search

wentaozhu / DeepLung

WACV18 paper "DeepLung: Deep 3D Dual Path Nets for Automated Pulmonary Nodule Detection and Classification"
Apache License 2.0
312 stars 143 forks source link

problem about evaluationScript #42

Closed jehovahxu closed 5 years ago

jehovahxu commented 5 years ago

how can i get a csv file like 3DRes18FasterR-CNN.csv

wentaozhu commented 5 years ago

You need to run 10 times. Use 1 fold as test, and other 9 folds as training. After that, concatenate 10 csv files.

jehovahxu commented 5 years ago

You need to run 10 times. Use 1 fold as test, and other 9 folds as training. After that, concatenate 10 csv files.

annotations_filename = './annotations/annotations.csv'  # path for ground truth annotations for the fold
annotations_excluded_filename = './annotations/annotations_excluded.csv'  # path for excluded annotations for the fold
seriesuids_filename = './annotations/seriesuids.csv'  # path for seriesuid for the fold
results_path = '/home/kejia/PycharmProjects/DeepLung-master/detector/results/res18/retrft969/val' #val'  # val' ft96'+'/val'#
sideinfopath = '/media/jehovah/Work/data/LUNA/propocess/subset9/'  # subset'+str(fold)+'/'  +str(fold)
datapath = '/media/jehovah/Work/data/LUNA/rowfile/subset9/'  # subset'+str(fold)+'/'

I can only run the frocwrtdetpepchluna16.py with such path . when i modified the subset9 to subset0 the error occur

image

jehovahxu commented 5 years ago

And i where can i get the output csv file about frocwrtdetpepchluna16.py . is that under every val fold ?

wentaozhu commented 5 years ago

For each fold, you need to use the fold's annotation to run the frocwrtdetpepchluna16.py

The froc w.r.t. detection probability and epoch on LUNA16 (frocwrtdetpepchluna16.py ) will tell you which detection probability and epoch is the best. Then you can collect the csv files of the epoch in the detection results folder. You are getting the csv file for the fold.