Feature | August 28, 2013 | Dianna M. E. Bardo, M.D.

Managing Pediatric Imaging and Dose

Producing high-quality diagnostic images with radiation dose as low as reasonably achievable

The issue of ionizing radiation in medical imaging is often discussed and is a hotly debated topic. Computed tomography (CT) is the major contributor to medical radiation dose exposure and has been vilified in lay and professional press as a danger to those exposed, potentially carcinogenic and most dangerous in children. As a pediatric radiologist, my primary concern is in producing high-quality diagnostic images with radiation dose as low as reasonably achievable (ALARA).

State-of-the-Art Imaging

The care and thought that precede any medical imaging examination of a child is extraordinary. The medical team that cares for each child includes not only the pediatrician and radiologist, but the manufacturer of medical imaging equipment as well. At Oregon Health & Science University (OHSU), Philips Healthcare is the maker of our CT scanners. Working in partnership with Philips, OHSU radiologists are able to utilize, test and influence the newest scanning, image reconstruction and image post-processing techniques. In this role, we are now working with a new knowledge- and model-based iterative image reconstruction technique, iterative model reconstruction (IMR). The knowledge-based iterative reconstruction algorithms of IMR differ from filtered back projection (FBP) methods in that the reconstruction is an iterative optimization process that uses the data statistics, image statistics and system models. The constrained optimization processes gives the user some amount of control over the desired image characteristics of the IMR images. CT images in our clinical patients are now undergoing an additional reconstruction step as we evaluate IMR. Subspecialty trained and board-certified radiologists are reviewing and evaluating CT images for quality, including assessment of image noise, diagnostic quality and visual appeal.

Often, CT images utilizing iterative reconstruction techniques are criticized for having a “plastic” or otherwise artificial look. In fact, one of our residents asked me how I knew iterative reconstructed images did not have “artifacts in them, compared to the real CT images” we are used to viewing. I guess it is all in what you are used to viewing, and we must remember that what we see in standard filtered back projection CT images is merely a representation of the patient’s anatomy.

At OHSU, we have been using Philips’ anatomic-based iterative reconstruction technique iDose4 since 2010. Since its installation, we have reduced CT dose by approximately 60 percent in all pediatric body examinations, relying on iDose4 to reduce image noise for these low-dose examinations. Our initial thoughts about using IMR were that we were happy with our level of dose reduction in pediatric exams, but we were looking to use IMR only for improving image quality. Having reviewed thousands of images now with IMR reconstruction, I am impressed by the quality of the images. The markedly reduced, or lack of, image noise in IMR CT images results in smoothness in visceral structures and muscle that mimics what is seen in natural, living tissues. With such remarkably improved image quality, our doses may just get a bit lower. 

Diagnoses With Confidence

Because IMR CT images more closely represent natural tissues, we have the opportunity to more accurately diagnose disease and be more confident in that diagnosis. Clinical CT images illustrate these concepts.

First is a 6-month-old infant with congenital heart disease. A CT scan was requested in order to visualize the anatomy of his coronary arteries, as an anomalous vessel was suspected on echo and would affect the surgical decisions for repairing his heart. A prospective ECG-gated (step-and-shoot) protocol designed to capture a motionless image of the coronary arteries was used. While the original images were diagnostic, the virtually noiseless IMR reconstruction elegantly shows only a single coronary artery (instead of the usual two), arising from the non-coronary cusp coursing over the right ventricular outflow tract (Figure 1). Reduced image noise of IMR also improves the quality of the 3-D surface rendered images. 

Low contrast tissue resolution is also optimized with IMR.  This effect is seen in a second child, a 6-year-old with a mass in the right lobe of the liver. Accounting for image noise, the mass, an embryonal sarcoma of liver, appears to be a homogenous texture on the FBP images (arrows).  The IMR images clearly show that the periphery of the mass is encapsulated and the internal detail of the varied tissue and architecture of the tumor because of improved low contrast resolution and marked reduction in image noise (arrows, Figure 2).

Image noise profoundly affects image quality in all patient populations, particularly in obese patients. IMR drastically reduces image noise even in the most noise degraded exams, improving diagnostic confidence. In a morbidly obese young man with chest pain, the noise standard FBP images limit my ability to thoroughly evaluate a complex plaque in the proximal left anterior descending artery (LAD). With IMR I can see that the plaque is composed of fibrous, fatty and calcified components, and that the degree of stenosis is severe (Figure 3).

The time required to reconstruct image data is important as well. IMR image reconstruction occurs within a matter of minutes — approximately two to three minutes for a chest, abdomen and pelvis scan — so images are readily available for diagnostic review. Workflow at the CT scanner is not affected by delayed or extended image reconstruction times, so patient care and throughput remains optimized.

The ground-breaking image quality and noise reduction of IMR iterative reconstruction fits our radiology practice at OHSU by maintaining efficient workflow, with improving patient care, and will likely lead to further reduction in CT radiation dose in children and adults.  itn

Dianna M. E. Bardo, M.D., is associate professor of radiology, pediatrics and cardiovascular medicine, director of cardiac radiology, pediatric neuroradiologist, at Oregon Health & Science University in Portland, Ore.

Related Content

Median Technologies and the Nice University Hospital to Use AI in Lung Cancer Screening
News | Artificial Intelligence | March 15, 2018
March 15, 2018 – Median Technologies, the industry-leading Imaging Phenomics Company and the Nice University Hospital
Siemens Healthineers Launches New Partners for Digital Ecosystem at HIMSS 2018
News | PACS Accessories | March 12, 2018
At the 2018 Healthcare Information and Management Systems Society (HIMSS) Annual Conference and Exhibition, March 5-9...
Imaging agent helps predict success of lung cancer therapy
News | Oncology Diagnostics | March 08, 2018
March 8, 2018 – Doctors contemplating the best therapy for...
Designed to optimize workflow, IntelliSpace Portal supports consistency across applications.
Sponsored Content | Case Study | Imaging | March 06, 2018
As one of the largest Catholic, not-for-profit healthcare systems in the United States, St. Louis-based SSM Health has...
TriHealth in Cincinnati

TriHealth in Cincinnati.

Sponsored Content | Case Study | Contrast Media Injectors | March 06, 2018
The continuing search for advantages to improve workflow has radiology departments constantly searching for new...
Carestream Supplies Wireless Digital X-ray Detectors for 2018 NFL Combine
News | Digital Radiography (DR) | February 27, 2018
In addition to demonstrating their speed, agility and strength at the 2018 National Football League (NFL) Scouting...
Mirada Medical Releases DLCExpert for Radiotherapy Treatment Planning
Technology | Treatment Planning | February 22, 2018
February 22, 2018 — U.K.-based medical imaging software provider Mirada Medical has released DLCExpert, the first com
ECR 2018 Spotlights Artificial Intelligence and Current Radiology Trends
News | Interventional Radiology | February 21, 2018
The European Congress of Radiology (ECR) announced the theme of its 2018 annual meeting, Feb. 28-March 4 in Vienna...
Aidoc Introduces AI Solution for Whole-Body CT Scan Analysis
News | Artificial Intelligence | February 20, 2018
Deep learning startup company Aidoc announced what it calls the world’s first and only comprehensive, full-body...
Overlay Init