The PhysioNet/Computing in Cardiology Challenge 2015: Reducing false arrhythmia alarms in the ICU

Abstract

High false alarm rates in the ICU decrease quality of care by slowing staff response times while increasing patient delirium through noise pollution. The 2015 Physio-Net/Computing in Cardiology Challenge provides a set of 1,250 multi-parameter ICU data segments associated with critical arrhythmia alarms, and challenges the general research community to address the issue of false alarm suppression using all available signals. Each data segment was 5 minutes long (for real time analysis), ending at the time of the alarm. For retrospective analysis, we provided a further 30 seconds of data after the alarm was triggered. A collection of 750 data segments was made available for training and a set of 500 was held back for testing. Each alarm was reviewed by expert annotators, at least two of whom agreed that the alarm was either true or false. Challenge participants were invited to submit a complete, working algorithm to distinguish true from false alarms, and received a score based on their program’s performance on the hidden test set. This score was based on the percentage of alarms correct, but with a penalty that weights the suppression of true alarms five times more heavily than acceptance of false alarms. We provided three example entries based on well-known, open source signal processing algorithms, to serve as a basis for comparison and as a starting point for participants to develop their own code. A total of 38 teams submitted a total of 215 entries in this year’s Challenge.

Enhancement of Life:threatening Arrhythmias Discrimination in the Intensive Care Unit with Morphological Features and Interval Features Extraction via Random Forest Classifier

“…In order to detect the pulse from the ABP and PPG signals, we used the open-source ABP pulse detection algorithm which is available in PhysioNet [6]. Since, ECG channels in the training set showed that the available ECG signals existing various disturbances such as clipping of the QRS complexes, and other noises, beats detected should be accomplished from different channels and signals and then compared them with each other.…”

“…In this section of algorithm, we focused on 5 life threatening arrhythmias, namely extreme tachycardia (ETC), ventricular fibrillation or flutter (VFB), asystole (ASY), extreme bradycardia (EBR), or ventricular tachycardia (VTA) and based on the clinical definition and different characteristics of each of them, signal quality features are used for them [6]. In our algorithm, the heuristic thresholding of each ABP pulse are computed with the function of abpfeature, and also, since blood pressure variation can occur during asystole and ventricular tachycardia, these features can detect these abnormalities accurately.…”

“…The task set by the Physionet "CinC Challenge 2015" was to create an algorithm to decline the false alarm ratio of 5 lifethreatening arrhythmias. If a given ICU alarm does not match into one of the 5 alarm types of data, the alarm will be categorized as a false alarm [6]. Since the quality and features of the different waveforms are a major issue in most cases, the ability of the algorithm to enhance the [7,8].…”

Assessing Effect of Beat Detector on Detection Dependent Signal Quality Indices

“…Computing For our analysis on ECG SQIs we used the ECG signals in the CinC 2015 training set [5]. We focused our analysis on asystole, bradycardia, and tachycardia records since ventricular tachycardia and ventricular fibrillation/flutter result in extreme modification of the ECG waveform.…”

“…Similar approaches can be used in automatic algorithms to reduce false alarms. For example, the Computing in Cardiology (CinC) 2015 challenge focused specifically on reducing false arrhythmia alarms in the ICU using patient monitoring algorithms which use multimodal physiological waveforms [5]. However combining information from multiple physiological signals introduces a new potential risk of adding noise artifacts from a low quality signal onto information from a high quality signal (e.g.…”