Diagnosing certain conditions can be difficult for many reasons, one of which is the similarity between symptoms of numerous diseases. Doctors can mistake one disease for another, which then impacts treatment and patient outcomes. Research is being done to help correct these misdiagnoses, and recently, a development has been made for two severe cardiac conditions.
Artificial intelligence (AI) has been applied to ultrasound videos of the heart in order to distinguish between cardiac amyloidosis, hypertrophic cardiomyopathy, and other forms of left ventricular (LV) hypertrophy. The learnings from this process will be very helpful in the diagnosis of these severe cardiac conditions in the future. You can read the full study in JAMA Cardiology.
About the Study
Conducted by researchers from Cedars-Sinai Medical Center, this study aimed to improve the diagnosis of life-threatening cardiac conditions through the use of AI. Researchers utilized the records from two healthcare systems, which supplied them with ultrasound videos from 23,745 patients.
The AI was able to classify between the different cardiac disorders by measuring three things:
- LV wall thickness
- Posterior wall thickness
- LV diameter
Upon validating the diagnoses from the AI, the research team found that their system was accurate. For each of the three signs listed above, there was a mean absolute error of 1.7 mm, 1.8 mm, and 3.8 mm, respectively. Both cardiac amyloidosis and hypertrophic cardiomyopathy were diagnosed correctly.
Looking Forward
This new diagnostic method is extremely helpful, as it is more accurate than even the well-trained eye of a clinician. Using it could lead to earlier diagnoses, earlier and proper treatment, and better patient outcomes.
Find the source article here.
