Health Imaging & IT | July 1, 2008 | Special Section: CAD’s Impact on Radiology
By Jonathan Batchelor
Fueled by clinical studies, integration with IT and reimbursement
The integration of computer-aided detection (CAD) software into the day-to-day practice of radiology has witnessed the enthusiastic adoption of the applications for some clinical indications. Mammography CAD, which was first greeted with skepticism and suspicion by radiologists, has since enjoyed a groundswell of support. Lung CAD, thanks to diligent efforts by early adopters, is showing signs that it may be the next area of medicine to embrace widespread utilization of this technology.
According to the American Cancer Society, lung cancer is difficult to treat because it is rarely caught in its earliest stages—when treatment options are most effective. As such, only 16 percent of patients diagnosed with lung cancer have a 50 percent chance of living beyond five years.
Given those statistics, a CAD application that can improve the diagnostic certainty of a primary evaluation for lung cancer and automatically track changes in nodule size during follow-up studies should be warmly welcomed by the clinical community. Although the technology was approved by the FDA in 2001 for chest radiography, and in 2004 for CT, it has not yet achieved the support and infiltration mammography CAD enjoys.
However, an ever-growing bibliography of peer-reviewed scientific journal articles demonstrating its efficacy, the integration of applications into enterprise healthcare systems, and reimbursement for its use by more payors signals that the technology may at last be poised for mainstream adoption.
Radiography CAD Gets Results
A recent prospective study published in Academic Radiology (May 2008) conducted by Edwin J.R. van Beek, MD, PhD, and colleagues in the department of radiology at the Carver College of Medicine, University of Iowa in Iowa City, Iowa, provides strong support for lung CAD utilization.
Chest radiography was performed using a Siemens Medical Solutions DR unit on 324 patients for surveillance of metastatic disease with known malignancy. Their exams were interpreted by experienced pulmonary radiologists via the PACS at the facility, and utilized IQQA®-Chest CAD software from EDDA Technology.
The study evaluated the sensitivity and specificity of the radiologists’ diagnostic interpretations with and without the use of CAD. For studies without the assistance of CAD, the radiologists’ diagnostic performance showed a sensitivity of 63.6 percent and a specificity of 98.1 percent. However, with the addition of CAD, the radiologists demonstrated a specificity of 92.7 percent and a specificity of 96.2 percent.
“This benefit [lung CAD utilization] might have important implications, related to the identification of patients for whom therapy is failing or in the identification of early metastatic disease where different treatment regimens are still available,” the authors wrote. “Both observations should result in important management decisions, thus improving overall patient care.”