Data Quality Assessment

Uploading Data

Once the dashboard application has been launched in your web browser, select your data file. PhysioView will accept raw European Data Formatted (.edf) files outputted by the Actiwave Cardio, archive (.zip) files outputted by the Empatica E4 or created for a batch of files, and comma-separated values (.csv) files outputted by other devices.

Single-File Uploads

In this example, a single CSV file containing raw ECG and accelerometer data for one subject is uploaded to the dashboard.

Specify the data type and sampling rate, and map the CSV headers to the corresponding sample/timestamp and signal variables.

Batch Uploads

PhysioView can also process multiple CSV files packaged in a ZIP archive for visualization and quality assessment on the dashboard. All CSV files in the ZIP archive must have identical headers.

Setting Data Parameters

In both the cardiac and EDA processing pipelines, filtering can be enabled and the segment size defined for quality assessment.

ECG/PPG Data

Select whether to filter your cardiac data and choose an artifact identification method and tolerance value (any floating-point value between 0 and 2) and beat detection algorithm. By default, beats in ECG signals are detected with a Shannon energy envelope approach [1] and beats in PPG signals are detected with an adaptive thresholding algorithm [2]. Artifacts are identified using the criterion beat difference test [3].

EDA Data

Specify the sampling rate for the EDA data and map the CSV headers to the corresponding timestamp and signal variables. Optionally, either load a separate CSV file containing a single temperature column (with at most one header row) or map a temperature variable directly from the primary CSV file.

Select whether to filter the EDA data and detect skin conductance response (SCR) peaks [4], specifying the minimum peak amplitude and the valid EDA value range.

Viewing Signal Quality Metrics

PhysioView’s main dashboard shows three panels.

Data Summary displays information about the loaded data file and signal quality metrics.

Data Quality shows interactive bar charts of quality metrics, such as numbers of missing and artifactual beats per segment in cardiac signals and proportions of invalid data points or values that are out of range or contain excessive slopes in EDA signals.

Signal View shows the primary signal (ECG/PPG/EDA), its associated signals (interbeat intervals or temperature), and, if given, the acceleration signal preprocessed from the uploaded data.

Exporting the Data Summary

The Export Summary button in the Data Summary panel allows you to download the data quality assessment summary for either a single file or a batch of files as a ZIP archive file or an Excel workbook. For batch exports in Excel mode, a separate workbook is created for each subject and all workbooks are packaged together in a ZIP archive.

Download an example Excel workbook with the SQA summary here.