Abstract:
The present invention relates to an apparatus for lesion segmentation in a medical image, which includes a first segmentation unit which obtains one or more first lesion segmentation results from a medical image using a predetermined first segmentation parameter with respect to at least one candidate lesion; and a second segmentation unit which determines at least one target lesion to segment among the candidate lesions based on the first lesion segmentation results, and obtains one or more second lesion segmentation results using a second segmentation parameter with respect to the determined target lesion to segment.
Abstract:
PURPOSE: A device for detecting errors of a lesion boundary, a method thereof, a device for modifying the errors, and a method thereof, and a device for checking the errors are provided to detect the errors of the lesion boundary by using an energy value corresponding to the lesion boundary. CONSTITUTION: An extracting unit(110) extracts a boundary of lesion included in 2D image frames composing a 3D image. An error determining unit(140) determines an error of the boundary based on one or more of prediction information about the lesion included in the 2D image frames and an energy value corresponding to the boundary. [Reference numerals] (110) Extracting unit; (120) Predicting unit; (130) Calculating unit; (140) Error determining unit
Abstract:
PURPOSE: A lesion apparatus and a method are provided to make a diagnosis accurately by diagnosing a diagnostic object lesion in the state of excluding other lesions in the surrounding of the diagnostic object lesion or an organ which affects a diagnosis of the lesion. CONSTITUTION: A lesion diagnostic apparatus(300) comprises a region decision unit(310), an analysis unit(320), and a diagnostic unit(330). The region decision unit divides and analyzes a domain based on a lesion or an organ which is included in a medical image. The region decision unit comprises a domain division part(311), a domain exclusion part(312), and a surrounding region decision part(313). The domain exclusion part excludes the domain including other lesions except a diagnostic object lesion or an organ among the divided domains by the domain division part. The surrounding region decision part decides the domains except a diagnosis domain among other domains except the domains excluded by the domain exclusion part as a surrounding domain for a lesion diagnosis. [Reference numerals] (311) Domain division part; (312) Domain exclusion part; (313) Surrounding region decision part; (321) Rear region analysis part; (322) Left-right region analysis part; (323) Surrounding region correction part; (330) Diagnostic unit
Abstract:
PURPOSE: An apparatus for detecting a lesion, a method for detecting the lesion and a lesion diagnosis apparatus are provided to accurately detect the lesion. CONSTITUTION: An extracting unit(110) extracts at least one texture region from a breast image. The extracting unit removes noise from the breast image. A setting unit(120) sets at least one lesion detection region from an extracted texture region. A detection unit(130) detects a lesion from the lesion detection region. [Reference numerals] (100) Lesion detecting apparatus; (110) Extracting unit; (120) Setting unit; (130) Detection unit