with familial hypercholesterolemia – a genetic disease that causes high blood cholesterol
  • The technology can accurately differentiate between different categories of patients with high cholesterol
  • It prevents unnecessary genetic testing of all patients, thereby saving both time and money

A new artificial intelligence (AI) based algorithm has been developed that can predict whether a patient is likely to have familial hypercholesterolemia (FH), reports a research team from Stanford University, California, USA.

Familial hypercholesterolemia (FH) is a genetic disease that results in drastic elevations in blood cholesterol that causes serious heart problems, which are sometimes fatal. FH is often misdiagnosed as garden-variety high cholesterol, which occurs due to an unhealthy diet and lack of exercise.


AI can Identify Risk of Genetic Disease Causing High Cholesterol

To tackle the problem of misdiagnosis of FH, the researchers came up with a novel AI-based technology that is capable of detecting a larger number of FH patients than conventional methods.

The study, published in npj Digital Medicine, a Nature Partner Journal (NPJ), was jointly led by Dr. Joshua W. Knowles, MD, PhD and Dr. Nigam H. Shah, MBBS, PhD.

Dr. Knowles is an Assistant Professor of Medicine (Cardiovascular Medicine) at the Stanford University Medical Center, California, USA. He also holds joint appointments at Bio-X, the Cardiovascular Institute, and the Maternal & Child Health Research Institute.

Dr. Shah is an Associate Professor of Medicine (Biomedical Informatics) and Biomedical Data Science at Stanford University, California, USA. He is also the Assistant Director of the Center for Biomedical Informatics Research at Stanford. He also holds joint appointments at Bio-X, the Stanford Cancer Institute, and the Wu Tsai Neurosciences Institute.

The lead author of the study was Dr. Juan Banda, PhD, who is currently an Assistant Professor of Computer Science at Georgia State University and a former Research Scientist at Stanford.

This project is part of a large initiative called Flag, Identify, Network, Deliver FH or FIND FH. This is a collaborative effort between Stanford Medicine and the Familial Hypercholesterolemia Foundation, which is a non-profit organization that focuses on AI, Machine Learning, and Big Data to engage FH patients and their families.

Familial Hypercholesterolemia and Risk of Heart Disease

The genetic disorder FH occurs due to a defect on chromosome 19. The defect prevents the body from clearing low-density lipoprotein (LDL), the so-called 'bad' cholesterol that clogs the arteries. It is believed that less than 10 percent of individuals actually realize that they have FH. An FH patient with high cholesterol is three-times more likely to develop heart disease at an earlier age than a patient having high cholesterol, but not FH. Moreover, an FH patient has a 10 times higher risk of developing heart disease compared to a normal individual.

Diagnosis of Familial Hypercholesterolemia: Drawbacks of Genetic Testing

Conventionally, FH is diagnosed by genetic testing to confirm whether a patient has inherited the FH mutation on chromosome 19. However, it is neither practicable, nor financially possible to test everyone for the FH mutation. In this regard, Shah said: "The problem is, the chance that someone seen in the Cardiology Clinic has this genetic condition is somewhere around 1 in 90 or 1 in 100, so it doesn't make sense to sequence every single person."

New AI Technology: A Game Changer

The new AI technology is a paradigm shift in the diagnosis of FH. The novelty of the new AI technology lies in its functionality. The technology works like a sieve, i.e. it filters all the medical data for each and every patient and captures only those patients who are likely to have the condition.

"Theoretically, when someone comes into the clinic with high cholesterol or heart disease, we would run this algorithm," Shah said. "If they're flagged, it means there's an 85 percent chance that they have FH. Those few individuals could then get sequenced to confirm the diagnosis and could start an LDL-lowering treatment right away."

How Was the AI Technology Developed?

The algorithm was developed based on patient data accessed from the electronic health database of Stanford's FH Clinic. The algorithm was trained to identify features that distinguish an FH patient from other patients. The health parameters that the algorithm analyzed included the following:
  • Family history
  • Current prescriptions
  • Lab test reports
  • Lipid profile
Thus, the sensitivity and predictive accuracy of the algorithm was increased manifold as it was trained on health data of actual FH patients. This data included those of 197 patients who had FH and 6,590 who didn't have FH. This allowed the program to accurately distinguish between the two datasets.

The outcome of this real patient data based training was that the algorithm was able to rank the patients according to their likelihood of having FH. A patient having a higher rank would be more likely to have the disease, compared to those having a lower rank.

Next Step: Taking the AI Technology to the Clinic

The clinical applicability of the new AI technology was rigorously tested in a real clinical setting. The algorithm was run on approximately 70,000 blinded medical records, which it had never encountered before. From the medical records flagged by the algorithm as being those of FH patients, the researchers sampled 100 records and found that the algorithm was 85 percent accurate in its prediction. Then the researchers, in collaboration with Geisinger Healthcare System, tested the algorithm on 466 FH patients and 5,000 patients without FH.

"The predictions came back with 85 percent accuracy, and we knew that many of the Geisinger patients had a confirmed FH diagnosis with genetic sequencing," Shah said. "So that's how we convinced ourselves that yes, this indeed works."


This new AI technology could fill the gap in the area of FH diagnostics, particularly when used in conjunction with family screening programs, which are regularly conducted by the FH Foundation.

This work is an excellent example of Stanford Medicine's commitment towards developing precision health for accurately diagnosing and preventing disease in healthy individuals, while treating those who are afflicted by the disease.

Reference :
  1. AI identifies risk of cholesterol-raising genetic disease - (http://med.stanford.edu/news/all-news/2019/04/ai-identifies-risk-of-cholesterol-raising-genetic-disease.html)

Source: Medindia

News A-Z
News Category
What's New on Medindia
Woman with Rare Spinal Cord Defect from Birth Sues Doctor
World AIDS Day 2021 - End Inequalities, End AIDS
View all

Medindia Newsletters Subscribe to our Free Newsletters!
Terms & Conditions and Privacy Policy.

More News on:
Oral Health And AIDS Cholesterol High Cholesterol AIDS/HIV AIDS/HIV - Epidemiology AIDS/HIV - Clinical Features AIDS/HIV - Health Education AIDS/HIV - Prevention And Transmission AIDS / HIV - Treatment AIDS/HIV- Lab Tests and Faqs 

Recommended Reading
Lifestyle Paths to Prevent Heart Disease
Heart disease can be of many types depending upon whether they involve the heart muscles or artery ....
High Density Lipoproteins - LDL/HDL and total/HDL ratio
Encyclopedia section of medindia gives details regarding LDL/HDL and total/HDL ratios...
Cholesterol is produced by the body (liver) and is essential for normal body functioning....
New Insights into Familial Hypercholesterolemia in Children and Adolescents
Treatment approaches for familial hypercholesterolemia, a hereditary genetic disorder in the ......
AIDS / HIV - Treatment
Encyclopedia section of medindia explains in brief about the treatment for AIDS/HIV...
"AIDS is an epidemic disease, a potentially preventable, deadly infection for which there is no cure...
AIDS/HIV - Clinical Features
Encyclopedia section of medindia gives general info about HIV Clinical Features...
AIDS/HIV - Epidemiology
AIDS or HIV is an epidemic disease, a potentially deadly infection that can be prevented with preca...
AIDS/HIV - Health Education
Encyclopedia section of medindia gives general info about AIDS information and health education....
AIDS/HIV - Prevention And Transmission
Encyclopedia section of medindia explains in brief about the prevention for AIDS/HIV...
High Cholesterol
High cholesterol level in blood are a risk for heart disease and stroke. They get deposited in the w...
Oral Health And AIDS
AIDS has taken on massive proportions in modern times. It is estimated that over 15 million people a...

Disclaimer - All information and content on this site are for information and educational purposes only. The information should not be used for either diagnosis or treatment or both for any health related problem or disease. Always seek the advice of a qualified physician for medical diagnosis and treatment. Full Disclaimer

© All Rights Reserved 1997 - 2021

This site uses cookies to deliver our services. By using our site, you acknowledge that you have read and understand our Cookie Policy, Privacy Policy, and our Terms of Use