Skip to content

Commit

Permalink
UPDATE 2024-07-02
Browse files Browse the repository at this point in the history
UPDATE 2024-07-02
  • Loading branch information
babbu3682 committed Jul 2, 2024
1 parent 88604ef commit c5c9933
Show file tree
Hide file tree
Showing 793 changed files with 13,048 additions and 16,370 deletions.
7 changes: 6 additions & 1 deletion .gitignore
Original file line number Diff line number Diff line change
Expand Up @@ -119,4 +119,9 @@ Notebook/
predictions/
백업용/
train.sh
test.sh
test.sh
create_datasets/Mayo_BACKUP.py
create_datasets/Sinogram_BACKUP.py
test_BACKUP.py
engine_BACKUP.py
잡다/
10 changes: 10 additions & 0 deletions BLIND_READER_STUDY/Newest_Version/README.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,10 @@
# Evaluation criteria for the blind reader study. <Newest Version>

The provided "Blind_Reader_Study_OMR.xlsx" file contains three sheets:
+ submission_sheet: This sheet contains the questionnaire given to participants, all anonymized.
+ answer_sheet: This sheet contains the answer key with labeled data.
+ expert_results: This sheet summarizes the results from three experts.s


A blind reader study was conducted using image slices from 15 randomly selected groups, each including brain and abdomen test sets with 14 pairs of images: an LDCT image, an NDCT image, and denoised LDCT images. Three board-certified radiologists, each with 15 years of experience, independently evaluated these images for noise reduction, contrast retention, and structure preservation using specific scoring systems. Noise reduction was assessed on a 1-14 point scale, with NDCT set at 7 points; scores above 7 indicated over-smoothing, while scores below 7 indicated insufficient noise reduction. Contrast retention and structural preservation were evaluated on a 1-5 point scale, with LDCT at 1 and NDCT at 5. Detailed evaluation criteria aimed to determine which images most closely resembled NDCT images, focusing on noise reduction, contrast retention, and structure preservation. The radiologists underwent preliminary reviews with non-test images to standardize their evaluations. The study found that MTD-GAN performed closest to NDCT in noise reduction for brain images and contrast retention for abdomen images, with competitive scores in other metrics. For comprehensive evaluation criteria and examples, see Appendix I-F. Table III shows the average and standard deviation of the blind test results per sample and per radiologist, indicating the superior performance of MTD-GAN in the evaluated metrics.

Binary file not shown.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Binary file not shown.
File renamed without changes
File renamed without changes
Loading

0 comments on commit c5c9933

Please sign in to comment.