PTB-XL (v1.0.1) Soft Segmentations (Delineation)

<p><strong>Overview</strong></p> <p>The dataset provides (soft) segmentation/delineation masks for the PTB-XL ECG dataset. The underlying original dataset (v1.0.1) is available from PhysioNet (<a href="https://doi.org/10.13026/kfzx-aw45">doi.org/10.13026/kfzx-aw45</a>) For further details on the orginal dataset, consult the dataset descriptor (Wagner, P., Strodthoff, N., Bousseljot, RD. et al. PTB-XL, a large publicly available electrocardiography dataset. Sci Data 7, 154 (2020). <a href="https://doi.org/10.1038/s41597-020-0495-6)">https://doi.org/10.1038/s41597-020-0495-6)</a>.</p> <p><strong>Dataset</strong></p> <p>The dataset was obtained by training a U-net segmentation model using</p> <p>ECGDeli (Pilia, N., Nagel, C., et al. ECGdeli&nbsp;- An open source ECG delineation toolbox for MATLAB. SoftwareX 13, 100639 (2021).<br> <a href="https://doi.org/10.1016/j.softx.2020.100639">https://doi.org/10.1016/j.softx.2020.100639</a>.).</p> <p>as initial labels increased by adding intermediate segments. Details on the model architecture will be described in a forthcoming article (Wagner, P., Mehari, T. et al.).</p> <p><strong>Dataset format</strong></p> <p>In the folder ptbxl_segmentations_8bit, we provide files in numpy format (8 bit unsigned integer)&nbsp;with filenames corresponding to the ECG ID of the corresponding sample in PTB-XL. Each file contains an array of shape 1000x12x24, where the first axis corresponds to time steps in the sequence (10s at 100 Hz), the second refers to the 12 ECG leads ([&#39;I&#39;, &#39;II&#39;, &#39;III&#39;, &#39;aVR&#39;, &#39;aVL&#39;, &#39;aVF&#39;, &#39;V1&#39;, &#39;V2&#39;, &#39;V3&#39;, &#39;V4&#39;, &#39;V5&#39;, &#39;V6&#39;]) and the third axis refers to [&#39;P-Onset&#39;, &#39;(p&#39;, &#39;P-Peak&#39;, &#39;p)&#39;, &#39;P-Offset&#39;, &#39;pq&#39;, &#39;QRS-Onset&#39;, &#39;(q&#39;, &#39;Q-Peak&#39;, &#39;qr&#39;, &#39;R-Peak&#39;, &#39;rs&#39;, &#39;S-Peak&#39;, &#39;q)&#39;, &#39;QRS-Offset&#39;, &#39;qj&#39;, &#39;J-point&#39;, &#39;jt&#39;, &#39;T-Onset&#39;, &#39;(t&#39;, &#39;T-Peak&#39;, &#39;t)&#39;, &#39;T-Offset&#39;, &#39;tp&#39;]. For every time step and channel it provides a soft output distribution over the 24 output classes of the segmentation model. &nbsp;&nbsp;</p>