Accepted for/Published in: JMIR Medical Informatics
Date Submitted: Sep 20, 2019
Open Peer Review Period: Sep 21, 2019 - Oct 16, 2019
Date Accepted: Jan 26, 2020
(closed for review but you can still tweet)
Diagnosis Performance of Computer-Aided Detection on Breast Lesions Less and More Than 2cm
ABSTRACT
Background:
Computer-aided Detection (CAD) is increasingly being used as an aid tool by radiologists on breast lesion diagnosis. Previous studies demonstrated that CAD can improve the diagnosis performance of radiologist. However, the optimal use of CAD in breast lesions less and more than 2cm has not been mentioned.
Objective:
To compare the performance of different radiologists by using CAD to detect breast cancer in size less and more than 2cm.
Methods:
We prospectively enrolled 261 consecutive patients (mean age, 43years; age range, 17–70 years), 398 lesions (148 more than 2cm, 79 with malignant, 69 with benign; 250 less or equal to 2cm, 71 with malignant, 179 with benign), with breast mass as prominent symptom. One novice radiologist with 1 year of US experience and one experienced radiologist with 5 years of US experience were assigned to read US images without CAD, and at a second reading combined with CAD. Diagnostic performances were compared by using analysis of variance.
Results:
For novice reader, the area under the receiver operating characteristic curve improved from 0.743 (95% confidence interval [CI]: 0.669, 0.817) at without-CAD mode to 0.884 (95% CI: 0.834, 0.933; P＜.001) at the combined-CAD mode in lesions less or equal to 2cm, for experienced reader those were from 0.838(95% CI :0.773,0.903) to 0.902(95% CI :0.862,0.942; P=.002). Those were from 0.806 to 0.797(novice reader) and 0.895 to 0.822 (experienced reader) in lesions more than 2cm. The sensitivity of novice and experienced reader in lesions less than 2cm improved from 61.97% and 73.23% at without-CAD mode to 90.14% and 97.18% at combined-CAD mode, respectively.
Conclusions:
Computer-aided detection (CAD) was helpful for novice and experienced reader to improve cancer detection at breast US in lesions less than 2cm. CAD was more sensitive for both novice and experienced reader. Clinical Trial: ChiCTR1800019649
Citation
Request queued. Please wait while the file is being generated. It may take some time.
Copyright
© The authors. All rights reserved. This is a privileged document currently under peer-review/community review (or an accepted/rejected manuscript). Authors have provided JMIR Publications with an exclusive license to publish this preprint on it's website for review and ahead-of-print citation purposes only. While the final peer-reviewed paper may be licensed under a cc-by license on publication, at this stage authors and publisher expressively prohibit redistribution of this draft paper other than for review purposes.