Computing oxygen-enhanced ventilation maps using correlation analysis.

Correlation maps of oxygen-enhanced ventilation were obtained in nine healthy volunteers using complete and selected image series. The complete series included all images acquired with the subjects alternately inhaling room air and 100% oxygen. The selected series were the subsets of the complete series and included only co-registered images that showed matched diaphragmatic position at maximal expiration. Cross-correlation was computed between the time response function of each pixel and the input function representing the alternation between periods of room air and 100% oxygen inhalation. The confidence level for the correlation analysis was set to 0.01. Pulmonary parenchymal anatomy was consistently reproduced throughout the lung, even in anterior slices where published data have reported correlation problems. The overall average correlation coefficient was 0.66 +/- 0.07 for the complete series and 0.75 +/- 0.08 for the selected series. It was concluded that correlation analysis could be used to reconstruct qualitative oxygen-enhanced ventilation maps.

Ventilation-perfusion ratio of signal intensity in human lung using oxygen-enhanced and arterial spin labeling techniques.

This study investigates the distribution of ventilation-perfusion (V/Q) signal intensity (SI) ratios using oxygen-enhanced and arterial spin labeling (ASL) techniques in the lungs of 10 healthy volunteers. Ventilation and perfusion images were simultaneously acquired using the flow-sensitive alternating inversion recovery (FAIR) method as volunteers alternately inhaled room air and 100% oxygen. Images of the T(1) distribution were calculated for five volunteers for both selective (T(1f)) and nonselective (T(1)) inversion. The average T(1) was 1360 ms +/- 116 ms, and the average T(1f) was 1012 ms +/- 112 ms, yielding a difference that is statistically significant (P < 0.002). Excluding large pulmonary vessels, the average V/Q SI ratios were 0.355 +/- 0.073 for the left lung and 0.371 +/- 0.093 for the right lung, which are in agreement with the theoretical V/Q SI ratio. Plots of the V/Q SI ratio are similar to the logarithmic normal distribution obtained by multiple inert gas elimination techniques, with a range of ratios matching ventilation and perfusion. This MRI V/Q technique is completely noninvasive and does not involve ionized radiation. A limitation of this method is the nonsimultaneous acquisition of perfusion and ventilation data, with oxygen administered only for the ventilation data.

Influence of oxygen flow rate on signal and T(1) changes in oxygen-enhanced ventilation imaging.

PURPOSE: To investigate the optimal oxygen flow rate for oxygen-enhanced MR ventilation imaging. MATERIALS AND METHODS: Using a cardiac-triggered nonselective inversion recovery (IR) half Fourier single-shot fast spin echo sequence, series of images were acquired with the subject alternately inhaling room air and 100% oxygen. Oxygen flow rates of 5 L/min, 10 L/min, 15 L/min, 20 L/min, and 25 L/min were studied, and signal intensity from the oxygen-enhanced ventilation images and T(1) of the lung were measured. RESULTS: The average signal intensity was 63.0 +/- 21.0 for 5 L/min, 98.7 +/- 26.8 for 10 L/min, 133.8 +/- 20.0 for 15 L/min, 138.7 +/- 19.7 for 20 L/min, and 139.2 +/- 37.9 for 25 L/min. The average T(1)'s of the lung were 1399 msec +/- 130 msec for room air, 1314 msec +/- 101 msec for 5 L/min, 1276 msec +/- 105 msec for 10 L/min, 1207 msec +/- 71 msec for 15 L/min, 1206 msec +/- 90 msec for 20 L/min, and 1207 msec +/- 42 msec for 25 L/min. CONCLUSION: The optimal flow rate is 15 L/min for oxygen-enhanced ventilation imaging.