US 12,488,469 B2
Segmentation method for lipid droplet in pathological slice image
Pau-Choo Chung, Tainan (TW); Hung-Wen Tsai, Tainan (TW); Jyun-Sin Wu, Tainan (TW); Yu-Hsuan Lin, Tainan (TW); and Kuo-Sheng Cheng, Tainan (TW)
Assigned to NATIONAL CHENG KUNG UNIVERSITY, Tainan (TW)
Filed by NATIONAL CHENG KUNG UNIVERSITY, Tainan (TW)
Filed on Feb. 22, 2023, as Appl. No. 18/173,042.
Claims priority of application No. 111108368 (TW), filed on Mar. 8, 2022.
Prior Publication US 2023/0289975 A1, Sep. 14, 2023
Int. Cl. A61B 5/00 (2006.01); G06T 7/12 (2017.01); G06T 7/155 (2017.01); G16H 30/40 (2018.01)
CPC G06T 7/12 (2017.01) [G06T 7/155 (2017.01); G16H 30/40 (2018.01); G06T 2207/20021 (2013.01); G06T 2207/20036 (2013.01); G06T 2207/20076 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30084 (2013.01); G06T 2207/30204 (2013.01)] 15 Claims
OG exemplary drawing
 
1. A segmenting method for a pathological slice image, wherein the method is performed by a computer system and comprises:
inputting the pathological slice image into a machine learning model to output a region map and a boundary map with respect to a plurality of lipid droplets, wherein the region map comprises a plurality of region probabilities, and the boundary map comprises a plurality of boundary probabilities; and
segmenting the lipid droplets from the pathological slice image according to the region probabilities and the boundary probabilities.