公开(公告)号：US20190298159A1

公开(公告)日：2019-10-03

申请号：US16420492

申请日：2019-05-23

Applicant: OLYMPUS CORPORATION

Inventor： Mitsutaka KIMURA ,  Takehito HAYAMI ,  Takashi KONO ,  Yamato KANDA ,  Ryoji TAKAMI

IPC: A61B1/04 ,  A61B5/06 ,  A61B8/08 ,  A61B6/00 ,  G06T7/00

Abstract: An image processing device includes a processor including hardware, the processor being configured to execute: estimating an operation state of a medical apparatus in a lumen of a subject based on information sequentially input from the medical apparatus, the medical apparatus including an imaging device, and an insertion portion inserted into the lumen, and the information including at least an intraluminal image obtained by the imaging device capturing an image inside the lumen; and performing processing corresponding to the operation state of the medical apparatus in the lumen based on an estimation result.

公开(公告)号：US20190311476A1

公开(公告)日：2019-10-10

申请号：US16433483

申请日：2019-06-06

Applicant: OLYMPUS CORPORATION

Inventor： Takehito HAYAMI ,  Yamato KANDA ,  Takashi KONO ,  Mitsutaka KIMURA ,  Ryoji TAKAMI

IPC: G06T7/00 ,  G06T7/20 ,  G06F3/01 ,  G06T7/60 ,  G06T7/11 ,  H04N5/225 ,  G06T7/70 ,  G06K9/62 ,  A61B1/00 ,  A61B5/00 ,  A61B1/06 ,  A61B1/31

Abstract: An image processing apparatus have a processor configured to: receive a plurality of images of a tissue including a lesion portion, the plurality of images being received from an endoscope; judge, based on a signal from the endoscope, whether or not to engage a differentiation classification operation; in response to judging that the differentiation classification operation is engaged, process the plurality of images to specify the lesion portion within one or more of the plurality of images; and perform differentiation classification on the one or more images to classify the lesion portion into at least one class of a plurality of classes; and in response to judging that the differentiation classification operation is not engaged, not process the plurality of images to specify the lesion portion within the one or more of the plurality of images.