scholarly journals Evaluation of the device independent nature of a photoplethysmography-deriving smartphone app

EP Europace ◽  
2021 ◽  
Vol 23 (Supplement_3) ◽  
Author(s):  
H Gruwez ◽  
S Evens ◽  
T Proesmans ◽  
C Smeets ◽  
P Haemers ◽  
...  
Keyword(s):  
Atrial Flutter ◽  
Sensitivity And Specificity ◽  
Heart Rhythm ◽  
Smartphone Application ◽  
Detection Algorithm ◽  
Interquartile Range ◽  
Detection Algorithms ◽  
Funding Sources ◽  
Successful Measurement ◽  
Sensitivity Specificity

Abstract Funding Acknowledgements Type of funding sources: None. Background Smartphone apps using photoplethysmography (PPG) technology enable digital heart rhythm monitoring through their built-in camera, without the need for additional, specific, or costly hardware. This may positively impact the availability and scalability of remote monitoring. However, the diversity of smartphone specifications on the consumer market may raise concerns regarding the robustness of AF detection algorithms between various devices. Purpose To study the device independency of AF detection performance by a PPG-based smartphone application. Methods Patients from the cardiology department were consecutively enrolled. Patients were handed 7 iOS models and 1 Android model and were asked to consecutively perform one PPG measurement per device. A 12-lead electrocardiogram (ECG) was collected during the same consultation and interpreted by a cardiologist as reference diagnosis. To allow an objective comparison across the devices, patients who failed to perform one successful measurement on each device were excluded. Additional exclusions were atrial flutter rhythms and insufficient quality results. Sensitivity, specificity and accuracy were calculated with respect to the reference diagnosis. McNemar’s analysis was used for the head-to-head comparison of the sensitivity and specificity of the proprietary algorithm on the different smartphone devices. Results A total of 150 patients participated in the study with a median CHA2DS2-VASc score of 3 (interquartile range: 1-5). The median age of the study population was 70 (interquartile range: 56-79) years. In total, 54.7% of the population was male and the AF-prevalence was 35.3%. After the exclusion of patients with atrial flutter (n = 14) and patients who did not successfully perform a PPG measurement on each device (n = 5), diagnostic-grade results of 131 patients were used to calculate the performance of the proprietary algorithm. The sensitivity and specificity of the AF detection algorithm ranged from 90.9% (95% CI 75.7-98.1) to 100.0% (95% CI 91.0-100) and 94.5% (95% CI 86.6-98.5) to 100.0% (95% CI 94.6-100), respectively. The overall accuracy across the devices ranged from 94.4% (95% CI 88.3-97.9) to 99.0% (95% CI 94.6-100). Head-to-head comparisons of the results did not reveal significant differences in sensitivity (P = 0.125-1.000) or specificity (P = 0.375-1.000) of the proprietary AF detection algorithm among the different devices. Conclusion This study demonstrated the device-independent nature of the PPG-deriving smartphone application with respect to 12-lead ECG diagnosis.

scholarly journals A comparison of over-the-counter available smartwatches and devices for electrocardiogram based detection of atrial fibrillation

2021 ◽  
Vol 2 (4) ◽  
Author(s):  
J Scholten ◽  
A Mahes ◽  
J R De Groot ◽  
M M Winter ◽  
A H Zwinderman ◽  
...  
Keyword(s):  
Atrial Fibrillation ◽  
Atrial Flutter ◽  
Heart Rhythm ◽  
Detection Algorithm ◽  
Left Knee ◽  
Funding Sources ◽  
Cardiology Department ◽  
Normal Heart Rate ◽  
Golden Standard ◽  
Sensitivity Specificity

Abstract Background There is an increasing number of smartwatches and devices commercially available that can generate and automatically interpret an electrocardiogram (ECG). Such devices have an enormous potential to improve population screening and telemonitoring of atrial fibrillation (AF). Purpose There is limited data on the sensitivity, specificity and interpretability of these devices and comparative studies are lacking. Our purpose was to compare three frequently used devices for AF detection. Methods We performed a single-center, prospective study in consecutive patients with AF presenting for electrical cardioversion (ECV). We collected a standard 12-lead ECG recording immediately followed by four times 30 seconds of ECG recordings from different devices for every patient prior to the ECV. These paired measurements were considered simultaneous. If the ECV was performed, the same measurements were repeated afterwards. The standard 12L-ECGs were interpreted by a cardiologist and used as golden standard for heart rhythm. The different devices used for the 30 second ECGs were: Withings Move ECG (lead I), Apple Watch series 5 (lead I), Kardia Mobile 6L (six leads) and Withings/Apple (1:1 ratio) on left knee (lead II). Sensitivity and specificity were determined for each AF detection algorithm excluding patients with atrial flutter (AFL) or uninterpretable ECGs. In addition, proportions of uninterpretable ECGs were determined including all patients and including only patients with sinus rhythm (SR) and compared between devices using McNemar's test. Results A total of 220 patients were included (age 70±10 years, female 35%, first ECV 44%) and in total 415 12-lead ECGs were performed (45% SR, 45% AF, 10% AFL). The sensitivity/specificity were overall similar for all devices (Withings 98%/95%, Apple 94%/98%, Kardia 99%/91%. P>0.05 for all). In detail, Kardia was the most sensitive test with highest proportion of suspected AF (57%) whereas Apple was the most specific, as shown by the highest proportion of normal heart rate results by the device (55%, P=0.003 compared to Kardia (43%)). Overall, Withings, Apple and Kardia had a comparable proportion of uninterpretable ECGs (20%, 20%, 24%, respectively. P>0.05 for all). Lead II had higher proportion of uninterpretable ECGs (32%, p<0.01 compared to all). More specifically, Kardia had a higher rate of uninterpretable ECGs in those with SR (P<0.05 compared to Withings (lead I) and Apple (lead I)). Conclusion In all devices, we found sensitivity/specificity for AF detection between 91%-99%, better than previous studies reported, and 20–24% of uninterpretable ECGs. Kardia was the most sensitive device, but less useful to rule out atrial fibrillation whereas Apple had numerically highest specificity. We aim to further evaluate both cardiologist interpretation and accuracy of atrial flutter detection using different leads to inform clinical use. Funding Acknowledgement Type of funding sources: Public hospital(s). Main funding source(s): Tergooi Cardiology department, J.P. Bokma was supported with a research grant by Amsterdam Cardiovascular Sciences Overview and comparison

scholarly journals Head-to-head comparison of proprietary PPG and single-lead ECG algorithms for atrial fibrillation detection

EP Europace ◽  
2021 ◽  
Vol 23 (Supplement_3) ◽  
Author(s):  
H Gruwez ◽  
S Evens ◽  
T Proesmans ◽  
C Smeets ◽  
P Haemers ◽  
...  
Keyword(s):  
Atrial Fibrillation ◽  
Atrial Flutter ◽  
Sensitivity And Specificity ◽  
Diagnostic Performance ◽  
Interquartile Range ◽  
Smartphone App ◽  
Smartphone Apps ◽  
Handheld Device ◽  
Detection Algorithms ◽  
Quality Measurements

Abstract Funding Acknowledgements Type of funding sources: None. Background Population based screening for atrial fibrillation (AF) has been suggested to reduce stroke. Photoplethysmography (PPG) deriving smartphone apps and single-lead electrocardiography (ECG) tools are attractive devices for screening due to their low cost, convenience, and accessibility. Automated algorithm analysis can serve as pre-screening or remote monitoring for AF, while confirmation on an ECG trace >30s is required to establish the diagnosis. This work directly compares the performance of proprietary algorithms on PPG vs single-lead ECG for the detection of AF. Purpose To evaluate and compare the diagnostic performance of a PPG-deriving smartphone app and a single-lead ECG-deriving handheld device for AF detection. Methods Patients were recruited from the cardiology ward. After obtaining written informed consent, demographic and medical information were collected. Patients were instructed to perform one measurement using a pulse-deriving smartphone app and one via a single-lead ECG handheld device. A 12-lead electrocardiogram (ECG) was collected and interpreted by a cardiologist as gold standard. Patients with atrial flutter were excluded, with additional exclusions for insufficient quality measurements and unsuccessful measurements resulting due to technical errors. Unclassified single-lead ECG measurements were handled as test-negative. Sensitivity, specificity and accuracy were calculated with respect to the reference diagnosis. McNemar’s analysis was performed to compare the sensitivity and specificity of the proprietary PPG and single-lead ECG AF detection algorithms. Results The median age in the study population (n = 300) was 70 years (interquartile range: 51-78), 56.3% were men, and the median CHA2DS2-VASc was 3 (interquartile range: 1-4) with an AF-prevalence of 32.3%. PPG signal and single‑lead ECG quality was sufficient in 272/300 (91.0%) and 278/298 (93.3%) participants, respectively. After excluding atrial flutter patients (n = 25) and insufficient quality measurements, the sensitivity and specificity were 97.6% (95% CI 93.8 to 99.3) and 94.1% (95% CI 86.8 to 98.1) for the PPG signal versus 95.7% (95% CI 91.4 to 98.3) and 91.1% (95% CI 83.2 to 96.1) for the single‑lead ECG signal, respectively. Results demonstrated a 96.4% (95% CI 93.2 to 98.3) accuracy for PPG and 94.1% (95% CI 90.4 to 96.6) for single-lead ECG. No significant differences in sensitivity (P = 0.453) or specificity (P = 0.219) between the proprietary PPG and single-lead ECG algorithms were found. Conclusion This study demonstrated equivalent diagnostic performance of PPG and single-lead ECG proprietary AF detection algorithms in smartphone apps.

scholarly journals Epitweetr: Early warning of public health threats using Twitter data

2021 ◽  
Author(s):  
Laura Espinosa ◽  
Ariana Wijermans ◽  
Francisco Orchard ◽  
Michael Hoehle ◽  
Thomas Czernichow ◽  
...  
Keyword(s):  
Public Health ◽  
Signal Detection ◽  
Early Warning ◽  
Detection Algorithm ◽  
Detection Algorithms ◽  
Twitter Data ◽  
Epidemic Intelligence ◽  
Health Threats ◽  
Correct Location ◽  
Sensitivity Specificity

Background: ECDC performs epidemic intelligence activities to systematically collate information from a variety of sources, including Twitter, to rapidly detect public health events. The lack of a freely available, customisable and automated early warning tool using Twitter data, prompted ECDC to develop epitweetr. The specific objectives are to assess the performance of the geolocation and signal detection algorithms used by epitweetr and to assess the performance of epitweetr in comparison with the manual monitoring of Twitter for early detection of public health threats. Methods: Epitweetr collects, geolocates and aggregates tweets to generate signals and email alerts. Firstly, we evaluated manually the tweet geolocation characteristics of 1,200 tweets, and assessed its accuracy in extracting the correct location and its performance in detecting tweets with available information on the tweet geolocation. Secondly, we evaluated signals generated by epitweetr between 19 October and 30 November 2020 and we calculated the positive predictive value (PPV). Then, we evaluated the sensitivity, specificity and timeliness of epitweetr in comparison with Twitter manual monitoring. Findings: The epitweetr geolocation algorithm had an accuracy of 30.1% and 25.9% at national and subnational levels, respectively. General and specific PPV of the signal detection algorithm was 3.0% and 74.6%, respectively. Epitweetr and/or manual monitoring detected 570 signals and 454 events. Epitweetr had a sensitivity of 78.6% [75.2% - 82.0%] and PPV of 74.6% [70.5% - 78.6%]; and the manual monitoring had a sensitivity of 47.9% [43.8% - 52.0%] and PPV of 97.9% [95.8% - 99.9%]. The median validation time difference between sixteen common events detected by epitweetr and manual monitoring was -48.6 hours [(-102.8) - (-23.7) hours]. Interpretation: Epitweetr has shown to have sufficient performance as an early warning tool for public health threats using Twitter data. Having developed epitweetr as a free, open-source tool with several configurable settings and a strong automated component, it is expected to increase its usability and usefulness to public health experts.

A dual U-Net algorithm for automating feature extraction from satellite imagery

2021 ◽  
pp. 154851292098354
Author(s):  
Samuel Humphries ◽  
Trevor Parker ◽  
Bryan Jonas ◽  
Bryan Adams ◽  
Nicholas J Clark
Keyword(s):  
Neural Networks ◽  
Object Detection ◽  
Satellite Images ◽  
Detection Algorithm ◽  
Military Operations ◽  
Detection Algorithms ◽  
Us Military ◽  
Standard Product ◽  
Different Types ◽  
Road Intersections

Quick identification of building and roads is critical for execution of tactical US military operations in an urban environment. To this end, a gridded, referenced, satellite images of an objective, often referred to as a gridded reference graphic or GRG, has become a standard product developed during intelligence preparation of the environment. At present, operational units identify key infrastructure by hand through the work of individual intelligence officers. Recent advances in Convolutional Neural Networks, however, allows for this process to be streamlined through the use of object detection algorithms. In this paper, we describe an object detection algorithm designed to quickly identify and label both buildings and road intersections present in an image. Our work leverages both the U-Net architecture as well the SpaceNet data corpus to produce an algorithm that accurately identifies a large breadth of buildings and different types of roads. In addition to predicting buildings and roads, our model numerically labels each building by means of a contour finding algorithm. Most importantly, the dual U-Net model is capable of predicting buildings and roads on a diverse set of test images and using these predictions to produce clean GRGs.

scholarly journals A novel algorithm to detect non-wear time from raw accelerometer data using deep convolutional neural networks

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Shaheen Syed ◽  
Bente Morseth ◽  
Laila A. Hopstock ◽  
Alexander Horsch
Keyword(s):  
Threshold Value ◽  
Detection Algorithm ◽  
Classification Model ◽  
Type I ◽  
Accelerometer Data ◽  
Deep Convolutional Neural Networks ◽  
Major Drawback ◽  
Wear Time ◽  
Detection Algorithms ◽  
Type Ii Errors

AbstractTo date, non-wear detection algorithms commonly employ a 30, 60, or even 90 mins interval or window in which acceleration values need to be below a threshold value. A major drawback of such intervals is that they need to be long enough to prevent false positives (type I errors), while short enough to prevent false negatives (type II errors), which limits detecting both short and longer episodes of non-wear time. In this paper, we propose a novel non-wear detection algorithm that eliminates the need for an interval. Rather than inspecting acceleration within intervals, we explore acceleration right before and right after an episode of non-wear time. We trained a deep convolutional neural network that was able to infer non-wear time by detecting when the accelerometer was removed and when it was placed back on again. We evaluate our algorithm against several baseline and existing non-wear algorithms, and our algorithm achieves a perfect precision, a recall of 0.9962, and an F1 score of 0.9981, outperforming all evaluated algorithms. Although our algorithm was developed using patterns learned from a hip-worn accelerometer, we propose algorithmic steps that can easily be applied to a wrist-worn accelerometer and a retrained classification model.

scholarly journals Kardia Mobile and ISTEL HR applicability in clinical practice: A comparison of Kardia Mobile, ISTEL HR and standard 12-lead electrocardiogram records in 98 consecutive patients of a tertiary cardiovascular care centre

2021 ◽  
Author(s):  
Bartosz Krzowski ◽  
Kamila Skoczylas ◽  
Gabriela Osak ◽  
Natalia Żurawska ◽  
Michał Peller ◽  
...  
Keyword(s):  
Atrial Fibrillation ◽  
Clinical Practice ◽  
Cardiovascular Diseases ◽  
Sinus Rhythm ◽  
Sensitivity And Specificity ◽  
Heart Rhythm ◽  
Everyday Practice ◽  
Portable Devices ◽  
Care Centre ◽  
Cardiovascular Care

Abstract Aims Mobile, portable ECG-recorders allow the assessment of heart rhythm in out-of-hospital conditions and may prove useful for monitoring patients with cardiovascular diseases. However, the effectiveness of these portable devices has not been tested in everyday practice. Methods and results A group of 98 consecutive cardiology patients (62 males [63%], mean age 69 ± 12.9 years) were included in an academic care centre. For each patient, a standard 12-lead electrocardiogram (SE), as well as a Kardia Mobile 6L (KM) and Istel (IS) HR-2000 ECG were performed. Two groups of experienced physycians analyzed obtained recordings. After analyzing ECG tracings from SE, KM, and IS, quality was marked as good in 82%, 80%, and 72% of patients, respectively (p < 0.001). There were no significant differences between devices in terms of detecting sinus rhythm (SE [60%, n = 59], KM [58%, n = 56], and IS [61%, n = 60]; SE vs KM p = 0.53; SE vs IS p = 0.76) and atrial fibrillation (SE [22%, n = 22], KM [22%, n = 21], and IS [18%, n = 18]; (SE vs KM p = 0.65; SE vs IS = 0.1). KM had a sensitivity of 88.1% and a specificity of 89.7% for diagnosing sinus rhythm. IS showed 91.5% and 84.6% sensitivity and specificity, respectively. The sensitivity of KM in detecting atrial fibrillation was higher than IS (86.4% vs. 77.3%), but their specificity was comparable (97.4% vs. 98.7%). Conclusion Novel, portable devices are useful in showing sinus rhythm and detecting atrial fibrillation in clinical practice. However, ECG measurements concerning conduction and repolarisation should be clarified with a standard 12-lead electrocardiogram.

Prognostic value of diastolic function assessment in patients undergoing cardiac resynchronization therapy

2021 ◽  
Vol 22 (Supplement_1) ◽  
Author(s):  
E Galli ◽  
OA Smiseth ◽  
JM Aalen ◽  
CK Larsen ◽  
E Sade ◽  
...  
Keyword(s):  
Diastolic Function ◽  
Primary Endpoint ◽  
Prognostic Value ◽  
Left Ventricular ◽  
Cardiac Resynchronization ◽  
Systolic Volume ◽  
Funding Sources ◽  
American Society ◽  
Sensitivity Specificity ◽  
Crt Response

Abstract Funding Acknowledgements Type of funding sources: None. Objective The best modality to assess diastolic function in CRT-candidates is an object of debate and the relationship between diastolic function, CRT-response and survival are not clearly understood. Purpose of the study: to assess diastolic patterns in patients undergoing CRT according to the 2016 recommendations of the American Society of Echocardiography/European Association of Cardiovascular Imaging and to evaluate the prognostic value of diastolic dysfunction (DD) in CRT candidates. Methods 193 patients (age: 67 ± 11 years, QRS width: 167 ± 21 ms) were included in this multicentre prospective study. Patients were stratified according to DD grades (grade I to III). CRT-response was defined as a reduction of left ventricular (LV) end-systolic volume >15% at 6-month follow-up (FU). The primary endpoint was defined as a composite of heart transplantation, LV assisted device implantation or all-cause death during FU. Results During FU, 132 (68%) patients were CRT-responders. CRT delivery was associated with diastolic function degradation in non-responders. Grade I DD was able to predict CRT-response with a sensitivity, specificity and accuracy of 70%, 65%, and 63%, respectively. After a median period of 35 months, the primary endpoint occurred in 29 (15%) patients. Grade I DD was associated with a better outcome [HR 0.26 95% CI: (0.10-0.66)], independently from ischemic cardiomyopathy, LV dyssynchrony and CRT-response (Table 1). Non-responders with grade II or grade III DD had the worse prognosis (HR 4.36, 95%CI: 2.10-9.06) Figure 1. Conclusions Grade I DD is associated with LV remodelling after CRT and is an independent predictor of prognosis in CRT candidates. Abstract Figure.

Validation of Michigan risk score and D-dimer to predict peripherally inserted central catheter-related thrombosis: A study of 206,132 catheter days

2021 ◽  
pp. 112972982110087
Author(s):  
Junren Kang ◽  
Wenyan Sun ◽  
Hailong Li ◽  
En ling Ma ◽  
Wei Chen
Keyword(s):  
Sensitivity And Specificity ◽  
Risk Score ◽  
Primary Objective ◽  
Class Iii ◽  
Operating Characteristics ◽  
Central Venous ◽  
Sensitivity Specificity ◽  
Catheter Related Thrombosis ◽  
D Dimer ◽  
Independent Cohort

Background: The Michigan Risk Score (MRS) was the only predicted score for peripherally inserted central venous catheters (PICC) associated upper extremity venous thrombosis (UEVT). Age-adjusted D-dimer increased the efficiency for UEVT. There were no external validations in an independent cohort. Method: A retrospective study of adult patients with PICC insertion was performed. The primary objective was to evaluate the performance of the MRS and age-adjusted D-dimer in estimating risk of PICC-related symptomatic UEVT. The sensitivity, specificity and areas under the receiver operating characteristics (ROC) of MRS and age-adjusted D-dimer were calculated. Results: Two thousand one hundred sixty-three patients were included for a total of 206,132 catheter days. Fifty-six (2.6%) developed PICC-UEVT. The incidences of PICC-UEVT were 4.9% for class I, 7.5% for class II, 2.2% for class III, 0% for class IV of MRS ( p = 0.011). The incidences of PICC-UEVT were 4.5% for D-dimer above the age-adjusted threshold and 1.5% for below the threshold ( p = 0.001). The areas under ROC of MRS and age-adjusted D-dimer were 0.405 (95% confidence interval (CI) 0.303–0.508) and 0.639 (95% CI 0.547–0.731). The sensitivity and specificity of MRS were 0.82 (95% CI, 0.69–0.91), 0.09 (95% CI, 0.08–0.11), respectively. The sensitivity and specificity of age-adjusted D-dimer were 0.64 (95% CI, 0.46–0.79) and 0.64 (95% CI, 0.61–0.66), respectively. Conclusions: MRS and age-adjusted D-dimer have low accuracy to predict PICC-UEVT. Further studies are needed.

scholarly journals Atrial Fibrillation Detection by the Combination of Recurrence Complex Network and Convolution Neural Network

2019 ◽  
Vol 2019 ◽  
pp. 1-9 ◽  
Author(s):  
Xiaoling Wei ◽  
Jimin Li ◽  
Chenghao Zhang ◽  
Ming Liu ◽  
Peng Xiong ◽  
...  
Keyword(s):  
Neural Network ◽  
Atrial Fibrillation ◽  
Complex Network ◽  
Individual Variation ◽  
Deep Neural Network ◽  
Detection Algorithm ◽  
Convolution Neural Network ◽  
Electrocardiogram Signal ◽  
Sensitivity Specificity ◽  
Atrial Fibrillation Detection

In this paper, R wave peak interval independent atrial fibrillation detection algorithm is proposed based on the analysis of the synchronization feature of the electrocardiogram signal by a deep neural network. Firstly, the synchronization feature of each heartbeat of the electrocardiogram signal is constructed by a Recurrence Complex Network. Then, a convolution neural network is used to detect atrial fibrillation by analyzing the eigenvalues of the Recurrence Complex Network. Finally, a voting algorithm is developed to improve the performance of the beat-wise atrial fibrillation detection. The MIT-BIH atrial fibrillation database is used to evaluate the performance of the proposed method. Experimental results show that the sensitivity, specificity, and accuracy of the algorithm can achieve 94.28%, 94.91%, and 94.59%, respectively. Remarkably, the proposed method was more effective than the traditional algorithms to the problem of individual variation in the atrial fibrillation detection.

Sign in / Sign up

Export Citation Format

Share Document