<div class="sc-kdrUpr eZtUed"><div class="sc-UEtKG dGqiYy sc-hDzlxo bEIZRR"><div class="sc-fqwslf gsqkEc"><div class="sc-cBQMlg kAHhUk"><h2 class="sc-dcKlJK sc-cVttbi gqEuPW ksnHgj">About Dataset</h2></div></div></div><div class="sc-fHzVOS cUYeeo"><div class="sc-davvxH nUNNB"><div style="min-height: 80px;"><div class="sc-etVRix jqYJaa sc-jCNfQM igJSrG"><h3>Context</h3>

<p>Respiratory sounds are important indicators of respiratory health and respiratory disorders. The sound emitted when a person breathes is directly related to air movement, changes within lung tissue and the position of secretions within the lung. A wheezing sound, for example, is a common sign that a patient has an obstructive airway disease like asthma or chronic obstructive pulmonary disease (COPD).</p>

<p>These sounds can be recorded using digital stethoscopes and other recording techniques. This digital data opens up the possibility of using machine learning to automatically diagnose respiratory disorders like asthma, pneumonia and bronchiolitis, to name a few. </p>

<h3>Content</h3>

<p>The Respiratory Sound Database was created by two research teams in Portugal and Greece. It includes 920 annotated recordings of varying length - 10s to 90s. These recordings were taken from 126 patients. There are a total of 5.5 hours of recordings containing 6898 respiratory cycles - 1864 contain crackles, 886 contain wheezes and 506 contain both crackles and wheezes. The data includes both clean respiratory sounds as well as noisy recordings that simulate real life conditions. The patients span all age groups - children, adults and the elderly.</p>

<p>This Kaggle dataset includes:</p>

<ul>

<li>920 .wav sound files</li>

<li>920 annotation .txt files</li>

<li>A text file listing the diagnosis for each patient</li>

<li>A text file explaining the file naming format</li>

<li>A text file listing 91 names (filename_differences.txt )</li>

<li>A text file containing demographic information for each patient</li>

</ul>

<p>Note:<br><br>

 filename_differences.txt is a list of files whose names were corrected after this dataset's creators found a bug in the original file naming script. It can now be ignored.</p>

<h3>General</h3>

<pre class="uc-code-block"><code> The demographic info file has <span class="hljs-number">6</span> columns:

  - Patient number

  - Age

  - Sex

  - Adult <span class="hljs-title function_">BMI</span> <span class="hljs-params">(kg/m2)</span>

  - Child <span class="hljs-title function_">Weight</span> <span class="hljs-params">(kg)</span>

  - Child <span class="hljs-title function_">Height</span> <span class="hljs-params">(cm)</span>

Each audio file name is divided into <span class="hljs-number">5</span> elements, separated with <span class="hljs-title function_">underscores</span> <span class="hljs-params">(_)</span>.

<span class="hljs-number">1.</span> Patient <span class="hljs-title function_">number</span> <span class="hljs-params">(<span class="hljs-number">101</span>,<span class="hljs-number">102</span>,...,<span class="hljs-number">226</span>)</span>

<span class="hljs-number">2.</span> Recording index

<span class="hljs-number">3.</span> Chest location 

      a. Trachea (Tc)

      b. Anterior <span class="hljs-title function_">left</span> <span class="hljs-params">(Al)</span>

      c. Anterior <span class="hljs-title function_">right</span> <span class="hljs-params">(Ar)</span>

      d. Posterior <span class="hljs-title function_">left</span> <span class="hljs-params">(Pl)</span>

      e. Posterior <span class="hljs-title function_">right</span> <span class="hljs-params">(Pr)</span>

      f. Lateral <span class="hljs-title function_">left</span> <span class="hljs-params">(Ll)</span>

      g. Lateral <span class="hljs-title function_">right</span> <span class="hljs-params">(Lr)</span>

<span class="hljs-number">4.</span> Acquisition mode 

     a. sequential/single <span class="hljs-title function_">channel</span> <span class="hljs-params">(sc)</span>, 

     b. simultaneous/multichannel (mc)

<span class="hljs-number">5.</span> Recording equipment 

     a. AKG C417L <span class="hljs-title function_">Microphone</span> <span class="hljs-params">(AKGC417L)</span>, 

     b. 3M Littmann Classic II SE <span class="hljs-title function_">Stethoscope</span> <span class="hljs-params">(LittC2SE)</span>, 

     c. 3M Litmmann <span class="hljs-number">3200</span> Electronic <span class="hljs-title function_">Stethoscope</span> <span class="hljs-params">(Litt3200)</span>, 

     d.  WelchAllyn Meditron Master Elite Electronic <span class="hljs-title function_">Stethoscope</span> <span class="hljs-params">(Meditron)</span>

The annotation text files have four columns:

- Beginning of respiratory <span class="hljs-title function_">cycle</span><span class="hljs-params">(s)</span>

- End of respiratory <span class="hljs-title function_">cycle</span><span class="hljs-params">(s)</span>

- Presence/absence of <span class="hljs-title function_">crackles</span> <span class="hljs-params">(presence=<span class="hljs-number">1</span>, absence=<span class="hljs-number">0</span>)</span>

- Presence/absence of <span class="hljs-title function_">wheezes</span> <span class="hljs-params">(presence=<span class="hljs-number">1</span>, absence=<span class="hljs-number">0</span>)</span>

The abbreviations used in the diagnosis file are:

- COPD: Chronic Obstructive Pulmonary Disease

- LRTI: Lower Respiratory Tract Infection

- URTI: Upper Respiratory Tract Infection

</code><div class="uc-code-block-copy-button-wrapper"><button class="uc-code-block-copy-button google-symbols" aria-label="Copy code">content_copy</button></div></pre>

<h3>Citation</h3>

<p>Paper: Α Respiratory Sound Database for the Development of Automated Classification<br><br>

Rocha BM, Filos D, Mendes L, Vogiatzis I, Perantoni E, Kaimakamis E, Natsiavas P, Oliveira A, Jácome C, Marques A, Paiva RP (2018) In Precision Medicine Powered by pHealth and Connected Health (pp. 51-55). Springer, Singapore.<br><br>

<a rel="noreferrer nofollow" aria-label="https://eden.dei.uc.pt/~ruipedro/publications/Conferences/ICBHI2017a.pdf (opens in a new tab)" target="_blank" href="https://eden.dei.uc.pt/~ruipedro/publications/Conferences/ICBHI2017a.pdf">https://eden.dei.uc.pt/~ruipedro/publications/Conferences/ICBHI2017a.pdf</a></p>

<h4>Ref Websites</h4>

<ul>

<li><a rel="noreferrer nofollow" aria-label="http://www.auditory.org/mhonarc/2018/msg00007.html (opens in a new tab)" target="_blank" href="http://www.auditory.org/mhonarc/2018/msg00007.html">http://www.auditory.org/mhonarc/2018/msg00007.html</a></li>

<li><a rel="noreferrer nofollow" aria-label="http://bhichallenge.med.auth.gr/ (opens in a new tab)" target="_blank" href="http://bhichallenge.med.auth.gr/">http://bhichallenge.med.auth.gr/</a> </li>

</ul>

<h3>Acknowledgements</h3>

<p>Many thanks to the research teams at the University of Coimbra, Portugal; the University de Aveiro, Portugal and the Aristotle University of Thessaloniki, Greece for making this dataset publicly available.</p>

<h3>Inspiration</h3>

<ul>

<li>Build a model to classify respiratory diseases.</li>

<li>Build a model to detect if a recording contains crackles, wheezes or both.</li>

<li>Annotation is a time consuming process. Create a model to automatically annotate respiratory sound recordings.</li>

<li>Deploy your model as a Tensorflow.js web app so it can be accessed from anywhere in the world.</li>

<li>Bioelectronics - Can you build your own digital stethoscope using an Arduino? If you are an aspiring inventor, this video will give you some valuable practical advice: <a rel="noreferrer nofollow" aria-label="https://www.youtube.com/watch?v=jo1cQ-ga2MI (opens in a new tab)" target="_blank" href="https://www.youtube.com/watch?v=jo1cQ-ga2MI">https://www.youtube.com/watch?v=jo1cQ-ga2MI</a></li>

</ul>

<p>Photo by voltamax on Pixabay.</p></div></div></div>