Output details
15 - General Engineering
University of Oxford
Automated QT Analysis That Learns from Cardiologist Annotations
The Food and Drug Administration (FDA) and European regulatory agencies require all new drugs to be evaluated in cardiac safety (Phase 1) studies. This paper describes a semi-supervised machine learning algorithm to quantify timing and shape changes in the electrocardiogram (ECG) waveform. The intellectual property (IP) described in the paper was protected by patent applications which have resulted in three granted US patents (US7941209, US8290575 and US8332017). It represents the core of the BioQT commercial system, which is now regularly used by pharmaceutical companies in their submissions to the FDA
(see http://www.obsmedical.com/products/bioqt).