<h1>About Dataset</h1>

<h2>Vertebral Column Data Set</h2>

<p>

<strong>Download:</strong> 

<a href="http://archive.ics.uci.edu/ml/machine-learning-databases/00212/" target="_blank">Data Folder</a> |

<a href="http://archive.ics.uci.edu/ml/machine-learning-databases/00212/" target="_blank">Data Set Description</a>

</p>

<h3>Abstract</h3>

<p>

This dataset contains values for six biomechanical features used to classify orthopaedic patients into either:

<ul>

  <li>Three classes: Normal, Disk Hernia, or Spondylolisthesis</li>

  <li>Or two classes: Normal or Abnormal</li>

</ul>

</p>

<h3>Dataset Characteristics</h3>

<ul>

  <li><strong>Type:</strong> Multivariate</li>

  <li><strong>Attributes:</strong> Real-valued</li>

  <li><strong>Associated Tasks:</strong> Classification</li>

  <li><strong>Number of Instances:</strong> 310</li>

  <li><strong>Number of Attributes:</strong> 6</li>

  <li><strong>Missing Values:</strong> None</li>

  <li><strong>Date Donated:</strong> 2011-08-09</li>

</ul>

<h3>Source</h3>

<p>

Guilherme de Alencar Barreto (guilherme '@' deti.ufc.br)<br>

Ajalmar Rêgo da Rocha Neto (ajalmar '@' ifce.edu.br)<br>

Department of Teleinformatics Engineering, Federal University of Ceará, Fortaleza, Brazil<br><br>

Henrique Antonio Fonseca da Mota Filho (hdamota '@' gmail.com)<br>

Hospital Monte Klinikum, Fortaleza, Brazil

</p>

<h3>Data Set Information</h3>

<p>

The biomedical data was collected during a medical residency in the Group of Applied Research in Orthopaedics (GARO) at the Centre Médico-Chirurgical de Réadaptation des Massues in Lyon, France. The dataset supports two classification tasks:

</p>

<ol>

  <li><strong>Three-class classification:</strong>

    <ul>

      <li>Normal – 100 patients</li>

      <li>Disk Hernia – 60 patients</li>

      <li>Spondylolisthesis – 150 patients</li>

    </ul>

  </li>

  <li><strong>Binary classification:</strong>

    <ul>

      <li>Normal – 100 patients</li>

      <li>Abnormal (Disk Hernia + Spondylolisthesis) – 210 patients</li>

    </ul>

  </li>

</ol>

<p>

Files formatted for use in the WEKA machine learning environment are also provided.

</p>

<h3>Attribute Information</h3>

<p>

Each patient is described by six biomechanical attributes, derived from the shape and orientation of the pelvis and lumbar spine:

</p>

<ul>

  <li>Pelvic incidence</li>

  <li>Pelvic tilt</li>

  <li>Lumbar lordosis angle</li>

  <li>Sacral slope</li>

  <li>Pelvic radius</li>

  <li>Grade of spondylolisthesis</li>

</ul>

<p>

Class labels include:

<ul>

  <li><strong>DH</strong> – Disk Hernia</li>

  <li><strong>SL</strong> – Spondylolisthesis</li>

  <li><strong>NO</strong> – Normal</li>

  <li><strong>AB</strong> – Abnormal (binary classification)</li>

</ul>

</p>

<h3>Relevant Papers</h3>

<ol>

  <li>Berthonnaud, E., Dimnet, J., Roussouly, P. & Labelle, H. (2005). “Analysis of the sagittal balance of the spine and pelvis using shape and orientation parameters.” <em>Journal of Spinal Disorders & Techniques</em>, 18(1):40–47.</li>

  <li>Rocha Neto, A. R. & Barreto, G. A. (2009). “On the Application of Ensembles of Classifiers to the Diagnosis of Pathologies of the Vertebral Column: A Comparative Analysis.” <em>IEEE Latin America Transactions</em>, 7(4):487–496.</li>

  <li>Rocha Neto, A. R., Sousa, R., Barreto, G. A. & Cardoso, J. S. (2011). “Diagnostic of Pathology on the Vertebral Column with Embedded Reject Option.” In <em>Proceedings of the 5th Iberian Conference on Pattern Recognition and Image Analysis (IbPRIA'2011)</em>, Lecture Notes on Computer Science, Vol. 6669, pp. 588–595.</li>

</ol>