The value of CAD in clinical trials
As vendors exhibit how CAD aids detection in clinical applications on the RSNA 2006 show room floor, a number of presentations are zeroing in on the impact CAD is having in the research arena. In particular, an informatics presentation hosted in the NCI caBIG Imaging Workspace area of the Lakeside Learning Center on Monday called “The Potential Value of CAD in Clinical trials” focused on the potential impact computer aided detection and diagnosis can have when used in clinical trials.
Imaging plays a key role in clinical trials, said Matthew Freedman, MD, associate professor of oncology and director of the division of advanced cancer imaging at the Imaging Science Research Center in Washington, D.C. “It helps to detect the occurrence or recurrence of disease, assure that any findings on an image is related to the disease under study and it also measures the size of the process,” explained Freedman.
The addition of computer aided detection and diagnosis to imaging in clinical trials helps to further strengthen the confidence levels of physicians when reading images to detect cancerous and non-cancerous lesions.
“This talk today is very much about limiting the variability of interpretation,” said Freedman. Variability in a clinical trial can also be caused by subject variability in genetic and environmental makeup as well as the variability in the initiation and the progression of disease.
The use of CAD in clinical settings has been shown to decrease such variability, said Freedman. At the same time, any decrease in variability within the experimental and control groups increases the likelihood of showing statistically significant results.
“You must decrease variability in all tasks related to image acquisition and interpretation,” said Freedman. “In a clinical trial, you want to get your groups at the lowest variable as possible. CAD can decrease the variability of interpretation and increase the ability to show a significant difference between your groups.”
Imaging plays a key role in clinical trials, said Matthew Freedman, MD, associate professor of oncology and director of the division of advanced cancer imaging at the Imaging Science Research Center in Washington, D.C. “It helps to detect the occurrence or recurrence of disease, assure that any findings on an image is related to the disease under study and it also measures the size of the process,” explained Freedman.
The addition of computer aided detection and diagnosis to imaging in clinical trials helps to further strengthen the confidence levels of physicians when reading images to detect cancerous and non-cancerous lesions.
“This talk today is very much about limiting the variability of interpretation,” said Freedman. Variability in a clinical trial can also be caused by subject variability in genetic and environmental makeup as well as the variability in the initiation and the progression of disease.
The use of CAD in clinical settings has been shown to decrease such variability, said Freedman. At the same time, any decrease in variability within the experimental and control groups increases the likelihood of showing statistically significant results.
“You must decrease variability in all tasks related to image acquisition and interpretation,” said Freedman. “In a clinical trial, you want to get your groups at the lowest variable as possible. CAD can decrease the variability of interpretation and increase the ability to show a significant difference between your groups.”