Computer analysis of ventilation-perfusion scans for detection and assessment of lung disease

A previously reported computer analysis has been used to provide numerical ventilation-perfusion lung scan data, for comparison with tests of airways function and results of arterial blood gas analysis in 11 patients with pulmonary embolism, 18 with asthma, and 37 with chronic obstructive lung disease. In pulmonary embolism an index of underperfusion showed high sensitivity, and an index of ventilation-perfusion mismatching correlated well with severity (hypoxaemia). In asthma an index of underventilation was sensitive and correlated well with severity of airways obstruction. In chronic obstructive lung disease the same index was sensitive but correlated poorly with severity. This was attributed to heterogeneity of the lung disease (airways obstruction plus emphysema) in chronic obstructive lung disease. Ventilation-perfusion mismatching was frequently present in airways disease, and was often of great severity in chronic obstructive lung disease. Discrimination between pulmonary embolism and either type of airways disease was possible by using a combination of underfusion and underventilation indices. The technique offers the prospect of increasing the information derived from lung scans and of automating the reporting of scans.