| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Developments |
1 From the Department of Radiology (S.K.C., J.P.F., C.E.B., R.L.E., H.H.H.), Section of Safety (K.L.C.), and Division of Biostatistics (T.L.H.), Mayo Clinic, 200 First St SW, Rochester, MN 55905. From the 2002 RSNA scientific assembly. Received November 20, 2002; revision requested January 13, 2003; revision received March 17; accepted April 2. Supported in part by the General Electric–Association of University Radiologists Radiology Research Fellowship Program. Address correspondence to S.K.C. (e-mail: scarlson@mayo.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMaterials and MethodsResults DiscussionREFERENCES |
|---|
© RSNA, 2003
Index terms: Computed tomography (CT), guidance, 60.12112 • Computed tomography (CT), technology, 60.12112 • Lung, biopsy, 60.12112 • Lung, CT, 60.12112
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResults DiscussionREFERENCES |
|---|
The major drawback to exclusive use of the intermittent mode is that respiratory motion can be a major problem during procedures performed in the lung or upper abdomen. Biopsy of small lung nodules is particularly challenging because of respiratory motion and inconsistent breath holding by the patient. During normal breathing, tumors in the lung can move 1–3 cm (7), and diaphragm motion can cause positional variation of the upper abdominal organs of 1.50–6.00 cm in the superior-inferior direction (8,9). Considerable variation in the degree of lung inflation and upper abdominal organ position can also occur in patients with no known lung abnormality despite consistent breath-hold instructions (10,11). Because intermittent-mode CT fluoroscopy allows imaging only in the transverse plane, inconsistent breath holding by the patient can cause target structures to move out of the field of view during intervention and reproducibility decreases. This results in prolonged procedure and CT fluoroscopy radiation exposure times. There is also the potential for decreased diagnostic yield of the biopsy specimen and higher complication rates.
To optimize CT fluoroscopy–guided biopsies of the lung or upper abdomen, techniques are needed to ensure accurate and reproducible breath holding by the patient. Breath-hold monitoring and feedback systems are being used successfully in radiation therapy protocols to enable selective delivery of higher and more effective doses of radiation to moving targets, thereby decreasing irradiation of adjacent healthy tissues (12). There has been extensive clinical use of breath-hold monitoring and feedback devices in magnetic resonance (MR) imaging to decrease image artifacts due to respiratory motion (13,14). These systems allow external monitoring of changes in body-wall girth or position, which strongly correlate with changes in diaphragm position, and thus internal lesion location (14,15). To our knowledge, the use of a breath-hold monitoring and feedback system in CT fluoroscopy–guided biopsies has not been investigated.
The purpose of this study was to develop and evaluate the feasibility of a bellows-based breath-hold monitoring and feedback system for use in intermittent-mode CT fluoroscopy–guided biopsies of the lung or upper abdomen.
|
Materials and Methods |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResults DiscussionREFERENCES |
|---|
Respiratory Bellows System
The respiratory bellows is a user-friendly system that was developed and modified by medical physicists and engineers at our institution. It consists of a hook and loop belt with expandable bellows that is wrapped around a patient’s upper abdomen or lower chest, depending on which area expands the most during inspiration (Fig 1). The bellows belt can be adjusted to fit any patient size. Variation in the length of the bellows during breathing, which corresponds to changes in chest or abdominal girth, causes pressure change within the tubing that is measured with a pressure-sensitive transducer (Mayo Clinic, Rochester, Minn). The transducer analog signal is digitized and displayed on a light-emitting diode (LED) monitor for patient feedback (Fig 2). The patient LED monitor is attached directly to the CT table on an articulated arm that can be adjusted to any position (Fig 3). The bellows is connected by rubber and silicone tubing to a central system control unit located next to the CT operator console. The system control unit consists of three switches: a power switch, a "zero" switch that allows manual adjustment of the display zero point (if needed) after the power-up phase, and a solenoid switch that closes the air flow to and from the bellows, which allows the operator to correlate the center red LED on the display with the preselected breath-hold level obtained during preprocedure localization CT (Fig 4).
|
|
|
|
Technique
The bellows belt is placed on the patient before CT scanning. Localization images are then obtained through the area of interest. During localization imaging, the solenoid switch of the bellows is turned off and on when the patient’s position has stabilized at a specific breath-hold level to be used as a reference. This correlates the pressure level associated with the patient’s reference breath-hold level to the center red diode on the display. The localization images are then reviewed to determine if the lesion is optimally visualized and percutaneously accessible at this reference breath-hold level. If the level is deemed suitable, the patient is then instructed to reproduce this same breath-hold level as needed throughout the procedure. By viewing the LED display, patients continuously monitor their breath-hold level and can consistently inhale or exhale to return to the reference level. When this is achieved, the center red diode is continuously lit and an intermittent CT fluoroscopy image is obtained to document target lesion position. The radiologist then places the needle, and further intermittent CT fluoroscopy images are obtained to document positioning of the needle tip within the target lesion.
Sensitivity and Reliability Testing of the Bellows System
Technical evaluation of sensitivity for detection of body-wall motion and reliability was performed by using a motor-driven mechanical body-wall phantom (Fig 5). The phantom was designed to simulate typical patient respiratory motion within a physiologic range of excursion and allowed for variable amplitude of excursion.
|
System reliability.—Repeated sinusoidal motion designed to mimic typical patient body-wall motion was used to measure signal repeatability over a range of respiratory amplitudes. Measurements during 150 repeated cycles with a period of 2.2 seconds and varied amplitudes (3.6, 6.0, 10.0, 15.0, and 20.0 mm) were recorded on a digital oscilloscope.
Patient and Radiologist Interaction with Bellows System
Human volunteer trials were performed to evaluate patient and radiologist interaction with the bellows system and adaptability of the system for use in the CT fluoroscopy setting. Trials with 10 healthy human volunteers (residents and staff members without known lung disease) were designed and conducted to mimic the clinical environment. The 10 volunteers comprised six women and four men (age range, 22–64 years; mean, 43 years). Each volunteer used the bellows system during 10 simulated CT acquisitions (no radiation) with suspended respiration at different specified levels. All volunteers were tested in three common physical positions for biopsy (supine, prone, and lateral decubitus) during three common levels of suspended respiration (small inspiration, medium inspiration, and expiration). Volunteers completed a questionnaire after their trial to address their opinion of the bellows system with regard to comfort and ease of use. A five-point grading scale was used: 1 = unacceptable, 2 = poor, 3 = acceptable, 4 = good, 5 = excellent. One radiologist (S.K.C.) conducted all the trials and also completed a questionnaire after each trial. In this questionnaire, the same grading scale was used to rank the interactive nature of the system and the practicality in the CT fluoroscopy setting. Specifically, these questions related to respiratory signal visualization by the radiologist signal response time, ability of the radiologist to specify a reference breath-hold level, ability of the volunteer to reproduce the breath-hold level, ease of system setup, flexibility of the system, and adaptability of the system to the CT fluoroscopy setting. Radiologist ratings were collected and summed across the seven questions to get an overall rating with a possible range from 7 to 35 (seven questions, each scored on a scale of 1–5).
Correlation of Breath-hold Level with Internal Target Location
Monitoring of external body-wall motion correlates strongly with diaphragm position and, thus, internal target lesion location (13–15). Before its use in the clinical setting, we wanted to confirm this finding with our bellows system. Two experiments were performed with three human volunteers (one woman and two men; age range, 22–39 years; mean, 32 years) with the bellows system and sagittal single-shot two-dimensional gradient-echo MR imaging (fast acquisition; repetition time msec/echo time msec, 34/1.8 (effective); bandwidth, ±31.25 kHz; excitation flip angle, 45°; matrix size, 512 x 160; section thickness, 10 mm; intersection gap, 2 mm; field of view, 40 cm; one signal acquired). MR imaging was used owing to the lack of ionizing radiation and the ability to acquire sagittal MR images, which allowed us to measure changes in diaphragm position more accurately than is possible with CT scans.
In the first experiment, the three volunteers were asked to hold their breath at different levels, starting with the center diode, then three diodes above center, three diodes below center, four diodes above the center, and four diodes below the center. By using MR software, the internal positions of the diaphragm and a specified blood vessel in the lower lung were measured and recorded with respect to an external chest wall marker during each breath-hold attempt. Measurements were performed by the principal investigator of the study (S.K.C.). This allowed us to correlate the breath-hold level (diode position) with the internal target location (motion in the superior-inferior direction) and to determine the approximate change in diaphragm and lower lung blood vessel position per single unit change in diode position on the visual display (eg, how many millimeters does each diode change represent?).
In the second experiment, the same volunteers were asked to perform 20 consecutive breath holds (10 on small inspiration and 10 on end expiration), with the goal of lighting only the center diode on the LED visual display. The diaphragm and blood vessel positions were again measured and recorded during each breath hold. This also allowed us to correlate the breath-hold level with internal target location and to further evaluate the reliability of the system.
Feasibility of Bellows System in Clinical Setting
We used the bellows system in seven biopsy procedures in mobile lung or upper abdominal lesions where respiratory motion was deemed to be a potential problem. Inclusion criteria included lung or upper abdominal lesions (including liver, spleen, adrenal, renal, stomach, and mesenteric masses) 3 cm or less in longest dimension that were not fixed to the abdominal wall or pleura. The seven patients comprised two women and five men (age range, 45–73 years; mean age, 58 years) who ranged in weight from 63 to 110 kg. Data from these procedures were collected during and after completion of the biopsy by using the questionnaire that was completed by the study coordinator, CT technologist, and radiologist who performed the biopsy. The questionnaire specifically addressed bellows setup and patient training time, ability of the patient to cooperate with the bellows system, the number of needle puncture attempts, lesion size, diagnostic yield of the biopsy specimen, CT fluoroscopy radiation exposure time, needle placement time, and procedure-related complications. All biopsies were performed with intermittent-mode CT fluoroscopy and either 17-18-gauge or 19-20-gauge coaxial automated core needle biopsy systems.
Statistical Methods
Reliability of the bellows system was assessed by using amplitude data measured at five displacement levels (range, 3.6–20.0 mm). The coefficient of variation was estimated at each displacement level by dividing the SD by the mean. The coefficient of variation is a measure of the relative variability of the repeated amplitude measurements with respect to the mean value at each displacement level.
Data from the first experiment with MR imaging were plotted, and a linear regression line was fit to the data. Diaphragm and lung blood vessel positions were each plotted versus the corresponding diode position. The slope of the regression line was then interpreted as the average change in diaphragm or blood vessel position for each one-diode change in the bellows system. A separate slope was calculated for each of three volunteers. The individual slopes and the mean of the three slopes were reported separately for diaphragm and blood vessel position. Ninety-five percent CIs were calculated for the mean slope. Data obtained during the second experiment (multiple breath holds to light only the center diode) were used to calculate the SDs for diaphragm and blood vessel positions as volunteers reproduced the reference breath-hold level (center diode). This portion of the study was performed primarily to aid the investigators in their clinical practice. Because data from these MR experiments were collected from only three volunteers, with multiple observations for each, these results are reported primarily for descriptive purposes.
Data collected from use of the bellows system in the clinical setting were summarized with means, ranges, frequencies, and proportions, as appropriate. CIs were calculated for estimated means and proportions.
|
Results |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResults DiscussionREFERENCES |
|---|
System reliability.—The system proved to be reliable, with extremely small coefficients of variation at multiple amplitudes (breath-hold levels). At the displacement levels of 3.6, 6.0, 10.0, 15.0, and 20.0 mm, the coefficients of variation were 1.9%, 1.4%, 0.7%, 0.8%, and 0.7%, respectively.
Patient and Radiologist Interaction with Bellows System
Volunteer ratings with regard to comfort and ease of use were consistently high (scores of 4 or 5 on a scale of 1–5). The volunteers were able to adjust their breathing and consistently reproduce the reference breath-hold level (center diode) with minimal instruction. Radiologist ratings with regard to the interactive nature of the system were also consistently high. In 26 of 30 trials during small inspiration (10 volunteers in three positions), the summation score reached the maximum of 35. The remaining four trials produced scores of 29, 30, 33, and 34, with the scores of 29 and 33 coming from the same volunteer in the supine and prone positions, respectively. The scores of 30 and 34 came from two other volunteers.
Correlation of Breath-hold Level with Internal Target Location
Data from the first experiment performed with MR imaging showed strong linear correlation between breath-hold level and internal target location (r2 = 0.84–0.94). Calculation of the slope of the linear regression line allowed us to determine the average number of millimeters each one-unit change in diode position represents. Measurements of diaphragm and lower lung blood vessel position showed an average change of 3.6 mm (95% CI: 1.5, 5.6) per diode and 2.5 mm (95% CI: 0.8, 4.1) per diode, respectively. The individual slopes for the three volunteers were 3.3, 3.0, and 4.5 for the diaphragm position and 2.5, 1.8, and 3.1 for the lower lung blood vessel position.
During the second experiment, the volunteers were able to light the center diode consistently during each breath-hold attempt. There was negligible variation in target structure (diaphragm and blood vessel) position measurements when the reference breath-hold level (center diode) was reproduced (Table).
|
|
|
|
|
|
|
Discussion |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResults DiscussionREFERENCES |
|---|
Accurate and safe CT fluoroscopy–guided percutaneous biopsy procedures in the lung or upper abdomen require precise and reproducible breath holding by the patient. Many patients, however, are unable to reproduce consistent levels of suspended inspiration or expiration. Use of a patient breath-hold monitoring and feedback system can significantly decrease the variability of diaphragm position (and thus internal target location) during consecutive suspended respirations. Findings in a study in the MR imaging literature show that average motion of the patient’s diaphragm can be reduced from 8.3 to 1.3 mm by using breath-hold monitoring and feedback (14). Findings in another study in the radiation oncology literature showed similar results, with a reduction in diaphragm variability from 1.4 to 0.3 cm (15).
In the early and middle 1980s, two articles described the use of similar breath-hold monitoring and feedback systems during diagnostic CT scanning to improve image quality by eliminating respiratory motion artifact, improve multiplanar reconstruction techniques, and assist in time-dependent delivery of contrast medium (16,17). Results in a study showed increased diagnostic accuracy with use of a respiratory gating system during conventional CT–guided biopsy of small pulmonary nodules (18). We modeled our study on these previous studies to show the usefulness of a breath-hold monitoring and feedback system in CT fluoroscopy–guided procedures.
Overall, the bellows system is easy to use, easily adapted to the CT fluoroscopy setting, and well received by patients. The system is sensitive and reliable for monitoring changes in body-wall girth. It provides instantaneous feedback to patients and allows them to adjust their breath-hold levels accordingly. Other breath-hold devices have been studied (13,15). The bellows system is comfortable and does not impair the patient’s normal breathing process. It is well suited for all CT scanners, requires minimal installation and space in the CT suite, and does not cause artifacts during CT scanning. Although we routinely place patients feetfirst into the CT gantry for scanning and procedures, the bellows system can also be used when the patient is placed headfirst into the gantry, as long as the patient’s LED monitor clears the gantry and is adjusted so that it is not directly between the CT beam and the region of interest.
Results obtained during the MR imaging experiments that demonstrated strong correlation between the reference breath-hold level and internal target location offer practical information for use of the bellows system in the clinical setting. Measurements showed an average change of diaphragm and lower lung blood vessel position of 3.6 mm per diode and 2.5 mm per diode, respectively. This information helped us guide the patient’s breath holding. For patients with small lesions (<1 cm), lighting of the center diode during breath holding is preferred. For patients with larger lesions, the target structure may be visualized adequately even if the patient is having difficulty lighting the center diode and is consistently lighting a diode above or below the center. Results also showed that there was negligible variation in target structure position when the reference breath-hold level was reproduced. This finding confirms our belief that if the patient can consistently reproduce the reference breath-hold level by lighting the center diode (or one diode level above or below the center diode), the target lesion should be visualized adequately during the procedure.
The bellows system setup and patient training times added little extra time to the procedure. Most patients have been abdominal breathers (ie, have greater abdominal excursion than chest excursion with inspiration), and ideal placement of the bellows belt in these patients is usually just below the xyphoid. We used the bellows system with the patients in the supine, lateral decubitus, and prone positions, and it performed well in all cases. As expected, there tends to be less respiratory excursion in the prone position than in the supine and lateral decubitus positions. This does not affect the bellows sensitivity to detect and monitor motion. The bellows belt is thin and has interfered with the intended needle biopsy site in only one instance so far. In this case, the belt was moved 2 cm superiorly, and a second set of localization images was obtained to reestablish the reference breath-hold level at the new bellows position.
In the clinical setting, we found the bellows system to be particularly beneficial in several instances. First, in biopsies of small peripheral lung lesions, access is often difficult because of overlying ribs. The bellows system can help the patient attain different levels of inspiration or expiration until the radiologist can find a level that allows adequate access to the lesion. Second, the bellows system can overcome changes in breath-hold level that occur between preprocedural CT scanning and the start of the procedure. The level of inspiration or expiration can change dramatically after the intravenous administration of conscious sedation medication. Breath-hold levels can also change once the procedure has begun, as a result of pain or anxiety. Bellows system monitoring and feedback can decrease the variation due to these factors and allow the patient to reproduce the reference breath-hold level obtained during localization imaging. In the sedated patient, this can also aid in keeping the patient’s percentage of oxygen saturation in the normal range by stimulating the patient to take in adequate amounts of air. Third, the bellows system can be helpful in parenchymal biopsies of the upper abdomen when the lesion cannot be visualized after washout of the intravenously administered contrast material. If the patient reproduces the same breath-hold level that was used for the contrast material–enhanced study when the lesion was visualized, the radiologist can use landmarks to access the lesion even though it is no longer visible on the delayed CT fluoroscopy images. Finally, the bellows system helps distract nervous patients during the procedure by giving them something to concentrate on other than the needle.
A limitation of our study is the small sample size of 10 healthy volunteers. The system may work differently with older or sicker patients or with patients who have a different body habitus. Another limitation is that during the phantom and volunteer trials to determine sensitivity, reliability, and patient interaction with the system, it was not feasible to dismantle the bellows system between trials because of time constraints. It is likely that the variability of the measurements would have been higher if we had been able to dismantle the device between trials and thus simulate patient-to-patient variability. However, in the seven clinical cases studied, we found no difference in the performance of the system with a range of patient ages and weights. A third limitation is that the bellows system requires patient cooperation with breath holding. There may be patients with chronic lung disease or other medical conditions who have difficulty holding their breath. The bellows system may be helpful in these patients to decrease the duration of the breath hold or to decrease the number of breath holds. In certain cases, percutaneous biopsy may not be the procedure of choice, and surgical biopsy may be warranted.
Before routine use of a breath-hold monitoring and feedback monitoring system can be implemented clinically, the efficacy of the system in the CT fluoroscopy setting should be demonstrated. A randomized controlled clinical trial is under way at our institution to determine the feasibility of the system in the clinical setting. The efficacy of bellows-assisted intermittent-mode CT fluoroscopy–guided biopsies of the lung or upper abdomen is compared with that of biopsies performed with intermittent-mode CT fluoroscopy without the aid of the bellows system. Our preliminary results suggest that this technique is feasible and useful. We believe that the bellows system will be a beneficial addition to the CT fluoroscopy interventional setting. The bellows system optimizes intermittent-mode CT fluoroscopy–guided procedures by helping the patient perform consistent breath holds at a preselected level, thus allowing target lesions to be visualized optimally throughout the procedure. We anticipate that there will be more applications for the bellows system in the future, including solitary pulmonary nodule enhancement studies, other dynamic perfusion studies, tumor ablation procedures, intratumoral injections, and possibly even robotic interventional procedures.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, S.K.C.; study concepts, S.K.C., J.P.F., R.L.E., C.E.B., K.L.C.; study design, S.K.C., J.P.F., H.H.H., T.L.H.; literature research, S.K.C.; clinical studies, S.K.C.; experimental studies, S.K.C., J.P.F., H.H.H., K.L.C., C.E.B.; data acquisition, S.K.C., J.P.F., H.H.H.; data analysis/interpretation, S.K.C., J.P.F., H.H.H., K.L.C., T.L.H.; statistical analysis, T.L.H., H.H.H.; manuscript preparation and definition of intellectual content, S.K.C.; manuscript editing, revision/review, and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMaterials and MethodsResults DiscussionREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  S. K. Carlson, J. P. Felmlee, C. E. Bender, R. L. Ehman, K. L. Classic, T. L. Hoskin, W. S. Harmsen, and H. H. Hu CT Fluoroscopy-guided Biopsy of the Lung or Upper Abdomen with a Breath-hold Monitoring and Feedback System: A Prospective Randomized Controlled Clinical Trial Radiology, November 1, 2005; 237(2): 701 - 708. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
