Artificial Intelligence Identifies HCM in Children and Adolescents

A study by researchers from Mayo Clinic recently published in the International Journal of Cardiology found that a deep learning artificial intelligence algorithm using a standard 12 lead electrocardiogram was able to detect hypertrophic cardiomyopathy in young people with impressive accuracy. This accuracy was particularly strong among adolescents aged 15 – 18.  

Mayo has been looking at artificial intelligence for its potential to screen populations for HCM for some time now.  Here is a previous HCMBeat story about Mayo’s work on artificial intelligence from February 2020. 

 

 

 

Different Treatment for Non-Genetic Hypertrophic Cardiomyopathy?

Many HCM patients, perhaps even the majority, are currently unable to identify the specific gene behind their HCM through genetic testing.  Despite this obvious difference, family screening, risk stratification and treatment standards are no different for patients who carry a HCM gene and those who do not have identified gene(s).

A recent article published in Circulation suggests that there ARE differences which should result in different treatment for this subset of patients.

In particular, non-sarcomere positive patients:

  • have a better prognosis, with lower rates of heart failure, sudden death, atrial fibrillation and stroke
  • Have lower incidence of family members affected by the disease
  • Are more likely to have additional medical conditions such as obesity, hypertension and diabetes

The article by Dr. Hugh Watkins, a British HCM and genetic expert, suggests that:

  • the risk to first degree relatives of this type of HCM patient is less than 50% and therefore, there is less need for repeated screening of relatives
  • Hypertension should be treated more aggressively in these patients.

Read more about non-genetic HCM here and more about screening these patients here on HCMBeat.

Risk Assessment in HCM Children

A recent study published in the European Journal of Preventative Cardiology found that a 12-lead electrocardiogram (EKG) was not useful as a screening tool to determine which children were at increased risk of sudden death and therefore, a candidate to receive an implantable defibrillator. 

The full article can be found here.

Crypts Sign of HCM? Study Says No

According to a recently published study by doctors in Copenhagen, Denmark, myocardial crypts (clefts, cracks or fissures in the myocardium) are found in the general population. Therefore, this article concludes that crypts seen on scans of the heart are not necessarily an indicator of HCM and do not warrant further investigation. 

This paper is a departure from a 2012 paper by doctors at Tufts, which concluded that myocardial crypts were associated with HCM, and that they were often found in relatives of HCM patients found to be gene positive for HCM, but lacking the hallmark thickening of the ventricle. 

Here is an example of what the crypts look like on MRI.

2020 AHA/ACC HCM Diagnosis & Treatment Guidelines Released – Updated With New Links

The highly anticipated 2020 American Heart Association/American College of Cardiology Guidelines for the Diagnosis and Treatment of Patients with Hypertrophic Cardiomyopathy have been released.

This document, drafted with reference to published HCM literature, and with input from a committee of HCM experts with broad expertise, updates the prior version published in 2011.  It contains clinical practice guidelines for the broad spectrum of issues which may confront medical professionals as they approach the diagnosis and treatment of patients and families affected by hypertrophic cardiomyopathy.

Continue reading “2020 AHA/ACC HCM Diagnosis & Treatment Guidelines Released – Updated With New Links”

Could Artificial Intelligence Be Useful for HCM Screening?

A study by researchers from Mayo Clinic published this week in the Journal of the American College of Cardiology found that an artificial intelligence algorithm was able to detect hypertrophic cardiomyopathy, commonly known as HCM, from EKG results with impressive accuracy, particularly among younger patients.

In order to “teach” the computer, the researchers used digital 12-lead ECGs from 2,448 patients with HCM along with 51,153 age- and sex-matched controls. The technology was then tested on 612 HCM patients and 12,788 controls.

The findings showed that the technology was able to identify HCM in a high number of cases, even where the EKG appeared “normal” to the human eye.

The researchers believe that this technology, when refined, may prove to be an efficient tool for HCM screening in the future. The team plans to continue testing the technology in greater subject samples in order to further refine its performance.

See also:

Mayo Clinic News Network

Health Analytics

 

Should Children from HCM Families be Screened Earlier?

A recent study by doctors at Toronto’s Hospital for Sick Children suggests that current screening guidelines for children from HCM families are inadequate and should instead recommend earlier screening exams. In the U.S., screening begins at age 12 pursuant to American College of Cardiology (ACC)/American Heart Association (AHA) guidelines.  In Europe, screening begins at age 10 pursuant to the European Society of Cardiology (ESC) guidelines.

Continue reading “Should Children from HCM Families be Screened Earlier?”

Can a Smartwatch Detect HOCM?

According to a limited study recently published in Nature, researchers were able to detect obstructive HCM (HOCM) using a noninvasive optical sensor contained in many commercial smartwatches.

How the Technology Works

These watches used photoplethysmography, a noninvasive optical method used to detect blood volume changes in the microvascular bed at the skin surface.  The same technology is used in clinical pulse oximeters and is now widely incorporated in commercial smartwatches that have heart rate detection.

Continue reading “Can a Smartwatch Detect HOCM?”

When Do You Screen Your Kids For HCM?

A recent study published in Circulation suggests that clinical testing of kids who are first degree family members of HCM patients (i.e. siblings and children of those who have already been diagnosed with HCM) could be improved by starting testing at a younger age. And, genetic testing should further improve diagnosis and treatment for this group.

Continue reading “When Do You Screen Your Kids For HCM?”

Do HCM Family Screening Protocols Need Adjustment?

A recent editorial published in Circulation: Genomic and Precision Medicine suggests that current HCM screening protocols may need adjustment to account for recent findings by a study by researchers in the Netherlands.  The Dutch study, published in the same journal, found that of 620 relatives of HCM patients who underwent genetic testing, 43% were found to be genetically positive for HCM, while 30% were diagnosed with HCM at the initial screening. 16% more went on to develop HCM during 7 years of repeated cardiac evaluation.

On the other hand, the 57% of relatives found to be genotype-negative were released from clinical HCM follow-up.

The Australian authors of the editorial, Semsarian and Ingles, note that current screening protocols would have failed to identify the 6 children (15%) who were diagnosed under the age of 12, half of which had a particularly malignant family history.

Additionally, few teens were diagnosed with HCM, which stands in contrast to current opinion that HCM is most likely to develop during adolescence. Indeed, most newly diagnosed family members were older than the age of 36, with 44% being over the age of 50.

Lastly, Semsarian and Ingles note their concern with general utilization of the Dutch practice of releasing a gene negative family member from serial follow up since the impact of all genes which have a role in causing HCM is not yet known while new genes which may cause HCM are still being identified. 

Semsarian and Ingles also note that the Dutch patient sample differs from more typical patient populations found in the U.S. and Australia where causes of HCM are more diverse and cannot be easily tied to a specific gene.