In-band artifact and noise present a major obstacle to extracting accurate, reliable, and repeatable information from ambulatory recordings of subcutaneous and surface ECGs. With more than 4 million patients a year experiencing ambulatory evaluations for cardiac arrhythmia, in-band artifact and noise is an urgent, wide-reaching challenge. In clinical care, for example, the high incidence of false positive arrhythmia detections can result in the need for expensive manual over-read, leading to poor operational efficiencies and higher cost of care. In addition, noise can mask P-waves, thereby preventing evaluation of atrial activity, an important factor in making the correct diagnosis and pursuing the most effective treatment strategy. This issue is also important in clinical research on drug safety and effectiveness (the basis for assessing public health impacts of new and approved medications) because noise introduces variability in interval measurements which, in turn, increases sample size and cost and also compromises the quality of information. Successful completion of this proposed multiphase SBIR research program will result in a universal ECG sensing lead that will provide relatively noise free signals for recording on commercially available ambulatory ECG monitors. This will result in a significant improvement in the quality of diagnostic information available to physicians that care for the 4 million patients that seek treatment for cardiac arrhythmias each year in the US. The improved diagnostic information will lead to more informed therapeutic decisions, improved patient quality of life, and higher quality care delivered at a lower cost. Specifically, Phase I will focus on researching approaches to optimization of a novel, proven, and patent- pending algorithm for removing in-band noise from ECG signals without compromising underlying signal morphology and fidelity, referred to as Multi-Domain FilteringTM. Innovative approaches will be researched to implement critical mathematical functions of this algorithm using efficient numerical methods and computational techniques. The algorithm will be implemented in embedded code and tested on targeted microprocessor systems to evaluate computational efficiency, power consumption, noise reduction performance, and fidelity. Testing will also be performed to estimate the impact of denoising on arrhythmia event detection accuracy when used with state-of-the-art ambulatory ECG monitoring devices.
Successful completion of this proposed multiphase SBIR research program will result in a universal ECG sensing lead that will provide relatively noise free signals for recording on commercially available ambulatory ECG monitors. This will result in a significant improvement in the quality of diagnostic information available to physicians that care for the 4 million patients that seek treatment for cardiac arrhythmias each year in the US. The improved diagnostic information will lead to more informed therapeutic decisions, improved patient quality of life, and higher quality care delivered at a lower cost.
| Brockway, Marina; Fossa, Anthony A; Mason, Jay W (2018) Comparison of Two Highly Automated ECG Algorithms for Detection of Drug-Induced Cardiac Ion Channel Block. Clin Pharmacol Ther 104:356-363 |
