Datasets:
Helen Ryan
Helen-Ryan/
MIT-BIH-and-PTB-Image
0
Downloads: 3
Diff of /README.md [fb86f9] .. [f35e1b]

Switch to unified view

a/README.md b/README.md 
...
 
...
4 
<p><strong><em>Abstract</em></strong></p>
 4 
<p><strong><em>Abstract</em></strong></p>
5 
<p>This dataset is composed of two collections of heartbeat signals derived from two famous datasets in heartbeat classification, <a rel="noreferrer nofollow" aria-label="the MIT-BIH Arrhythmia Dataset  (opens in a new tab)" target="_blank" href="https://www.physionet.org/physiobank/database/mitdb/">the MIT-BIH Arrhythmia Dataset </a>and The <a rel="noreferrer nofollow" aria-label="PTB Diagnostic ECG Database (opens in a new tab)" target="_blank" href="https://www.physionet.org/physiobank/database/ptbdb/">PTB Diagnostic ECG Database</a>. The number of samples in both collections is large enough for training a deep neural network.</p>
 5 
<p>This dataset is composed of two collections of heartbeat signals derived from two famous datasets in heartbeat classification, <a rel="noreferrer nofollow" aria-label="the MIT-BIH Arrhythmia Dataset  (opens in a new tab)" target="_blank" href="https://www.physionet.org/physiobank/database/mitdb/">the MIT-BIH Arrhythmia Dataset </a>and The <a rel="noreferrer nofollow" aria-label="PTB Diagnostic ECG Database (opens in a new tab)" target="_blank" href="https://www.physionet.org/physiobank/database/ptbdb/">PTB Diagnostic ECG Database</a>. The number of samples in both collections is large enough for training a deep neural network.</p>
6 
<p>This dataset has been used in exploring heartbeat classification using deep neural network architectures, and observing some of the capabilities of transfer learning on it. The signals correspond to electrocardiogram (ECG) shapes of heartbeats for the normal case and the cases affected by different arrhythmias and myocardial infarction. These signals are preprocessed and segmented, with each segment corresponding to a heartbeat.<br>
 6 
<p>This dataset has been used in exploring heartbeat classification using deep neural network architectures, and observing some of the capabilities of transfer learning on it. The signals correspond to electrocardiogram (ECG) shapes of heartbeats for the normal case and the cases affected by different arrhythmias and myocardial infarction. These signals are preprocessed and segmented, with each segment corresponding to a heartbeat.<br>
7 
Content</p>
 7 
Content</p>
8 
<p><strong>Arrhythmia Dataset</strong></p>
 8 
<p><strong>Arrhythmia Dataset</strong></p>
9 
<pre class="uc-code-block"><code><span class="hljs-attr">Number of Samples:</span> <span class="hljs-number">109446</span>
 9 
<pre class="uc-code-block"><span class="hljs-attr">Number of Samples:</span> <span class="hljs-number">109446</span>
10 
<span class="hljs-attr">Number of Categories:</span> <span class="hljs-number">5</span>
 10 
<span class="hljs-attr">Number of Categories:</span> <span class="hljs-number">5</span>
11 
<span class="hljs-attr">Sampling Frequency:</span> <span class="hljs-string">125Hz</span>
 11 
<span class="hljs-attr">Sampling Frequency:</span> <span class="hljs-string">125Hz</span>
12 
<span class="hljs-attr">Data Source:</span> <span class="hljs-string">Physionet's</span> <span class="hljs-string">MIT-BIH</span> <span class="hljs-string">Arrhythmia</span> <span class="hljs-string">Dataset</span>
 12 
<span class="hljs-attr">Data Source:</span> <span class="hljs-string">Physionet's</span> <span class="hljs-string">MIT-BIH</span> <span class="hljs-string">Arrhythmia</span> <span class="hljs-string">Dataset</span>
13 
<span class="hljs-attr">Classes:</span> [<span class="hljs-attr">'N':</span> <span class="hljs-number">0</span>, <span class="hljs-attr">'S':</span> <span class="hljs-number">1</span>, <span class="hljs-attr">'V':</span> <span class="hljs-number">2</span>, <span class="hljs-attr">'F':</span> <span class="hljs-number">3</span>, <span class="hljs-attr">'Q':</span> <span class="hljs-number">4</span>]
 13 
<span class="hljs-attr">Classes:</span> [<span class="hljs-attr">'N':</span> <span class="hljs-number">0</span>, <span class="hljs-attr">'S':</span> <span class="hljs-number">1</span>, <span class="hljs-attr">'V':</span> <span class="hljs-number">2</span>, <span class="hljs-attr">'F':</span> <span class="hljs-number">3</span>, <span class="hljs-attr">'Q':</span> <span class="hljs-number">4</span>]
14 
</code>
 14 





  15
  
<p><strong>The PTB Diagnostic ECG Database</strong></p>

  15
  
<p><strong>The PTB Diagnostic ECG Database</strong></p>





  16
  
<pre class="uc-code-block"><code><span class="hljs-attribute">Number of Samples</span><span class="hljs-punctuation">:</span> <span class="hljs-string">14552</span>

  16
  
<pre class="uc-code-block"><span class="hljs-attribute">Number of Samples</span><span class="hljs-punctuation">:</span> <span class="hljs-string">14552</span>





  17
  
<span class="hljs-attribute">Number of Categories</span><span class="hljs-punctuation">:</span> <span class="hljs-string">2</span>

  17
  
<span class="hljs-attribute">Number of Categories</span><span class="hljs-punctuation">:</span> <span class="hljs-string">2</span>





  18
  
<span class="hljs-attribute">Sampling Frequency</span><span class="hljs-punctuation">:</span> <span class="hljs-string">125Hz</span>

  18
  
<span class="hljs-attribute">Sampling Frequency</span><span class="hljs-punctuation">:</span> <span class="hljs-string">125Hz</span>





  19
  
<span class="hljs-attribute">Data Source</span><span class="hljs-punctuation">:</span> <span class="hljs-string">Physionet's PTB Diagnostic Database</span>

  19
  
<span class="hljs-attribute">Data Source</span><span class="hljs-punctuation">:</span> <span class="hljs-string">Physionet's PTB Diagnostic Database</span>





  20
  
</code>

  20
  





  21
  
<p><strong>Data Files</strong></p>

  21
  
<p><strong>Data Files</strong></p>





  22
  
<p>This dataset is created by saving the each ECG arrhythmia into the image form. Then total images from each classes are divided into train and test data where training samples are a80% of the total data and test samples are 20%. </p>

  22
  
<p>This dataset is created by saving the each ECG arrhythmia into the image form. Then total images from each classes are divided into train and test data where training samples are a80% of the total data and test samples are 20%. </p>





  23
  
<p><strong>Acknowledgements</strong></p>

  23
  
<p><strong>Acknowledgements</strong></p>





  24
  
<p>Mohammad Kachuee, Shayan Fazeli, and Majid Sarrafzadeh. "ECG Heartbeat Classification: A Deep Transferable Representation." <a rel="noreferrer nofollow" aria-label="arXiv preprint arXiv:1805.00794 (2018) (opens in a new tab)" target="_blank" href="https://arxiv.org/abs/1805.00794">arXiv preprint arXiv:1805.00794 (2018)</a>.</p>

  24
  
<p>Mohammad Kachuee, Shayan Fazeli, and Majid Sarrafzadeh. "ECG Heartbeat Classification: A Deep Transferable Representation." <a rel="noreferrer nofollow" aria-label="arXiv preprint arXiv:1805.00794 (2018) (opens in a new tab)" target="_blank" href="https://arxiv.org/abs/1805.00794">arXiv preprint arXiv:1805.00794 (2018)</a>.</p>





  25
  
<p><strong>Inspiration</strong></p>

  25
  
<p><strong>Inspiration</strong></p>