| Summary: | This paper presents a machine learning based approach for the discrimination of malignant and benign microcalcification (MC) clusters in digital mammograms. A series of morphological operations was carried out to facilitate the feature extraction from segmented microcalcification. A combination of morphological, texture, and distribution features from individual MC components and MC clusters were extracted and a correlation-based feature selection technique was used. The clinical relevance of the selected features is discussed. The proposed method was evaluated using three different databases: Optimam Mammography Image Database (OMI-DB), Digital Database for Screening Mammography (DDSM), and Mammographic Image Analysis Society (MIAS) database. The best classification accuracy (<inline-formula> <math display="inline"> <semantics> <mrow> <mn>95.00</mn> <mspace width="0.166667em"></mspace> <mo>±</mo> <mspace width="0.166667em"></mspace> <mn>0.57</mn> </mrow> </semantics> </math> </inline-formula>%) was achieved for OPTIMAM using a stack generalization classifier with 10-fold cross validation obtaining an A<inline-formula> <math display="inline"> <semantics> <msub> <mrow></mrow> <mi>z</mi> </msub> </semantics> </math> </inline-formula> value equal to <inline-formula> <math display="inline"> <semantics> <mrow> <mn>0.97</mn> <mspace width="0.166667em"></mspace> <mo>±</mo> <mspace width="0.166667em"></mspace> <mn>0.01</mn> </mrow> </semantics> </math> </inline-formula>.
|
|---|
