Multislice CT for trauma leaves much to be desired

By O. Clark West, MD
pdf path

Dr. West is an Associate Professor and the Chief of Emergency and Trauma Radiology, The University of Texas Medical School at Houston, TX. He is also a member of the Editorial Board of this journal.

We had the privilege and challenge of being an early adopter of multislice helical CT for use in our trauma center. In August 1999, Memorial Hermann Hospital, Houston, TX, installed a GE LightSpeed QX/i CT scanner (GE Healthcare, Waukesha, WI). This is the first version of the 4-channel multislice scanner. A model with the same name is currently on the market, but it is a radically different scanner than the first-gener-ation multislice scanner that we are still using.

Our initial 2 years with the scanner were spent learning how to use it optimally. In the beginning, image quality was substantially worse than that of our previous single-slice helical scanner. We had to unlearn the rules that apply to single-slice CT. Specifically, slice thickness and pitch are not the important parameters in understanding how to use multislice CT. Instead, detector configuration is of paramount importance. Once we began to think of scanning in the 4 × 1.25 mm, 4 × 2.5 mm, 4 × 3.75 mm, or 4 × 5 mm detector configurations, we began to make sense of our image quality issues. For example, we learned that the highest resolution detector configuration with a slow table speed resulted in outstanding quality images of the cervical spine. We learned that displaying 2.5-mm transverse images made with the bone algorithm was sufficient for detecting and evaluating most cervical spine fractures, but that we absolutely needed high-quality sagittal and coronal reformations. In order to get these, it was necessary to make a second set of axial images at 1.25 mm thickness using the standard algorithm that smoothes images and reduces noise. We gradually learned to optimize parameters for all body parts in a top-of-the-head to bottom-of-the pelvis approach to imaging the multitrauma patient.

Technical advances in multislice CT technology have focused on hardware improvements. Sixteen-channel detectors are coming into widespread use and 40- and 64-row devices are due on the market within the next few months. The X-ray tube rotates much faster now with rotation times of 400 to 500 msec compared with the 800 msec on our current scanner. These technical advancements offer numerous possibilities, including rapid scanning from the top of the head to the bottom of the pelvis in the 16 × 1.25 mm configuration and creating images optimized for multiple purposes, including screening for thoracic and lumbar spinal trauma, inspecting transverse images for parenchymal organ injuries in the abdomen, and creating angiograms of the thoracic aorta. These technical advances are truly exciting and will improve our ability to conduct rapid, efficient, and accurate head, neck, and torso examinations.

While faster hardware and more data channels are desirable improvements, many factors continue to limit the efficiency of CT in the multitrauma patient. These include the absence of an integrated patient monitoring system in the scanner gantry; with such a system, electrical cables, which are a source of streak artifacts that seriously degrade our image quality, would not need to be imaged with the patient.

Second, to achieve optimal image quality, we are in the habit of scanning the head and neck with the arms at the patient's sides and then stopping and repositioning the arms above the patient's head for imaging the torso. This repositioning is time-consuming and potentially risky, since intravenous catheters tend to be displaced during this maneuver and we may be moving undetected upper extremity injuries. Advancement in reconstruction technology that would eliminate streak artifacts created by the arms at the side is a much needed improvement.

Third, our most serious limitation in fully utilizing the rich data from our 4-slice multidetector scanner is the time it takes for the technologist to create sagittal and coronal reformations of the spine, CT angiograms, three-dimensional images of the pelvis, etc. Axial images appear on our picture archiving and communication system (PACS) very promptly, ie, sufficiently fast that we can provide online interpretation of these images and do not need to monitor the examination from the CT control room. However, we can wait 20 minutes to 2 hours to view sagittal and coronal reformations of the cervical spine, which are required for interpretation. This problem represents a major deficiency in the current generation of multislice scanners, one that really has not been addressed well in the latest releases. We absolutely need automation that allows the technologist to prescribe the sagittal and coronal reformatted images as part of the initial examination setup. Since we need it the same way every time, the computer should do it for us automatically. This development would truly revolutionize imaging of the multi-trauma patient, substantially reducing technologists' workload and would allow less skilled technologists to successfully utilize relatively complex imaging protocols. With the shortage of technologists that many areas of the country experience and the difficulty in finding highly skilled CT technologists to work off-hour shifts, a more automated scanner control software package is a high priority in selecting a new system for our trauma center. I am much less concerned about whether I have 16 or 64 detector channels than I am about having reformatted images of the spine and aorta created automatically.

The major CT vendors have been slow to respond to the specific needs of the trauma/emergency radiology patient. Multislice CT is the major diagnostic tool in the trauma and emergency setting and requirements in this environment must be specifically considered. To my knowledge, no vendor offers a "multitrauma package" for their scanners, which automatically creates sagittal and coronal reformatted images and CT angiograms without the need for extra technologist interaction. If emergency and trauma radiologists start requesting such specifications for CT that perform some or all of the functions described above, we are likely to see positive developments along these lines.

When requests were made for decreased radiation exposures for CT of pediatric patients, CT manufacturers responded promptly. While the emergency and trauma community is not as well organized nor as large as that of pediatric radiology, we may obtain similar results by applying consistent pressure from multiple sites. Please join me in promoting development of a "multitrauma - critical care package" for multislice CT.

Back To Top

Multislice CT for trauma leaves much to be desired.  Appl Radiol. 

July 09, 2004

Copyright © Anderson Publishing 2019