The Technology
Diagnosis of the cardiac condition involves the analysis of an ECG image through a visual examination of the printed physical output of the recording. The technology enables an automated analysis employing innovative deep-learning methods. The research, enables immediate analysis of ECG images, providing a list of detectable heart diseases and their likelihoods. The computational learning-based system utilizes convolutional neural networks, The technology can identify more than 52 different diseases with an accuracy of 90% , surpassing other methods and the accuracy achieved by cardiologists. The system, developed based on this method, is entirely general and capable of processing new data files. It can incorporate work for additional diseases, learn from new ECG formats not previously trained on, and utilize advanced image processing capabilities to directly process prints obtained from ECG machines. The system can function even when the captured image is taken by a mobile camera (possibly a smartphone), accommodating shadows, distortions, and interferences.
Advantages
-
Can work with distorted and shadow pictures.
-
Automatic improvement of the system with more data, including automatic adaption to new formats.
-
High accuracy
Applications and Opportunities
-
AI-assisted and automatic ECG analysis and disease classification
Business Development Contacts