1 From the Departments of Clinical Radiology (C.J.M., J.S., J.W., M.F.R.) and Internal Medicine I (M.S.), Klinikum Grosshadern, University of Munich, Marchioninistrasse 15, D-81377 Munich, Germany; and Siemens Medical Systems, Erlangen, Germany (M.D., R.B.L.). Received May 2, 2000; revision requested June 29; revision received April 5, 2001; accepted August 2, 2001.
PURPOSE: To determine differences in the signal intensity (SI)
time courses at oxygen-enhanced magnetic resonance (MR) lung
imaging in healthy volunteers and patients with pulmonary diseases
and to correlate these differences with pulmonary diffusing
capacity.
MATERIALS AND METHODS: Seventeen patients with pulmonary diseases and 11 healthy volunteers underwent oxygen-enhanced MR imaging while they breathed room air and 100% oxygen. A turbo spin-echo sequence with global or section-selective inversion pulses was used. For postprocessing, SI slope maps during the breathing of 100% oxygen were calculated. Mean SI slope and SI change values were compared with the diffusing capacity of the lung for carbon monoxide (DLCO).
RESULTS: The SI slopes were significantly different for patients and volunteers (P .05, Mann-Whitney U test). Linear correlations were detected between the DLCO and SI slopes for the section-selective inversion pulse (r2 = 0.81) and the global inversion pulse (r2 = 0.74). A lower correlation was associated with the SI change for the section-selective pulse (r2 = 0.04; global pulse, r2 = 0.81). Regional differences were seen in the SI slope and SI change maps. These differences correlated with findings on radiographs and computed tomographic scans.
CONCLUSION: The SI slope during the breathing of 100% oxygen allows spatially resolved assessment of the pulmonary diffusion capacity.
Index terms: Lung, function, 60.919 • Lung, MR, 60.12143 • Magnetic resonance (MR), contrast enhancement, 60.12143 • Magnetic resonance (MR), functional imaging, 60.12143, 60.12144
A variety of pulmonary disorders are associated with impaired
oxygen diffusion from the alveoli into the capillaries. Currently,
calculation of the diffusing capacity of the lung for carbon
monoxide (D
LCO) is used to represent oxygen diffusion from the
alveoli into the capillaries (
1,
2). However, this procedure
yields only global values, and local differences in the diffusing
capacity cannot be detected. Thus the ability to perform spatially
resolved evaluations of oxygen transfer in different sections
of the lungs is highly desirable to detect areas in the lungs
with a limited lung function and to determine whether a diffuse
or localized pathologic condition exists. The information acquired
with this type of examination may be especially beneficial for
lung surgery, including lung transplantation, and for the diagnosis
of diseases with limited oxygen transfer from the alveoli to
the capillaries, such as pulmonary emphysema and fibrosis.
Oxygen-enhanced magnetic resonance (MR) lung imaging allows spatially resolved visualization of oxygen diffusion from the alveoli into the capillaries of the lungs. Edelman et al (3), Stock et al (4), and Chen et al (5) proposed the use of 100% oxygen as a contrast agent in MR imaging of the lung. Use of molecular oxygen, a weak paramagnetic agent with a magnetic moment of 2.8 Bohr magnetons (6–8), leads to a reduction in the T1 of blood (9,10). Therefore, the breathing of 100% oxygen results in signal intensity (SI) changes in those areas of the lungs in which oxygen diffuses from the alveoli into the capillaries with use of a T1-weighted pulse sequence. Although the MR signal is influenced by the paramagnetic property of deoxygenated hemoglobin, it has been shown that there is little effect on T1 (11,12).
The purpose of this study was to compare the SI time courses of healthy volunteers and patients with a pathologic diffusing capacity during the breathing of 100% oxygen and to correlate the differences with the pulmonary diffusing capacity.
The study population comprised 11 volunteers (one woman, 10
men; age range, 26–31 years; mean age, 28 years ±
1 [SD]) and 17 patients (11 women, six men; age range, 29–56
years; mean age, 43 years ± 12) with various pulmonary
diseases. The patients were recruited consecutively into the
study. Patient inclusion criteria were patient age between 18
and 70 years, histologic diagnosis of interstitial lung disease,
and stable phase of the disease (no acute bronchopulmonary infection
within the 2 weeks before the examination). Exclusion criteria
included ischemic heart disease, skeletal muscle disease, malignant
disease, alcohol and/or drug addiction, pregnancy, and a contraindication
to MR imaging. The study protocol was approved by the university
review board, and all subjects gave informed consent before
participating in the study.
Twelve of the 17 patients had idiopathic lung fibrosis, one had extrinsic allergic alveolitis, two had pulmonary emphysema, one had adult respiratory distress syndrome, and one had pulmonary hypertension. Three of the 17 patients were examined after unilateral lung transplantation to assess diffusing capacity related to emphysema (n = 1) or fibrosis (n = 2) in the native lung. Findings in the transplanted lung lobes were excluded from statistical analysis. The volunteers did not have any known lung or heart disease.
MR imaging examinations were performed with a 1.5-T whole-body imager (Magnetom Vision; Siemens Medical Systems, Erlangen, Germany). Measurements were performed with a single-shot inversion-recovery turbo spin-echo sequence with repetition time of 7,000 or 2,000 msec, echo time of 4.2 msec, section thickness of 8 mm, and echo train length of 537.6 msec. The sequence was centrically reordered so the k-space lines critical for contrast were measured first, since T2 in the lungs is only 20–40 msec at 1.5 T. An optimized inversion time of 1.3 seconds was used (13). The field of view was 400 mm, and the matrix was 128 x 256 (phase-encoding x readout directions). The standard sequence protocol included a global inversion pulse. For patients with severely reduced lung function, a section-selective inversion pulse was used to reduce the examination time. Three volunteers were also examined with this section-selective pulse scheme to compare their SI time courses with those of the patients.
To allow recovery of the magnetization, the delay between two images was 7 seconds when a global inversion pulse was used. In contrast, the delay was only 2 seconds for the section-selective inversion pulse in the interleaved acquisition mode, which resulted in an acquisition time of 8 seconds for four images. End-diastolic electrocardiographic triggering was used, and four coronal images were acquired with both versions of the sequence. The images were obtained within the R-R interval. The start of the inversion pulse was optimized to perform data acquisition at the end of the R-R interval. This was done individually for each patient and volunteer depending on the length of the R-R interval. The signal-to-noise ratio for the acquired base images was 1.0.
Eight volunteers and 12 patients were examined with the sequence with the global inversion pulse. Examinations performed with the sequence with the section-selective inversion pulse were performed in five patients and three volunteers.
The patients and volunteers alternately breathed room air and 100% oxygen. For the volunteers, the acquisition paradigm was as follows: seven measurements during the breathing of room air (to check for baseline variations), then nine measurements during the breathing of 100% oxygen, and finally seven more measurements during breathing of room air. Measurement meant the acquisition of two images with the global inversion pulse and four images with the section-selective inversion pulse. For the patients, the corresponding numbers of measurements were four for baseline followed by four to eight during the breathing of 100% oxygen or room air.
When the sequence with the global inversion pulse was used, the paradigm of breathing room air and then 100% oxygen had to be applied twice to obtain four MR images, since only two MR images could be acquired in one breath hold. When the sequence with a section-selective inversion pulse was used, however, the paradigm had to be applied only once, since all four images could be acquired in one breath hold. This allowed use of a longer paradigm for the section-selective inversion pulse: Four baseline images were acquired in the first (room air) time slot, 12 images in the second (100% oxygen) time slot, and 12 images in the third (room air) time slot.
Whenever possible, all examinations were performed while the patients held their breath. Because this application is limited in patients with severe lung disease who are unable to hold their breath, the three patients with severe lung fibrosis underwent imaging without a breath hold. Since the section-selective inversion pulse may be associated with more artifacts, which result from respiratory motion, than the global inversion pulse in these patients, the global inversion pulse was used in their examinations. For comparison, four volunteers underwent imaging with this same protocol.
Each MR examination took approximately 30–45 minutes, including patient positioning, measurements, and data acquisition. No adverse reactions due to the breathing of 100% oxygen were noticed in the volunteers and patients. The examinations with the pulse sequences with the section-selective inversion pulse were well tolerated by the patients because of the shorter measurement time.
A special breathing system was used for the inhalation of room air and 100% oxygen (14,15). To avoid motion artifacts, the volunteers and patients were not allowed to take the mouthpiece out. A high flow of 20–25 L/min was used to administer the room air and 100% oxygen to achieve a sharp transition when the gases were switched (15).
For postprocessing, only images with a diaphragm position within a certain tolerance level were used. The tolerance level was defined as a diaphragm motion of plus or minus 10 pixels. These images were selected manually by one author (C.J.M.). This procedure was necessary even though we used a breath-hold technique because the diaphragm was not in the same position for all breath holds. Despite this fact, we continued to use the breath-hold technique because it reduced diaphragm motion more than free breathing did.
A user-developed software (Advanced Visual Systems, Waltham, Mass) was implemented to calculate maps of the slopes of the SI time course after the switch to 100% oxygen and the SI change between breathing room air and 100% oxygen. This procedure is similar to the approach presented for the assessment of myocardial perfusion (16). A 9 x 9-pixel kernel was used to spatially smooth the images. The measurements (performed by two of the authors [C.J.M., M.S.]) in the first segment of room air breathing were averaged and taken as baseline values. A low-pass Butterworth filter with a cutoff value of 10 and a power of 100 smoothed the SI time course over the entire experiment. The purpose of the filter was to reduce high-frequency noise in the SI time curves. In the next step, the software automatically detected the maximum SI during the breathing of 100% oxygen. The slope of the SI change (SI) was then calculated with the following formula: SI = (SIM - SIB)/(tM - tB), where SIM represents the maximum SI during the breathing of 100% oxygen, tM represents the corresponding time, SIB represents the value of the SI of the averaged baseline measurements, and tB represents the time when the breathing of 100% oxygen started.
The percentage SI change between the breathing of 100% oxygen and room air was calculated pixel by pixel. Mean values for the SI slopes and SI changes were calculated for six defined regions of interest (three for each lung) and for each image acquired. The regions of interest were placed by one author (C.J.M.). The mean size of the regions was 180 pixels ± 23, which corresponds to an area of 879 mm2 ± 112. The SI slopes and SI changes were translated into color-coded maps of the lungs.
All patients underwent routine lung function testing, which comprised several examinations. The lung function tests included analysis of dynamic and static lung volumes and arterialized blood gas. In the comparisons of patients and volunteers, only the DLCO was used. The time between the pulmonary function test and the MR imaging examination was 1 day to 1 week. The mean values for the SI slopes and SI changes averaged over all regions of interest and all images were compared with DLCO (2). In addition to the evaluation of the SI slopes, SI changes were calculated by averaging the images acquired during the breathing of 100% oxygen with the images acquired during the breathing of room air at the beginning of the experiment. SI changes were calculated for the same regions of interest used for the calculation of SI slopes. Percentage SI change maps were calculated as follows: (IO2 - IB)/IB, where IO2 is the averaged image for the images acquired during the breathing of 100% oxygen, and IB is the averaged images acquired during the breathing of room air before the inhalation of 100% oxygen. The first two images acquired during the breathing of 100% oxygen were eliminated in the calculation of IO2 to avoid the averaging of images acquired in the initial upslope of the SI.
SI slopes and SI changes were compared statistically between volunteers and patients (nonparametric Mann-Whitney U test). Differences with a P value of .05 or less were considered statistically significant. However, since no adjustment for multiple testing was made, the P values were used for descriptive purposes only.
The effect of breathing 100% oxygen was very weak on the base
images acquired during the breathing of room air and oxygen
(
Fig 1). SI changes averaged over all images for the patients
ranged from 1.1% ± 1.9 to 21.1% ± 6.5 (
Fig 2).
For the volunteers for the sequence with a global inversion
pulse, the corresponding mean values (
Table) were 16.7% ±
7.2 with a breath hold and 14.7% ± 5.8 without a breath
hold. For the volunteers examined with the sequence with the
section-selective pulse, the corresponding value for the SI
change averaged over all images was 14.0% ± 5.2. Comparison
of the mean, median, and range values for the sequences with
the section-selective inversion pulse and the global inversion
pulse (
Table) revealed similar results with both sequences.
Therefore, the two groups used in the statistical comparisons
consisted of patients and volunteers examined with the section-selective
or global sequence, respectively, with a breath hold. The data
for the non–breath-hold examinations were compared without
a statistical test because of the small number of volunteers
and patients. Differences in SI changes and SI slopes in patients
and volunteers were significant (
P .05). For the patients and
volunteers examined without a breath hold, the median for the
SI changes was similar, whereas that for the SI slopes was different
(
Table).
| fig.ommitted |
Figure 1. Coronal base MR images of the lungs in a volunteer were acquired with the sequence with the global inversion pulse during the breathing of room air (A) or 100% oxygen (B). Pulmonary parenchyma and pulmonary vessels can be seen clearly.
| |
| fig.ommitted |
Figure 2a. Graphs depict the mean SI change values for patients during the breathing of 100% oxygen compared with DLCO during MR imaging with (a) the global inversion pulse or (b) the section-selective inversion pulse. In a, the solid line indicates the linear regression for the breath-hold data, whereas the dashed line indicates that for the non-breath-hold data. The regression model for the former was y = 0.1899x + 5.179 (R2 = 0.81) and for the latter was y = 0.1622x + 5.9353 (R2 = 0.9073). The SI change for the patient with pulmonary emphysema (DLCO = 42%) shows a very large SD (in a). This is due to the very distinct regional difference in SI change between the left and right lungs (see text). Also, the other large SDs (in a and b) are due to large local differences in SI changes within the lung. In b, the solid line indicates the linear regression for the breath-hold data. The regression model is y = 0.0754x + 1.3482 (R2 = 0.039).
| |
| fig.ommitted |
Figure 2b. Graphs depict the mean SI change values for patients during the breathing of 100% oxygen compared with DLCO during MR imaging with (a) the global inversion pulse or (b) the section-selective inversion pulse. In a, the solid line indicates the linear regression for the breath-hold data, whereas the dashed line indicates that for the non-breath-hold data. The regression model for the former was y = 0.1899x + 5.179 (R2 = 0.81) and for the latter was y = 0.1622x + 5.9353 (R2 = 0.9073). The SI change for the patient with pulmonary emphysema (DLCO = 42%) shows a very large SD (in a). This is due to the very distinct regional difference in SI change between the left and right lungs (see text). Also, the other large SDs (in a and b) are due to large local differences in SI changes within the lung. In b, the solid line indicates the linear regression for the breath-hold data. The regression model is y = 0.0754x + 1.3482 (R2 = 0.039).
| |
| fig.ommitted |
SI Change Values with Two Pulse Sequences at Oxygen-enhanced MR Imaging in the Lung
| |
Only a small correlation was detected between the values for
SI change and D
LCO with the section-selective inversion pulse.
Large SI changes tended to be associated with large D
LCO values.
For the section-selective inversion pulse, the Pearson correlation
coefficient of the linear regression was
r = 0.20 (
r2 = 0.04).
For the global inversion pulse, the Pearson correlation coefficient
was
r = 0.90 (
r2 = 0.81) with a breath hold and
r = 0.95 (
r2 = 0.91) without a breath hold.
On the SI change maps for the volunteers, the SI change due to the breathing of 100% oxygen can be clearly seen within the lung area, in the spleen and aorta (Fig 3). In contrast, local decreases in SI change were found in two patients: one with pulmonary emphysema mainly in the right upper lung lobe (Fig 4a) and one with pulmonary fibrosis (Fig 5).
| fig.ommitted |
Figure 3. Left: Coronal SI change (enhancement) map in a volunteer was obtained during a breath hold with the sequence with a global inversion pulse. Right: Corresponding color-coding map. Relatively large and homogeneous SI change (green to red) can be seen in the map.
| |
| fig.ommitted |
Figure 4a. Images in a 40-year-old female patient with pulmonary emphysema were obtained with the sequence with a global inversion pulse with a breath hold. (a) Left: Coronal SI change (enhancement) map for the left and right lungs. Right: Corresponding color-coding map. A striking difference in the SI change due to bullous changes between the two lungs can be clearly seen. Almost no SI change is seen in the right lung (blue), whereas the left lung clearly shows a positive SI change (green). (b) Anteroposterior radiograph of the chest depicts the right lung with bullous changes (arrows) due to emphysema that are more extensive than those seen in the left lung. Findings in a correlate well with those in b. (c) Left: Coronal SI slope map shows a very low SI slope (blue) in the right lung and a nearly normal SI slope (green) in the left lung. Right: Corresponding color-coding map.
| |
| fig.ommitted |
Figure 4b. Images in a 40-year-old female patient with pulmonary emphysema were obtained with the sequence with a global inversion pulse with a breath hold. (a) Left: Coronal SI change (enhancement) map for the left and right lungs. Right: Corresponding color-coding map. A striking difference in the SI change due to bullous changes between the two lungs can be clearly seen. Almost no SI change is seen in the right lung (blue), whereas the left lung clearly shows a positive SI change (green). (b) Anteroposterior radiograph of the chest depicts the right lung with bullous changes (arrows) due to emphysema that are more extensive than those seen in the left lung. Findings in a correlate well with those in b. (c) Left: Coronal SI slope map shows a very low SI slope (blue) in the right lung and a nearly normal SI slope (green) in the left lung. Right: Corresponding color-coding map.
| |
| fig.ommitted |
Figure 4c. Images in a 40-year-old female patient with pulmonary emphysema were obtained with the sequence with a global inversion pulse with a breath hold. (a) Left: Coronal SI change (enhancement) map for the left and right lungs. Right: Corresponding color-coding map. A striking difference in the SI change due to bullous changes between the two lungs can be clearly seen. Almost no SI change is seen in the right lung (blue), whereas the left lung clearly shows a positive SI change (green). (b) Anteroposterior radiograph of the chest depicts the right lung with bullous changes (arrows) due to emphysema that are more extensive than those seen in the left lung. Findings in a correlate well with those in b. (c) Left: Coronal SI slope map shows a very low SI slope (blue) in the right lung and a nearly normal SI slope (green) in the left lung. Right: Corresponding color-coding map.
| |
| fig.ommitted |
Figure 5. Left: Coronal SI change map in a 57-year-old female patient with pulmonary fibrosis. Right: Corresponding color-coding map. A strong effect (red) due to the breathing of 100% oxygen can be seen in the aorta. A regional difference in the severity of the fibrosis in this patient was evaluated with computed tomography. The pathologic alterations were most advanced in the cranial portions of the lungs. (The SI change map in Fig 4a depicts such regional differences in the disease process.) The more cranial parts of the lungs show a very small (blue) SI change, whereas the more caudal parts show a larger (green) SI change.
| |
After the switch to 100% oxygen, the SI time courses showed
a clearly delayed SI increase in the lungs of patients compared
with the volunteers (
Table). For all volunteers, the mean values
for the SI slopes averaged over all regions of interest were
13.8% per minute ± 4.8 with a breath hold and 9.8% per
minute ± 1.1 without a breath hold with use of the pulse
sequence with a global inversion pulse. For the same sequence,
the mean values for the SI slopes of the patients were 6.5%
per minute ± 2.2 and 4.3% per minute ± 2.0, respectively.
The corresponding values for D
LCO ranged from 23% to 80% (percentage
of reference value) (
Fig 6). The correlation coefficient for
the linear regression for the SI slopes and the D
LCO values
was 0.86 (
r2 = 0.74) for the patients examined with the breath-hold
technique. In some cases, the SDs for the values of the SI slopes
were relatively large (
Fig 6). This was due to strong spatial
variations in the SI slope value over the lung area.
| fig.ommitted |
Figure 6. In patients examined with the sequence with a global inversion pulse, graph depicts the values for SI slopes in the lungs after the switch to 100% oxygen compared with DLCO (as a percentage of the reference value)= without a breath hold. The linear regression model (solid line) for the examinations with a breath hold is y = 0.1131x + 1.678 (R2 = 0.74). The somewhat large SDs are due to a pronounced spatial heterogeneity in the SI slopes. All but four patients had pulmonary fibrosis. Diagnoses for those four patients were (a) extrinsic allergic alveolitis (DLCO = 23%), (b) emphysema (DLCO = 42%), (c) adult respiratory distress syndrome (DLCO = 40%), and (d) pulmonary hypertension (DLCO = 80%).
| |
The Pearson correlation coefficient for the linear regression
between D
LCO and SI slope data was 0.90 (
r2 = 0.81) (
Fig 7).
The mean value of the SI slope for the volunteers was 12.3%
per minute ± 4.2 (section-selective inversion pulse).
Differences in the SI slopes were significant (
P = .05), which
suggests that the SI slopes of patients were smaller than those
of volunteers.
| fig.ommitted |
Figure 7. In patients examined with the sequence with a section-selective inversion pulse, graph depicts the values for SI slope in the lungs after the switch to 100% oxygen compared with DLCO (as a percentage of the reference value). The linear regression model (solid line) for the examinations with a breath hold is y = 0.1371x - 1.638 (R2 = 0.8099). The large SDs are due to a strong spatial heterogeneity in the SI slopes. Four of these patients had pulmonary fibrosis (DLCO = 37%, 45%, 52%, and 53%, respectively). The remaining patient had emphysema in the native lung and had undergone unilateral lung transplantation; for this patient, only SI slope values in the native lung were averaged.
| |
The SI slope maps for the volunteers showed relatively high
and homogeneous SI slopes (
Fig 8) compared with those for the
patients (
Figs 4c,
9). In the patient with pulmonary emphysema,
a difference in the SI slope between the right and left lungs
after the switch from room air to 100% oxygen can be clearly
seen (
Fig 4c). This finding is consistent with the SI change
map for this patient (
Fig 4a). The SI slope for the lungs of
the patient with pulmonary fibrosis was smaller after the switch
to 100% oxygen than that for the volunteers (
Figs 8,
9).
| fig.ommitted |
Figure 8. Left: Coronal SI slope map for the lungs in a 28-year-old male volunteer examined with the sequence with a global inversion pulse. Right: Corresponding color-coding map. In the SI slope map, the slope is relatively high (green to yellow).
| |
| fig.ommitted |
Figure 9. Left: Coronal SI slope map in a 46-year-old female patient with lung fibrosis examined with the sequence with a global inversion pulse without a breath hold. Right: Corresponding color-coding map. The values for the SI slopes are smaller (blue) than those in Figure 8.
| |
Results of the statistical test of the differences in the SI
changes and SI slopes suggested that the values for the patients
were smaller than those for the volunteers. The SI change and
SI slope maps revealed local differences in lung function. The
mean values of the SI slopes had a good to excellent correlation
with the D
LCO values for the patients. However, this study is
somewhat limited by the small number of patients and their various
lung diseases.
In comparison with SI change maps, the postprocessing method with SI slope values allowed determination of the SI dynamics when subjects started to breathe 100% oxygen. Results of the statistical test suggested differences in the SI slope after the switch to 100% oxygen between patients with pathologic diffusion of oxygen (emphysema and interstitial diseases, such as fibrosis and extrinsic allergic alveolitis) and volunteers. Therefore, a delayed increase in the SI after the switch to 100% oxygen may be a parameter to help detect lung disease, which is related to the pathologic diffusing capacity of the lungs. As a main advantage of the MR angiographic technique used in this study over DLCO, regional differences in diffusion were observed; therefore, variations in the disease could be visualized. The values of the SI slopes after the switch to 100% oxygen correlate with the values of DLCO. This method allows the assessment and imaging of ventilation, perfusion, and oxygen transfer through the alveolar-capillary membrane (15), which is distinctly different from other MR imaging techniques. For example, ventilation can be determined with MR imaging by using hyperpolarized noble gases (17–20), and perfusion can be assessed with contrast material–enhanced MR imaging (21–23) or other techniques, such as arterial spin labeling (24,25).
A similar postprocessing procedure for oxygen-enhanced MR imaging has been proposed by Hatabu et al (26), who investigated ventilation defects in a patient with bronchial narrowing caused by lung cancer. Postprocessing was performed by means of exponential fitting of the data for the SI time course. In contrast, the aim of this study was to determine the oxygen transfer from the alveoli into the capillaries; therefore, we used a simplified linear approach that is tolerant of noisy data.
To obtain temporally resolved data about the inflow of oxygen into the lung capillaries, the SI slope maps representing the upslope in SI after the switch to 100% oxygen are helpful. For this issue, as well as for the assessment of regional differences in lung function, both the SI slope and SI change offer valuable information.
A strong correlation was found between the SI slope values after the switch to 100% oxygen and the corresponding values for DLCO. A weaker correlation was found between the SI change and DLCO values in the case of the section-selective inversion pulse. These results suggest that the calculation of SI change and SI slope maps may help determine regional differences in the lung. The calculation of SI slopes and SI slope maps may represent a modality that can help determine a spatially resolved DLCO value. In support of this theory is the finding that the statistical test revealed no difference in the mean SI change between the volunteers and patients in three cases.
In the volunteers, the percentage SI change decreased slightly (2.7%) with the section-selective compared with the global inversion pulse (both with a breath hold), which indicates that the section-selective sequences with a 1.3-second inversion time were insensitive to inflow effects, even though the mean transit time of red blood cells is approximately 1 second at rest. This may be explained by the fact that if inflow effects existed, they were measured similarly during the breathing of room air or 100% oxygen, which led to a reduction in the SI.
A comparison of findings with the sequences with global or section-selective inversion pulses (both with a breath hold) showed that the values and SDs for the SI slopes and SI changes for the volunteers were similar. This indicates that the same results were obtained regardless of the pulse sequence used. This also indicates that the inflow effects associated with the section-selective inversion pulse did not play an important role.
The SI of the pulmonary veins did not always increase, although the technique measures the oxygenation of blood. An explanation of this finding may be that minimal diaphragm motion existed, despite the fact that the only images used showed a diaphragm position within a certain tolerance span. This motion could reduce the SI, especially in the pulmonary veins. Therefore, images obtained during the breathing of 100% oxygen could show the same or, in the worst case, even lower SI than the images obtained during the breathing of room air.
A difference of about 4% per minute (13.8% per minute with a breath hold, 9.8% per minute without a breath hold) could be detected between the SI slope values calculated for breath-hold and non–breath-hold measurements. Theoretically, no difference should have been found, because oxygen diffusion should remain the same with or without a breath hold. However, the motion of the diaphragm was definitely more pronounced without a breath hold. Therefore, the SI varied in a larger range for the measurements obtained without a breath hold. This could have effected the calculation of the SI slopes.
The purpose of this study was to compare SI slopes and SI changes with DLCO. Future investigations should be performed to determine whether this method allows the differentiation of pulmonary disease. Although the current study included patients with a variety of diseases, the number of patients with lung diseases other than pulmonary fibrosis was too small to allow assessment of differences in SI slope and SI changes in different diseases.
Our findings demonstrate that normal and pathologic lungs exhibit definite differences in SI slopes and SI changes during the breathing of 100% oxygen. Moreover, major differences between different regions of the lung were visualized in patients. Both the SI change and SI slope maps allowed determination of regional differences. However, only the SI slope maps and corresponding values were reliable parameters that represented a spatially resolved DLCO.
The authors are very grateful to T. Seitz, M. Giessel, F. Greil,
and M. Felbinger from the technical anesthesia department (University
of Munich) for volunteering their time and expertise in the
construction of the breathing system and to Martin Dugas, MD,
(University of Munich) for his excellent statistical assistance.
- Fraser RG, Pare JA, Pare PD, Fraser RD, Genereux GP, eds. Methods in clinical, laboratory, and functional investigation In: Diagnosis of diseases of the chest. 3rd ed. Philadelphia, Pa: Saunders, 1988; 388-457.
- Ogilvie CM, Forster RE, Blakemore WS, Morton JW. A standardized breath holding technique for the clinical measurement of the diffusing capacity of the lung for carbon monoxide. J Clin Invest 1957; 36:1-17.
- Edelman RR, Hatabu H, Tadamura E, Li W, Prasad PV. Noninvasive assessment of regional ventilation in the human lung using oxygen-enhanced magnetic resonance imaging. Nat Med 1996; 2:1232-1236.
- Stock KW, Chen Q, Morrin M, Hatabu H, Edelman RR. Oxygen-enhanced magnetic resonance ventilation imaging of the human lung at 0.2 and 1.5 T. J Magn Reson Imaging 1999; 9:838-841.
- Chen Q, Jakob PM, Griswold MA, Levin DL, Hatabu H, Edelman RR. Oxygen enhanced MR ventilation imaging of the lung. MAGMA 1998; 7:153-161.
- Young IR, Clarke GJ, Bailes DR, Pennock JM, Doyle FH, Bydder GM. Enhancement of relaxation rate with paramagnetic contrast agents in NMR imaging. J Comput Tomogr 1981; 5:543-546.
- Gore JC, Doyle FH, Pennock JM. Relaxation rate enhancement observed in vivo by NMR imaging. In: Partain CL, James AE, Rollo FD, Price RR, eds. Nuclear magnetic resonance imaging. Philadelphia, Pa: Saunders, 1983; 94-106.
- Paulling L, Coryell C. The magnetic properties and structure of the hemochromogens and related substances. Proc Natl Acad Sci U S A 1936; 22:159-163.
- Tadamura E, Hatabu H, Li W, Prasad PV, Edelman RR. Effect of oxygen inhalation on relaxation times in various tissues. J Magn Reson Imaging 1997; 7:220-225.
- Tripathi A, Bydder GM, Hughes JMB. Effect of oxygen tension on NMR spin-lattice relaxation rate of blood in vivo. Invest Radiol 1984; 19:174-178.
- Brooks RA, Chiro GD. Magnetic resonance imaging of stationary blood: a review. Med Phys 1987; 14:905-913.
- Thulborn KR, Waterton JC, Matthews PW, Radda GK. Oxygenation dependence of the transverse relaxation time of water protons in whole blood at high field. Biochim Biophys Acta 1982; 714:265-270.
- Müller C, Löffler R, Penzkofer H, et al. Breathhold MR-imaging of the human lungs combined with the calculation of T1-maps during breathing of room air and 100% oxygen (abstr). MAGMA 1998; 6(suppl):100.
- Penzkofer H, Löffler R, Peller M, et al. Multisection ventilation imaging of the human lung after breathing of 100% oxygen (abstr). In: Proceedings of the Sixth Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 1998; 1:454.
- Löffler R, Müller CJ, Peller M, et al. Optimization and evaluation of the signal intensity change in multisection oxygen enhanced MR lung imaging. Magn Reson Med 2000; 43:860-866.
- Penzkofer H, Wintersperger BJ, Knez A, Weber J, Reiser M. Assessment of myocardial perfusion using multisection first-pass MRI and color-coded parameter maps: a comparison to 99mTc SESTA MIBI SPECT and systolic myocardial wall thickening analysis. Magn Reson Imaging 1999; 17:161-170.
- Kauczor HU, Ebert M, Kreitner KF, et al. Imaging of the lungs using 3He MRI: preliminary clinical experience in 18 patients with and without lung disease. J Magn Reson Imaging 1997; 7:538-543.
- Kauczor HU, Surkau R, Roberts T. MRI using hyperpolarized noble gases. Eur Radiol 1998; 8:820-827.
- Mugler JP, III, Driehuys B, Brookeman JR, et al. MR imaging and spectroscopy using hyperpolarized 129Xe gas: preliminary human results. Magn Reson Med 1997; 37:809-815.
- Kauczor HU, Hofmann D, Kreitner KF, et al. Normal and abnormal pulmonary ventilation: visualization at hyperpolarized He-3 MR imaging. Radiology 1996; 201:564-568.
- Hatabu H. MR pulmonary angiography and perfusion imaging: recent advances. Semin Ultrasound CT MR 1997; 18:349-361.
- Hatabu H, Tadamura E, Levin DL, et al. Quantitative assessment of pulmonary perfusion with dynamic contrast-enhanced MRI. Magn Reson Med 1999; 42:1033-1038.
- Berthezène Y, Croisille P, Wiart M, et al. Prospective comparison of MR lung perfusion and lung scintigraphy. J Magn Reson Imaging 1999; 9:61-68.
- Mai VM, Berr SS. MR perfusion imaging of pulmonary parenchyma using pulsed arterial spin labeling techniques: FAIRER and FAIR. J Magn Reson Imaging 1999; 9:483-487.
- Mai VM, Hagspiel KD, Christopher JM, et al. Perfusion imaging of the human lung using flow-sensitive alternating inversion recovery with an extra radiofrequency pulse (FAIRER). Magn Reson Imaging 1999; 17:355-361.
- Hatabu H, Tadamura E, Li W, Prasad PV, Edelman RR. Dynamic evaluation of pulmonary ventilation by MR imaging with inhalation of molecular oxygen (abstr). Radiology 1997; 205(P):268.
作者：
Christian J. Müller MS Martin Schwaiblmair MD 2007-5-12