| CPC G16H 30/40 (2018.01) [G06T 7/0012 (2013.01); G06T 7/33 (2017.01); G06T 2207/10132 (2013.01); G06T 2207/30096 (2013.01)] | 8 Claims |
|
1. An ultrasonic diagnostic device, comprising:
a detector having a machine learned detection model, the detector being configured to detect, using the machine learned detection model, a lesion candidate in an ultrasonic image;
a database configured to contain features of lesion candidates to be excluded;
a determiner configured to determine whether the detected lesion candidate is an exclusion target, by collating a feature of the detected lesion candidate with the features registered in the database; and
a controller configured to display a mark for indicating the detected lesion candidate on the ultrasonic image in a case where the detected lesion candidate is not an exclusion target, the controller being configured to restrict display of the mark in a case where the determiner determines that the detected lesion candidate is an exclusion target;
an inputter configured to receive an exclusion instruction while the lesion candidate and the mark are displayed on the ultrasonic image; and
a register configured to register, in the database, a feature of a lesion candidate addressed by the exclusion instruction, wherein
the detector has a function of obtaining the feature of the lesion candidate, and the registered feature of the lesion candidate addressed by the exclusion instruction is the feature obtained by the detector, and
lesion candidates to be marked are correctable by the exclusion instruction without relearning of the machine learned detection model.
|