Method, computer program product, and system for the automated analysis of lesions in magnetic resonance, mammogram and ultrasound images

A method for the analysis of a lesion in an anatomy, comprising:

• obtaining plural image data, representative of plural images of a same portion of the anatomy, derived from at least one imaging modality selected from the group consisting of magnetic resonance imaging, x-ray imaging, and ultrasound imaging, said plural images derived at different times during introduction of a contrast agent into said anatomy;
• identifying from the image data a possible lesion in said plural images;
• extracting at least one feature related to inhomogeneity of uptake of a lesion from said plural image data corresponding to the identified possible lesion ;
• and
• characterizing said possible lesion based at least in part on the at least one feature extracted from the plural images derived from said at least one imaging modality.

A system for the analysis of a lesion in an anatomy, comprising:

• means for obtaining plural image data, representative of plural images of a same portion of the anatomy, derived from at least two imaging modalities selected from the group consisting of magnetic resonance imaging, x-ray imaging, and ultrasound imaging;
• means for identifying from the image data a possible lesion in said plural images;
• means for extracting, for each of said plural images derived from said at least two imaging modalities, at least one feature related to characterization of a lesion from said plural image data corresponding to the identified possible lesion;
• and
• means for merging in a common image classifier a plurality of extracted features, including at least one feature extracted from each of the plural images derived from said at least two imaging modalities to characterize said possible lesion based on the merged extracted features and yield a corresponding classification.

A system for the analysis of a lesion in an anatomy, comprising:

• means for obtaining plural image data, representative of plural images of a same portion of the anatomy, derived from at least one imaging modality selected from the group consisting of magnetic resonance imaging, x-ray imaging and ultrasound imaging, said plural images derived at different times during introduction of a contrast agent into said anatomy;
• means for identifying from the image data a possible lesion in said plural images;
• means for extracting at least one feature related to inhomogeneity of uptake of a lesion from said image data corresponding to the identified possible lesion;
• and means for characterizing said possible lesion based at least in part on the at least one feature extracted from the plural images derived from said at least one imaging modality.

A method for the analysis of a lesion in an anatomy, comprising:

• obtaining plural image data, representative of plural images of a same portion of the anatomy, derived from at least one imaging modality selected from the group consisting of magnetic resonance imaging, x-ray imaging, and ultrasound imaging;
• identifying from the image data a possible lesion in said plural images;
• extracting a radial gradient peak distinction and a variance of a gradient along a margin of said possible lesion;
• and
• characterizing said possible lesion based at least in part on said radial gradient peak distinction and said variance of a gradient along a margin extracted from the plural images derived from said at least one imaging modality.