Perfusion deficit versus anatomic visualization in detection of pulmonary emboli
via electron-beam CT: validation in swine
E. A. Hoffman, J. K Tajik, G. Petersen, T. J. Reiners, B.H. Thompson, W. Stanford
Departments of Radiology and Physiology, Univ. of Iowa College of Medicine, Iowa City, IA 52242 USA
Recently, volumetric X-ray CT scanning occurring during i.v. contrast drip has been reported to be a promising technique in detecting pulmonary emboli by allowing visualization of the clot as an unenhanced portion of the reconstructed vascular lumen. To determine whether functional imaging via the dynamic mode of electron beam CT (Imatron) might aid in diagnosing pulmonary emboli by detecting pulmonary perfusion deficits, we have scanned 32 pigs before and after autologous clot injection. Pigs were anesthetized with a combination of pentobarbital, inovar, and succinylcholine, intubated and mechanically ventilated. Catheters were placed in the inferior vena cava for drug delivery, contrast infusion, and clot delivery. 5cc of blood were drawn and allowed to coagulate for subsequent re-injection into the inferior vena cava near the right ventricle. The clot was divided into two portions and each half was tagged with an embedded length of suture material. Three scans were performed both before and after injection of one of the clots (the most stable one):
- 1. 60 3mm contiguous slices during slow infusion of contrast with the ventilator on during scanning.
- 2. 40 3mm contiguous slices (100 msec scan aperture), no contrast infusion and cephalad-caudal extent of scanning selected to cover the carina to the base of the lung, respiration suspended at FRC.
- 3. Functional imaging (8 8mm slices, 10 time points each) scanning each slice at multiple time points during bolus contrast injection to sample regional parenchymal "time-intensity" curves. Contrast was injected into the right ventricle at 10cc/sec over 2 sec and scanning commenced one heart beat prior to contrast injection. Images were acquired during a breath hold at FRC and scanning was gated to the ECG.
Following CT scanning, the pigs were overdosed with pentobarbital, frozen and evaluated by pathology for clot location (based upon location of suture imbedded in injected clot).
Image data from scan 3 were used to evaluate and quantitate the distribution of pulmonary perfusion in the pig before and after clot injection. Scan 2 allowed us to link the perfusion data with high resolution volumetric image data sets of the anatomic distribution of vessels feeding the region of flow deficit and thereby to determine the anatomic location of the embolus. Image data from scan 1 were reviewed by two radiologists in a double-blinded protocol.
Results: Functional deficits were present either: 1. before and after injection of the clot; or 2. only after injection of the clot. Deficits detected before injection of the clot correlated with areas of pneumonia diagnosed via visual inspection of the CT slices. Edges of the flow deficit were sharply defined when the flow deficit was due to embolus and poorly defined when due to pneumonia. Pathology findings and functional image based assessment of emboli agreed in 13 out of 14 studies. In one case, a clot was located by pathology which was not found by functional imaging because scanning did not cover the region of the embolus. Interestingly, in addition to locating the position of the primary clot found by pathology, functional imaging identified an additional 8 regions of perfusion deficits, presumably caused by clot fragmentation. Double-blinded review of images from scan 1 identified 5 of the clots spotted by functional imaging and missed 17. 7 of 17 missed clots were also missed by the pathologic evaluation and thus represented clot fragments other than the fragment containing the suture. In one of the 32 studies, the post clot functional scan was not performed because of premature death of the pig. 17 (of the 32) pigs scanned more recently have yet to be evaluated.
Conclusions: Functional imaging via dynamic CT to date appears to provide a significant improvement in aiding detection of pulmonary emboli. The amount of contrast required for such a study is significantly less than used during a continuous drip and thin slice, volumetric scanning. Combining a non-contrast volumetric scan with a functional scan, provides a method to link structure and function data using a single scanner.
