Accuracy Measures of 3D Automated Breast Ultrasound in Detection of Breast Lesions in Comparison to 2D Ultrasound and Digital Mammography

Background: Breast screening imaging tools have anim-portant role in prevention as well as early detection of breast cancer. Despite that mammography is the first screening tool, its sensitivity decreased by higher breast density resulting in increasing breast cancer risk by 15-25%. In addition that hand-held ultrasound screening can determine early-stage cancer breast than mammographic screening alone, however screening with HHUS alone is time consuming and has a high false pos-itive results. Automated breast ultrasound is one of the latest break throughs in detecting breast lesions and as well high res-olution images production. Aim of Study: To study better accuracy measures of ABUS in diagnosis and characterization of different breast lesions than HHUS and digital mammography. Patients and Methods: This prospective study was per-formed in the Female Imaging Unit, Radiology Department, Faculty of Medicine, Cairo University and approved by its Research and Ethical committee; all the included cases gave informed consent. The study was conducted on 42 patients presented from surgery clinic and early screening clinic with either: Annual screening or Breast lump. Was done between January 2022 and January 2023. Their ages ranged from 27 to 66 years (mean age: 48.48±11.526 SD years). All of the cases (n=42) will be examined with mammography and sonography, and further analyzed with ABUS. Results: Digital mammography accuracy measures in de-tecting breast lesions were 86.7% sensitivity, specificity 88.9%, 80% NPV, 92.9% PPV, and accuracy of 87.5%. Mammography was under estimated in about 13.3% of the lesions in the breast and overestimated in 11.1% of breast lesions while Handheld ultrasound sensitivity in detecting breast lesions was about 93.3%, 94.4% specificity, NPV 89.5%, 96.6% PPV, and 93.7% accuracy. Handheld ultrasound was under estimated in 6.6% of breast masses and overestimated in 5.5% of breast lesions. As well as the specificity of Automated breast ultrasound was 88.9%, sensitivity was 96.7%, 93.5% PPV, 94.1% NPV and 93.7% accuracy. Automated ultrasound was over estimated in about 11.1% of cases and under estimated in 1.1% of cases. Conclusion: Automated breast ultrasound is a promising imaging modality that can be used in addition to mammography and instead of HHUS, in screening as well as early breast can-cer detection, to reduce the patient radiation exposure eliminate human factor, save scan time and provide 3D images.