Can computed tomography classifications of chronic obstructive pulmonary disease be identified using Bayesian networks and clinical data?

  • Authors:

  • Lars P. Thomsen;Ulla M. Weinreich;Dan S. Karbing;Vanja G. Helbo Jensen;Morten Vuust;Jens B. FrøKjæR;Stephen E. Rees

  • Affiliations:

  • Respiratory and Critical Care Group (RCARE), Center for Model-based Medical Decision Support, Department of Health Science and Technology, Aalborg University, Fredrik Bajers Vej 7E, DK-9220 Aalbor ...;Department of Pulmonary Medicine, Aalborg University Hospital, Mølleparkvej 4, DK-9000 Aalborg, Denmark;Respiratory and Critical Care Group (RCARE), Center for Model-based Medical Decision Support, Department of Health Science and Technology, Aalborg University, Fredrik Bajers Vej 7E, DK-9220 Aalbor ...;Department of Radiology, Aalborg University Hospital, Hobrovej 18-22, DK-9000 Aalborg, Denmark;Department of Radiology, Sygehus Vendsyssel, Barfredsvej 83, DK-9000 Fredrikshavn, Denmark;Department of Radiology, Aalborg University Hospital, Hobrovej 18-22, DK-9000 Aalborg, Denmark;Respiratory and Critical Care Group (RCARE), Center for Model-based Medical Decision Support, Department of Health Science and Technology, Aalborg University, Fredrik Bajers Vej 7E, DK-9220 Aalbor ...

  • Venue:
  • Computer Methods and Programs in Biomedicine
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Diagnosis and classification of chronic obstructive pulmonary disease (COPD) may be seen as difficult. Causal reasoning can be used to relate clinical measurements with radiological representation of COPD phenotypes airways disease and emphysema. In this paper a causal probabilistic network was constructed that uses clinically available measurements to classify patients suffering from COPD into the main phenotypes airways disease and emphysema. The network grades the severity of disease and for emphysematous COPD, the type of bullae and its location central or peripheral. In four patient cases the network was shown to reach the same conclusion as was gained from the patients' High Resolution Computed Tomography (HRCT) scans. These were: airways disease, emphysema with central small bullae, emphysema with central large bullae, and emphysema with peripheral bullae. The approach may be promising in targeting HRCT in COPD patients, assessing phenotypes of the disease and monitoring its progression using clinical data.