Participants in the annual challenges, which cover topics ranging from physiological signal processing and analysis to forecasting and modelling clinically important events and processes, discuss their approaches to the posed problems during dedicated scientific sessions at the Annual International Computing in Cardiology (CinC) Conference.
In 2011, the PhysioNet/CinC Challenge was entitled "Improving the quality of ECGs [electrocardiograms] collected using mobile phones". A newly published focus issue of Physiological Measurement brings together papers presented at the 38th CinC Conference, hosted in Hangzhou, China, in September 2011 by Zhejiang University. The focus issue, Signal quality in cardiorespiratory monitoring, also includes research addressing the broader question of signal quality metrics in cardiorespiratory monitoring.
Improving ECGs
The PhysioNet/CinC Challenge 2011 called for the development of an efficient algorithm that can run in near real-time within a mobile phone and that can provide useful feedback during the acquisition of a diagnostically usable 12-lead ECG.
As a minimum requirement, the algorithm should be able to indicate within a few seconds, while the patient is still present, whether the ECG is of adequate quality for interpretation or whether another recording needs to be made. Ideally, the software should identify common problems (such as misplaced electrodes, poor skin-electrode contact, external electrical interference and patient motion artefacts) and either compensate for these deficiencies or provide guidance for correcting them.
Writing in an Editorial for the Physiological Measurement focus issue, Gari Clifford from the UK's University of Oxford and George Moody from Massachusetts Institute of Technology (Cambridge, MA) explained the rationale behind this challenge:
"The ECG is among the most useful tools for diagnosing cardiovascular diseases (CVD), the most frequent cause of death worldwide. Although both CVD and mobile phones are ubiquitous, adequate primary healthcare is not. Many rural populations around the world rely on clinics staffed by lay volunteers to identify those in need of secondary care by healthcare professionals in distant city hospitals. It is increasingly feasible to provide rural clinics with inexpensive medical instruments such as electrocardiographs that transmit digital ECGs to smartphones for storage and display.
"These devices extend the reach of diagnosticians to remote areas, but without some means of quality control, technology alone cannot deliver consistently usable information to those able to interpret it. Methods that improve the quality of data collected result in better usage of the scarcest resource, clinical expertise. The growing interest in mHealth to provide point-of-care diagnostics to underserved populations is also driving the desire to leverage the power of smartphones to insert intelligence into medical data acquisition."
For the challenge, PhysioNet provided a data set of 2000 12-lead ECG records. Clifford and Moody point out that the use of shared data sets in many of the papers in this focus issue enables readers to make objective comparisons of the performance of the described methods. Such comparisons can point the way to further advances towards assessment of quality in ECGs and other physiologic signals.
Seven challenge participants, including the authors of several of the Physiological Measurement papers, have also contributed their algorithms as open-source software for further study; these can be found here.
Source:
medicalphysicsweb.org
