CAD acquisition parameters prove critical to accuracy

Article

Subtle changes to imaging acquisition parameters can dramatically affect the accuracy of computer-aided detection systems, according to research conducted at the University of Maryland and the Baltimore VA Medical Center.

Subtle changes to imaging acquisition parameters can dramatically affect the accuracy of computer-aided detection systems, according to research conducted at the University of Maryland and the Baltimore VA Medical Center.

While studies evaluating the effect of CAD on image interpretation continue to be presented at the RSNA meeting every year, surprisingly little attention has been accorded the image acquisition parameters themselves, said Dr. Eliot Siegel, chief of radiology and nuclear medicine at the University of Maryland/Baltimore VA Medical Center.

Siegel and colleagues examined the effect that two reconstruction kernels, soft tissue and bone, would have on lung nodule detection. They found trade-offs in sharpness and image quality between the two kernels. The soft-tissue kernel provides a smoother image with increased contrast and decreased noise, but poor edge detection. The bone kernel provides improved spatial resolution and better edge detection, but increased noise.

Siegel's group tested the two reconstruction parameters on 21 chest CT exams at 120 kVp, 12 mAs, and 0.75 collimation. After acquisition, images were reconstructed using either the B40f (soft tissue) or the B60f (bone) kernel.

Nodules were defined as noncalcified with a size greater than or equal to 4 mm. Truth was established by a consensus of thoracic radiologists.

The investigators detected a total of 34 lung nodules between 3 mm and 3 cm in size. While both kernels had the same 77% sensitivity, the bone kernel had significantly fewer (2.4) false positives compared with the soft-tissue kernel (2.8).

"The superior performance of the bone kernel suggests that it should be the preferred technique for research and perhaps even clinical studies," Siegel said.

Dr. Bruce Reiner, director of radiology research at the University of Maryland/Baltimore VA, presented information from a sister study on collimation parameters. That study examined the same 21 chest CT exams and used the same acquisition parameters as the reconstruction kernel study, but collimation was set at 0.75 mm, 1.5 mm, and 3 mm.

The highest sensitivity and specificity were seen at the 0.75-mm collimation slice thickness and the lowest at 3-mm collimation-an unexpected finding, Reiner said.

The number of false positives increased with an increase in slice thickness, which can lead to higher costs, increased morbidity, and decreased productivity, he said.

"The accuracy of CAD is dependent on the CT reconstructed slice thickness, and it is best to use narrow collimation and thin reconstruction," Reiner said.

Recent Videos
Combining Advances in Computed Tomography Angiography with AI to Enhance Preventive Care
Study: MRI-Based AI Enhances Detection of Seminal Vesicle Invasion in Prostate Cancer
What New Research Reveals About the Impact of AI and DBT Screening: An Interview with Manisha Bahl, MD
Can AI Assessment of Longitudinal MRI Scans Improve Prediction for Pediatric Glioma Recurrence?
A Closer Look at MRI-Guided Adaptive Radiotherapy for Monitoring and Treating Glioblastomas
Incorporating CT Colonography into Radiology Practice
What New Research Reveals About Computed Tomography and Radiation-Induced Cancer Risk
What New Interventional Radiology Research Reveals About Treatment for Breast Cancer Liver Metastases
New Mammography Studies Assess Image-Based AI Risk Models and Breast Arterial Calcification Detection
Can Deep Learning Provide a CT-Less Alternative for Attenuation Compensation with SPECT MPI?
Related Content
© 2025 MJH Life Sciences

All rights reserved.