-
Notifications
You must be signed in to change notification settings - Fork 0
/
gamma3d.py
235 lines (195 loc) · 10.8 KB
/
gamma3d.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
'''
Tested with Python 3.11.1 & pymedphys 0.36.1
'''
import pymedphys, pydicom, os, time
from pymedphys import gamma
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import subprocess
import multiprocessing
def gamma_3d(dpt, dta, path_to_ref, path_to_eval, cutoff=10, interp=10, max=1.1):
# Load the two DICOM dose files to be compared
try:
ref = pymedphys.dicom.zyx_and_dose_from_dataset(pydicom.read_file(path_to_ref))
eval = pymedphys.dicom.zyx_and_dose_from_dataset(pydicom.read_file(path_to_eval))
except: # if datasets are already loaded by pydicom
ref = pymedphys.dicom.zyx_and_dose_from_dataset(path_to_ref)
eval = pymedphys.dicom.zyx_and_dose_from_dataset(path_to_eval)
# Get the dose grids and coordinate arrays from the DICOM files
dose_ref_grid, coord_ref = ref[:2]
dose_eval_grid, coord_eval = eval[:2]
# Define the gamma analysis parameters
dose_percent_threshold = dpt
distance_mm_threshold = dta
lower_percent_dose_cutoff = cutoff
interp_fraction = interp
maximum_gamma = max
# Calculate the gamma index matrix
gamma_analysis = gamma(
dose_ref_grid, coord_ref,
dose_eval_grid, coord_eval,
dose_percent_threshold,
distance_mm_threshold,
lower_percent_dose_cutoff,
interp_fraction,
maximum_gamma,
quiet=False
)
# Filter the gamma index matrix to remove noise
filtered_gamma = gamma_analysis[~np.isnan(gamma_analysis)]
# Generate return values
passed = np.sum(filtered_gamma <= 1) / len(filtered_gamma)
max_gamma = np.max(filtered_gamma)
mean_gamma = np.mean(filtered_gamma)
std_gamma = np.std(filtered_gamma)
return passed, max_gamma, mean_gamma, std_gamma
def voxel_wise(dpt, path_to_ref, path_to_eval, cutoff=10):
# Load the two DICOM dose files to be compared
ref = pymedphys.dicom.zyx_and_dose_from_dataset(pydicom.read_file(path_to_ref))
eval = pymedphys.dicom.zyx_and_dose_from_dataset(pydicom.read_file(path_to_eval))
# Get the dose grids and coordinate arrays from the DICOM files
dose_ref_grid, coord_ref = ref[:2]
dose_eval_grid, coord_eval = eval[:2]
ref_max = np.nanmax(coord_ref)
mask = (coord_ref < (cutoff / 100) * ref_max) & (~np.isnan(coord_ref))
# coord_ref[mask] = np.nan
percent_difference = np.abs(coord_ref - coord_eval) / ref_max * 100
passed = np.count_nonzero((percent_difference < dpt) & (~np.isnan(percent_difference))) / np.count_nonzero((~np.isnan(percent_difference)))
print(passed)
def run_chunk(chunk):
for arg in chunk:
full_eval(arg)
def parallelize(args):
n_cores = multiprocessing.cpu_count() - 2
chunks = [args[i::n_cores] for i in range(n_cores)]
print('Queueing jobs on', n_cores, 'threads...')
with multiprocessing.Pool(n_cores) as p:
p.map(run_chunk, chunks)
def full_eval(id):
root_dir = r'N:\fs4-HPRT\HPRT-Data\ONGOING_PROJECTS\AutoPatSpecQA\02_cCTPatients\Logfiles\converted'
plan_dir = r'N:\fs4-HPRT\HPRT-Data\ONGOING_PROJECTS\AutoPatSpecQA\02_cCTPatients\Logfiles\DeliveredPlans'
# print(f'>>> START ID {id}')
doses = os.path.join(plan_dir, id + '\Doses\WaterIso1mmLateral')
log_dir = os.path.join(root_dir, '..', 'extracted')
for file in os.listdir(log_dir):
if file.__contains__(str(id)) and file.__contains__('record') and file.endswith('csv'):
record_df = pd.read_csv(os.path.join(log_dir, file), index_col='TIME', dtype={'FRACTION_ID':str, 'BEAM_ID':str})
elif file.__contains__(str(id)) and file.__contains__('delta') and file.endswith('csv'):
delta_df = pd.read_csv(os.path.join(log_dir, file), index_col='UNIQUE_INDEX', dtype={'FRACTION_ID':str, 'BEAM_ID':str})
if len(record_df) == len(delta_df):
record_df.reset_index(inplace=True), delta_df.reset_index(inplace=True)
df = pd.concat([record_df, delta_df], axis=1)
df = df.loc[:,~df.columns.duplicated()].copy()
else:
print(' /x\ DF length mismatch!')
return None
# accept only more than 15 fx
for beam in df.BEAM_ID.drop_duplicates():
beam_df = df.loc[df.BEAM_ID == beam]
if len(beam_df.FRACTION_ID.drop_duplicates()) < 15:
df.drop(beam_df.index, inplace=True)
# prepare output
beams = df.BEAM_ID.drop_duplicates().to_list()
fractions = df.FRACTION_ID.drop_duplicates().to_list()
# target_df = pd.DataFrame(index=fractions, columns=[f'Beam_{bid}_LOGPASS' for bid in beams] + ['TOTAL_LOGPASS'] + [f'Beam_{bid}_GAMMAPASS' for bid in beams] + ['TOTAL_GAMMAPASS'])
gamma_crits = [(3, 3), (2, 2), (1, 1)] # dpt, dta
log_crits = [2, 1.5, 1] # mm distance to plan spot
target_df = pd.DataFrame(columns=['FRACTION_ID'] + [f'LOGPASS_{crit}_mm' for crit in log_crits] + [f'GAMMAPASS_{crit[0]}_{crit[1]}' for crit in gamma_crits])
df['DISTANCE'] = np.sqrt(df['DELTA_X(mm)'].copy() ** 2 + df['DELTA_Y(mm)'].copy() ** 2)
# calculate physical spot pass rate
for i, fx in enumerate(fractions):
if fx != fractions[-1]:
print(f' > Calculating physical passrates ({i + 1}/{len(fractions)})..', end='\r')
else:
print(f' > Calculating physical passrates ({i + 1}/{len(fractions)})..', end='\n')
for crit in log_crits:
fx_df = df.loc[df.FRACTION_ID == fx]
total_log_pass = fx_df.loc[fx_df.DISTANCE <= crit].MU.sum() / fx_df.MU.sum() * 100
target_df.loc[i, 'FRACTION_ID'] = fx
target_df.loc[i, f'LOGPASS_{crit}_mm'] = total_log_pass
# calculate gamma dose pass rate
for i, fx in enumerate(fractions):
if fx != fractions[-1]:
print(f' > Calculating gamma-3D passrates ({i + 1}/{len(fractions)})..', end='\r')
else:
print(f' > Calculating gamma-3D passrates ({i + 1}/{len(fractions)})..', end='\n')
for crit in gamma_crits:
for bid in beams:
has_ref_plan, has_ref_beam, has_eval_plan, has_eval_beam = False, False, False, False
# get reference doses
for file in os.listdir(doses):
if file.startswith('RD') and file.endswith('.dcm'): # accept only RS export dose dicoms (these are the only dose files in /Doses)
ds = pydicom.read_file(os.path.join(doses, file))
if ds.DoseSummationType == 'PLAN':
ref_plan_dose = ds
has_ref_plan = True
elif ds.DoseSummationType == 'BEAM':
if ds.InstanceNumber == int(bid):
ref_beam_dose = ds
has_ref_beam = True
if has_ref_plan and has_ref_beam:
break
# get evaluation doses
for file in os.listdir(os.path.join(doses, 'eval')):
if file.startswith('RD') and file.endswith('.dcm'):
ds = pydicom.read_file(os.path.join(doses, 'eval', file))
if ds.SeriesDescription.__contains__(str(fx)):
if ds.DoseSummationType == 'PLAN':
eval_plan_dose = ds
has_eval_plan = True
elif ds.DoseSummationType == 'BEAM':
if ds.InstanceNumber == int(bid):
eval_beam_dose = ds
has_eval_beam = True
if has_eval_plan and has_eval_beam:
break
# dpt, dta = crit
# beam_gamma_pass, _, _, _ = gamma_3d(dpt, dta, ref_beam_dose, eval_beam_dose)
# target_df.loc[fx, f'Beam_{bid}_GAMMAPASS'] = beam_gamma_pass
dpt, dta = crit
target_df.loc[i, 'FRACTION_ID'] = fx
if has_eval_plan and has_ref_plan:
plan_gamma_pass, _, _, _ = gamma_3d(dpt, dta, ref_plan_dose, eval_plan_dose)
target_df.loc[i, f'GAMMAPASS_{dpt}_{dta}'] = plan_gamma_pass * 100
else:
if crit == gamma_crits[0]: print(f' /!\ Missing evaluation dose for fraction {fx}')
target_df.loc[i, f'GAMMAPASS_{dpt}_{dta}'] = np.nan
written = False
while not written:
try:
target_df.to_csv(os.path.join(doses, f'{id}_H2O_isoshift_1mm_lat_3Dgamma.csv'))
break
except PermissionError:
input(f' /!\ Permission denied for target CSV, close file..')
except FileNotFoundError:
print(f' /!\ Target location {doses} not found')
print(' ... Waiting for drive mapping ...')
time.sleep(5)
if __name__ == '__main__':
# root_dir = r'N:\fs4-HPRT\HPRT-Data\ONGOING_PROJECTS\AutoPatSpecQA\02_cCTPatients\Logfiles\DeliveredPlans\1663630\Doses'
# eval_dir = os.path.join(root_dir, 'eval')
# criteria = [(2, 2), (1, 1), (1, 0.5), (0.5, 1), (0.5, 0.5)]
ponaqua_qualified = [id.strip('\n') for id in open(r'N:\fs4-HPRT\HPRT-Data\ONGOING_PROJECTS\AutoPatSpecQA\02_cCTPatients\qualified_IDs.txt', 'r').readlines()]
all_fx = 0
for id in ponaqua_qualified:
pat_fx = full_eval(id)
print(id, pat_fx)
all_fx += pat_fx
print('ALL', all_fx)
# if int(id) in [671075,1230180,1635796,1683480,1676596,280735,1367926]:
# ponaqua_qualified.remove(id)
# to_do = ["1669130"]
# parallelize(to_do)
# ref = os.path.join(root_dir, 'RD_R1_HNO.dcm')
# evals = [os.path.join(eval_dir, dcm) for dcm in os.listdir(eval_dir) if dcm.__contains__('RD') and dcm.endswith('.dcm')]
# data = ['Fx\tPass\tMax\tMean\tStd\n']
# for fx, eval in enumerate(evals):
# print(f'>> STARTING 3D gamma evaluation for {fx+1}/{len(evals)} with {dpt}%/{dta}mm..')
# voxel_wise(1, ref, eval)
# passed, max, mean, std = gamma_3d(dpt, 1000, ref, eval)
# print(passed)
# # data.append(f'{fx+1}\t{np.round(passed, 5)}\t{np.round(max, 5)}\t{np.round(mean, 5)}\t{np.round(std, 5)}\n')
# with open(os.path.join(root_dir, f'gamma3D_{dpt}p_{dta}mm.txt'), 'w+') as file:
# file.writelines(data)
# file.close()