I read Greg Freiherr's column "Why the NEJM study on CAD is wrong" on your website. Please always keep in mind the following about computer-aided detection:
I read Greg Freiherr's column "Why the NEJM study on CAD is wrong" on your website. Please always keep in mind the following about computer-aided detection:
The potential value (i.e., cancer detection rate and tumor size at detection) of CAD will always be inversely proportional to the performance (i.e., cancer detection rate and tumor size at detection) of the radiologist.
Current cost for screening mammography cancer detection (using Medicare reimbursement rates for screening mammography only and cancer detection rate of four cancers/1000 patients) is $86 x 1000/4 = $21,500 per cancer detection.
If we wish to remain cost equivalent for the value of CAD for screening mammography cancer detection ($18/CAD examination), CAD will need to aid the radiologist in the detection of one additional cancer among 1194 patients (or result in a significant reduction in the average size of an invasive tumor for an equivalent number of cancer cases): $21,500/$18 = 1194 patients.
Several CAD reports within our current medical literature note that with experienced breast imagers, CAD helped identify one additional cancer per 4000 to 6000 patients. CAD cost/value ratio in this group would range: 4000 to 6000 x $18 = $72,000 to $108,000 per cancer detection.
Given my experience and research with CAD (most recent article, scheduled for publication in Oct./Nov. issue of Radiology, compared R2 and iCAD ability to detect ~220 invasive cancers without associated calcification measuring less than 16 mm), I believe CAD remains in its infancy but, with the correct research and development, could in the future potentially become the primary reader for screening mammography. Currently, CAD marketing and PR far exceed real clinical results for experienced breast imagers.
Richard L. Ellis, M.D.
Codirector, Norma J. Vinger
Center for Breast Care
La Crosse, WI
Can Emerging AI Software Offer Detection of CAD on CCTA on Par with Radiologists?
May 14th 2025In a study involving over 1,000 patients who had coronary computed tomography angiography (CCTA) exams, AI software demonstrated a 90 percent AUC for assessments of cases > CAD-RADS 3 and 4A and had a 98 percent NPV for obstructive coronary artery disease.
Contrast-Enhanced Mammography Study Reveals 24 Percent Lower Sensitivity with Moderate/Marked BPE
April 30th 2025In comparison to minimal or mild background parenchymal enhancement on contrast-enhanced mammography (CEM), researchers found that moderate or marked BPE was associated with a 12 percent lower AUC for breast cancer detection.
Could AI-Powered Abbreviated MRI Reinvent Detection for Structural Abnormalities of the Knee?
April 24th 2025Employing deep learning image reconstruction, parallel imaging and multi-slice acceleration in a sub-five-minute 3T knee MRI, researchers noted 100 percent sensitivity and 99 percent specificity for anterior cruciate ligament (ACL) tears.