Visualization of model prediction results (heatmaps) indicating the classification confidence of each pixel.

Visualization of model prediction results (heatmaps) indicating the classification confidence of each pixel.

<p>(A) Malignant images and (B) corresponding heatmap. (C) Benign image and (D) corresponding heatmap.</p>

Saved in:

Bibliographic Details
Main Author:	Qingyang Wang (460264) (author)
Other Authors:	Yazhi Luo (20648632) (author), Ying Zhao (16546) (author), Shuhao Wang (4637881) (author), Yiru Niu (11141535) (author), Jinxi Di (20648635) (author), Jia Guo (156736) (author), Guorong Lan (20648638) (author), Lei Yang (102719) (author), Yu Shan Mao (20648641) (author), Yuan Tu (11231151) (author), Dingrong Zhong (740693) (author), Pei Zhang (232846) (author)
Published:	2025
Subjects:	Medicine Cell Biology Biotechnology Cancer Biological Sciences not elsewhere classified Mathematical Sciences not elsewhere classified Information Systems not elsewhere classified xlink "> compared provides better preservation low interobserver consistency deep neural networks pulmonary lesions based model &# 8217 intraoperative cytological segmentation mean pixel accuracy pixel level mean intersection yolov8 algorithm test set strongly correlated sections individually proposed method pathologist diagnoses model predictions model achieved microscopic images make clinical lung cancer level decisions less invasive initial diagnosis impressive speed histological examination gold standard experienced pathologist detection model deemed suitable consuming process complete morphology automated recognition
Tags:	Add Tag No Tags, Be the first to tag this record!

Be the first to leave a comment!

Cannot write session to /tmp/vufind_sessions/sess_fluu8d82o6t47em7lhs1p3m6b2