Aims and objectives
♦ Computer-aided diagnosis (CAD) system in breast ultrasound
Plays a role in interpreting lesions that are found by the examiner.
♦ S-detectTM (Samsung Medison Co.
Ltd.,
Seoul,
Korea)
Morphologic analysis according to the US BI-RADS lexicons and final assessment.
Useful for less experienced radiologists.
No report about multi-reader analysis for same lesions.
→ The purpose of our study is to investigate the reliability of CAD system for ultrasound using breast phantom with multi-reader analysis for identical lesions.
Methods and materials
♦ From March 2016 to February 2017
♦US (Samsung Ultrasound RS80A),
including the CAD system: S-detectTM (Samsung Medison Co.
Ltd.,
Seoul,
Korea)
♦ Breast phantom: Customized Breast Ultrasound Examination Phantom
“Breast FAN” (Kyoto Kagaku Co.
Ltd.).
14 lesions ; 5 suspicious lesion,
6 benign lesions,
3 axillary lymph nodes.
♦ 6 readers (3 senior and 3 junior residents in radiology department)
; 3 senior residents
(3rd grade,
more than 1month training for breast US)
; 3 junior residents
(1st grade without breast US training)
♦The...
Results
♦ Table 1 shows the inter-and intra-reader reliability of breast lesion based on modalities and reader’s experiences.
The kappa values of final assessment on US is higher than those on CAD.
In the subjective combined conclusion,
the kappa value is improved on Junior group.
The kappa values of final decision on senior and junior group are more variable on CAD than US.
Espicially,
in the junior group,
kappa value of the 1st line study of the CAD was 0.694,
and that of second line study...
Conclusion
Better agreement of lexicons and final assessment in US than CAD.
The diagnostic performance of CAD in senior group was better than that of junior group like in US.
The combination US with CAD improves the reliability and diagnostic performance in junior group (starter,
beginner).
Several studies have shown that CAD is useful for inexperienced radiologists.
But,
CAD results of junior group (beginner or starter) was variable and inconsistent in this study.
Therefore,
minimum training and experience for breast US is indispensable for using breast...
References
ADDIN EN.REFLIST 1.
Wang Y,
Wang H,
Guo Y,
Ning C,
Liu B,
Cheng HD,
et al.
Novel computer-aided diagnosis algorithms on ultrasound image: effects on solid breast masses discrimination.
Journal of digital imaging.
2010;23(5):581-91.
2.
Wang X,
Guo Y,
Wang Y.
Automatic detection of regions of interest in breast ultrasound images based on local phase information.
Bio-medical materials and engineering.
2015;26 Suppl 1:S1265-73.
3.
Yang HC,
Chang CH,
Huang SW,
Chou YH,
Li PC.
Correlations among acoustic,
texture and morphological features for breast ultrasound...