Dynamic Lung Volume Tracker and Controller

Inventors: Binh Q. Tran, Ph.D., Jehangir K. Tajik, Rajendra A. Chiplunkar, and Eric A. Hoffman

Overview

Pulmonary volumetric imaging using high resolution (i.e. less than 3mm slice thickness) x-ray CT provides a mechanism to quantify disease, standardize measurements across patients, and in a single patient, track progression or treatment of disease over time by repeatably scanning at a fixed lung volume. Advanced image processing techniques for quantifying pulmonary disease such as fractals, run-length encoding, etc. as well as efforts to characterize normal and abnormal lung tissue parameters are highly dependent upon the level of lung inflation at the time of data acquisition. Proper lung volume control during scanning also permits identification and comparison of the same parenchymal region over several lung inflation levels by following fiducial markers represented by airway branch points. This technique can then be used for example, to study airway reactivity due to methacholine (MCh) challenges which results in broncho-constriction.[1] By scanning at the same lung inflation level, airways before and after MCh inhalation can be identified and cross-sectional areas can be compared to study airway response to MCh. Volumetric imaging with appropriate control of lung volume also permits superimposing separately gathered ventilation and perfusion scans to study ventilation/perfusion matching. Present methods of accounting for lung inflation level require the patient to self-control the lung inflation level at each breathhold. Because scanning of the lung requires several breathhold maneuvers to complete an apex-to-base due to long image acquisition times and cannot be performed in one continuous breathhold, volumetric images are highly dependent upon patient cooperation and is subject to errors caused by patient repeatability and fatigue due to hyperinflation efforts. In patients demonstrating respiratory disease, these efforts are found to be unreliable and thus, lead to mismatch during volume rendering occurring at the slice level where two sequential breathhold maneuvers were performed. Additionally, we have found prolonged periods of apnea to be undesirable and led to decreased lung volume over time due to oxygen uptake. [2] Similar imaging problems related to respiratory motion arise when performing abdominal imaging protocols(i.e. liver, diaphragm, etc.).

Research and Development

To address these difficulties, we have developed a pneumotachometer-based device which continuously monitors patient inspiration and expiration in real-time for the purpose of tracking lung volume changes for the entire imaging duration. Figure 1 contains a picture containing the vital components of the volume controller equipment. Prior to scanning, the desired lung inflation level at which to perform scanning is input to the software and when the desired level is reached, the expiratory path is occluded and imaging begins. Real-time sampling of the pneumotach flow signal permits dynamic calculation of lung volume changes. This equipment utilized separate subject specific calibration coefficients for inspiratory and expiratory measurements, automatically determined by custom developed software algorithms, to correct for temperature, humidity and respiratory quotient, as well as flow directional non-uniformities. The humidified expiratory gas posed special problems in the development of this equipment.

Figure 1: Lung volume monitoring and control equipment as used on a test subject. The subject has the option of interrupting scanning with the push button switch in the event that further breathholding is not possible. Vital components consist of: (top) display monitor, transducers, occlusion software, (middle) pneumotach heater, power supply, electronic amplifiers, (bottom) keyborad, PC. Normally, equipment shown except for transducers and occlusion hardware reside in the controller room.

Condensation on the pneumotach changed caused flow measurement drift and led to errors in volume measurements. For acquisitions at multiple lung volumes, duration of scanning often exceeded an hour or more and it was initially found that appreciable condensation did indeed occur. Several methods were implemented to alleviate these problems. First, the pneumotach was heated to a level above body temperature (37 C) in order to raise the dew point and prevent condensation from occurring. This had the effect of maintaining measurement stability for a longer duration than previously, but still inadequate for the lengthy scan times necessary to acquire data sets at several lung volumes. Second, to further address the condensation problem, we tried several different filter media directly upstream of the pneumotach and finally settled upon utilization of a heat-and-moisture exchange (HME) filter. The addition of the HME filter greatly enhanced the measurement stability and system performance.

Another major development portion of this project involved controlling the hardware remotely from a personal computer (PC) in another room so as not to expose the operator to harmful x-ray unnecessarily. The operator monitored lung volume changes and controlled when scanning would occur from the CT controller room while subject, transducers and other hardware remained in the scanner room. Another aspect of this device permitted the subject to interrupt or pause scanning via a push button switch whose output was connected to the CT scanner and also the controlling PC. Depressing this switch simultaneously had the effect of opening the airway occlusion apparatus, pausing the scanning sequence, and resuming tracking of lung volume changes. Without this mechanism, the patient's alternative would be to remove the occlusion valve and invalidate imaging data acquired previously for that lung volume.

Validation

Validation of equipment performance was performed using three distinct tests: 1) volume measurements as compared vs. a calibrated super-syringe used to simulate respiratory maneuvers, 2) scanning of a subject at the same lung slice level fixed to a universal plane while breathholding at the same fixed lung volume before and after taking intervening breaths, and 3) volumetric imaging of subjects, each at 3 different lung volumes between residual volume (RV) and total lung volume (TLC).

Results

Simulation with Calibrated Super-syringe

Simulated breathing using a calibrated super-syringe with room air resulted in volume error of less than <2% for a delivered volume of 1, 2, and 3 liters. Tests were repeated with syringe injections performed at various delivery flow rates with similar results.

Single Slice Scan Before and After Intervening Breaths

Scanning a single lung slice at a fixed level of lung inflation before and after taking several intervening breaths permits testing that the equipment properly and repeatably returns the patient to a desired lung inflation level. Digital image subtraction revealed lung volumes before and after taking intervening breaths matched well with one another. Computation of regional percent air content (shown in Figure 2) in seven region of interests (ROIs) further indicated proper matching of lung volume before and after breathing. Mean percent air content error was found to be 0.41 + 0.56%. Table 1 shows the raw % air content data before and after taking intervening breaths for each of seven ROI regions.

Volumetric Imaging at Different Lung Volumes

Each subject was imaged at 3 different lung volumes spanning RV to TLC. Figure 3 showed good visual matching between anatomic structures across slices representing scanning breaks where intervening breaths were taken before resuming scanning.

Figure 3: Top row depicts volumetric lung images at three volumes: RV, 40% VC, TLC. Bottom row superimposes extracted airway tree upon the lung at each inflation level..

References:

1. Casale, T.B., R. Chiplunkar, S. Reed, M. Collins, and E. Hoffman, Methacholine-induced airway constriction in vivo mimics in vitro cholinergic innervation. Am. J. of Resp. and Crit. Care Med., Vol. 156 (4), A877, 1996.

2. Tajik, J.K., B. Tran, and E. Hoffman, Xenon enhanced CT imaging of local pulmonary ventilation. SPIE Medical Imaging Proceedings 1996, vol 2709, pp.40-254.