公开(公告)号：US11424021B2

公开(公告)日：2022-08-23

申请号：US16868742

申请日：2020-05-07

Applicant: National Taiwan University

Inventor： Wei-Chih Liao ,  Wei-Chung Wang ,  Kao-Lang Liu ,  Po-Ting Chen ,  Ting-Hui Wu ,  Holger Roth

IPC: G06K9/00 ,  G16H30/40 ,  G16H50/20 ,  G06T7/00 ,  G06N3/08 ,  G06T7/11

Abstract: Provided are a medical image analyzing system and a method thereof, which mainly crop a plurality of image patches from a processed image including a segmentation label corresponding to a location of an organ, train a deep learning model with the image patches to obtain prediction values, and plot a receiver operating characteristic curve to determine a threshold which determines whether the image patches are cancerous, thereby effectively improving the detection rate of cancer.

公开(公告)号：US12067717B2

公开(公告)日：2024-08-20

申请号：US17507948

申请日：2021-10-22

Applicant: National Taiwan University

Inventor： Wei-Chung Wang ,  Wei-Chih Liao ,  Kao-Lang Liu ,  Po-Ting Chen ,  Po-Chuan Wang ,  Ting-Hui Wu

IPC: G06T7/00 ,  A61B5/00 ,  G06N3/08 ,  G06T7/70 ,  G06T7/73 ,  G06V10/774 ,  G06V10/776 ,  G06V10/82 ,  G16H30/20

CPC classification number: G06T7/0012 ,  A61B5/425 ,  A61B5/4887 ,  A61B5/726 ,  A61B5/7267 ,  G06N3/08 ,  G06T7/70 ,  G06T7/73 ,  G06V10/774 ,  G06V10/776 ,  G06V10/82 ,  G16H30/20 ,  G06T2207/20016 ,  G06T2207/20064 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/30096 ,  G06T2207/30204 ,  G06V2201/031

Abstract: A medical image analyzing system and a medical image analyzing method are provided and include inputting at least one patient image into a first model of a first neural network module to obtain a result having determined positions and ranges of an organ and a tumor of the patient image; inputting the result into a plurality of second models of a second neural network module, respectively, to obtain a plurality of prediction values corresponding to each of the plurality of second models and a model number predicting having cancer in the plurality of prediction values; and outputting a determined result based on the model number predicting having cancer and a number threshold value. Further, processes between the first model and the second models can be automated, thereby improving identification rate of pancreatic cancer.

公开(公告)号：US20220130038A1

公开(公告)日：2022-04-28

申请号：US17507948

申请日：2021-10-22

Applicant: National Taiwan University

Inventor： Wei-Chung Wang ,  Wei-Chih Liao ,  Kao-Lang Liu ,  Po-Ting Chen ,  Po-Chuan Wang ,  Ting-Hui Wu

IPC: G06T7/00 ,  G06T7/70 ,  G06V10/82 ,  G06V10/776 ,  G06V10/774 ,  G16H30/20 ,  G06N3/08 ,  A61B5/00

Abstract: A medical image analyzing system and a medical image analyzing method are provided and include inputting at least one patient image into a first model of a first neural network module to obtain a result having determined positions and ranges of an organ and a tumor of the patient image; inputting the result into a plurality of second models of a second neural network module, respectively, to obtain a plurality of prediction values corresponding to each of the plurality of second models and a model number predicting having cancer in the plurality of prediction values; and outputting a determined result based on the model number predicting having cancer and a number threshold value. Further, processes between the first model and the second models can be automated, thereby improving identification rate of pancreatic cancer.