Abstract | ||
---|---|---|
In computer-aided diagnosis (CADx), displaying a set of lesions similar to the one being evaluated has the potential to improve radiologists' diagnostic accuracy. In this work, we investigate whether an automated retrieval CADx system can effectively assist radiologists in diagnosis of clustered microcalcifications (MCs). We first develop a retrieval system for relevant cases by taking into account both perceptually similar image features and the likelihood of malignancy of the lesion under consideration. We then conduct an observer study with a group of 12 breast radiologists to evaluate the diagnostic value of the proposed retrieval system on a set of 100 test cases (50 malignant, 50 benign). Based on receiver-operating characteristic analysis, the results demonstrate that the proposed retrieval CADx system can significantly improve the readers' diagnostic accuracy of MC lesions in terms of both likelihood of malignancy and BI-RAD assessment. |
Year | DOI | Venue |
---|---|---|
2016 | 10.1109/ISBI.2016.7493452 | 2016 IEEE 13TH INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING (ISBI) |
Keywords | Field | DocType |
Clustered microcalcifications, computer-aided diagnosis, and content-based image retrieval | Computer vision,Aided diagnosis,Pattern recognition,Computer science,Feature (computer vision),Computer-aided diagnosis,Image retrieval,Malignancy,Test case,Artificial intelligence,Content-based image retrieval | Conference |
ISSN | Citations | PageRank |
1945-7928 | 0 | 0.34 |
References | Authors | |
6 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Juan Wang | 1 | 109 | 27.00 |
Yongyi Yang | 2 | 1409 | 140.74 |
Miles N. Wernick | 3 | 595 | 61.13 |
Robert M Nishikawa | 4 | 599 | 58.25 |