Summary
Background: Lung microangiopathy is a little known negative influence of diabetes mellitus on
the functioning of the lungs. In current medical practice lung microangiopathy is
diagnosed by comparing two measurements of lung diffusing capacity – once with the
subject standing and once with the subject lying down. The necessity to take two measurements
is inconvenient.
Objectives: The aim of this study is to design a supportive method for diagnosing lung microangiopathy.
This will be based on routinely performed pulmonary measurements as well as on investigation
of process modelling and data processing.
Methods: A model of the diffusion of oxygen from the alveoli to the blood has been described
with a set of differential equations. The idea of the proposed model is based on the
physiological analysis of the oxygen flow (caused by a concentration gradient) and
on general knowledge regarding the kinetics of associating oxygen with haemoglobin.
The model parameters are estimated using diffusing capacity and alveolar volume measurements
– routinely performed in pulmonary tests.
Results: The model parameter estimates proved good candidates for the binary classification
of the presence or absence of micro-angiopathy. The proposed classification procedure,
based on parameter values and established diagnostic thresholds, gives sensitivity
Sens = 79.34% and specificity Spec = 87.08%. The results of classification with the use of diffusing capacity measurement
are worse: Sens = 62.12% and Spec = 79.89%.
Conclusions: The proposed classification procedure is based on the model parameters. These have
proved to be sensitive indicators of lung microangiopathy. Close to 80% of micro-angiopathy
cases have been classified as such. Less than 20% were false alarms. The oxygen pathway
model allows for simulations. Blood saturation and oxygen partial pressure have been
simulated for the organism’s various needs for oxygen, both for the normal and the
impaired alveoli-capillary barrier.
Keywords
Models - decision support - lung - diffusion
