Computer-aided diagnosis in breast tomosynthesis and chest CT

Area of Science:

• Medical Imaging
• Computer-Aided Diagnosis
• Radiology

Background:

• Computer-aided diagnosis (CAD) aids in detecting and characterizing abnormalities in medical images.
• CAD has proven effective for breast cancer detection in mammography.
• Research is expanding CAD applications to digital breast tomosynthesis (DBT) and chest computed tomography (CT).

Purpose of the Study:

• To review recent advancements in CAD for DBT and chest CT.
• To discuss the development of image segmentation and feature extraction techniques for mass detection in DBT.
• To present CAD methods for lung nodule detection and characterization in chest CT scans.

Main Methods:

• Developing and comparing CAD systems using 2D, 3D, and combined approaches for DBT.
• Implementing prescreening, segmentation, and feature extraction techniques for lung nodule detection in CT.
• Conducting observer ROC studies to evaluate CAD's impact on radiologist accuracy.

Main Results:

• DBT CAD development presents unique challenges and opportunities due to its quasi-3D nature.
• Chest CT CAD methods show promise for improved lung nodule detection and characterization.
• CAD systems demonstrated potential to enhance radiologists' accuracy in detecting abnormalities.

Conclusions:

• CAD development for 3D modalities like DBT and CT shares common approaches.
• Experiences in DBT and CT CAD can facilitate development for other 3D imaging applications.
• Continued research in CAD is crucial for advancing medical image analysis and diagnosis.