Abstract
Background and purpose
The ICH-E14/S7B Q&As recommend individual correction for QT prolongation assessments in in vivo safety pharmacology studies. The template matching method is used in QT analysis, but the template waveforms are selected manually, and this can be time consuming when there are large variations and changes in electrocardiogram (ECG) waveforms. We developed a new QT analysis method using ECG templates generated automatically by artificial intelligence (AI) and performed continuous 24-hour QT analysis after moxifloxacin dosing in cynomolgus monkeys.
Methods
Moxifloxacin (100 mg/kg) was administered orally to 4 male telemetered cynomolgus monkeys. ECGs were recorded continuously from 2 hours before dosing until 24 hours after dosing, and QT intervals were analyzed using the new method, which involved algorithmic exclusion of waveforms with noise or arrhythmias and automatic generation of 12 templates by AI for QT analysis. The results obtained using the new method were compared with those obtained using the conventional method. We assessed practicality of our algorithm by measuring the time required by a technologist to perform QT analysis.
Results and Conclusion
The results by the newly developed method were equivalent to the results by manual analysis. In addition, the time required by technologist was within 30 minutes per case. As a result, this QT analysis method using AI-generated ECG templates suggest a potential for significantly reducing manual labor in this study.
