-
Notifications
You must be signed in to change notification settings - Fork 8
/
Copy pathMimic.py
246 lines (193 loc) · 11.7 KB
/
Mimic.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
236
237
238
239
240
241
242
243
244
245
246
import os
import pandas as pd
import dask.dataframe as dd
from datetime import datetime
import sys
from google.colab import drive
drive.mount('/content/drive')
sys.path.append('..')
from utils.util import generic_utils
class MIMIC:
def __init__(self, dir_path) -> None:
self._dir_path = os.path.abspath(dir_path)
self._mimic_structure = [
'admissions', 'patients', 'transfers', 'd_hcpcs', 'd_icd_diagnoses',
'd_icd_procedures', 'd_labitems', 'diagnoses_icd', 'drgcodes', 'emar',
'emar_detail', 'hcpcsevents', 'labevents', 'microbiologyevents', 'pharmacy',
'poe', 'poe_detail', 'prescriptions', 'procedures_icd', 'services', 'chartevents',
'd_items', 'datetimeevents', 'icustays', 'inputevents', 'outputevents',
'procedureevents'
]
try:
assert os.path.exists(self._dir_path)
print(f"MIMIC dataset found at path : {self._dir_path}")
except AssertionError:
print(f"Dataset directory is empty. Exiting system.")
sys.exit(0)
def read_data(self, file_path):
if os.path.exists(file_path):
df = dd.read_parquet(file_path)
return df
else:
print(f"Given file path {file_path} doesn't exist.")
def join_path(self, file_name):
return os.path.join(self._dir_path, f'{file_name}.parquet')
def read_admissions(self):
return self.read_data(self.join_path('admissions'))
def read_patients(self):
return self.read_data(self.join_path('patients'))
def read_transfers(self):
return self.read_data(self.join_path('transfers'))
def read_d_hcpcs(self):
return self.read_data(self.join_path('d_hcpcs'))
def read_d_icd_diagnoses(self):
return self.read_data(self.join_path('d_icd_diagnoses'))
def read_d_icd_procedures(self):
return self.read_data(self.join_path('d_icd_procedures'))
def read_d_labitems(self):
return self.read_data(self.join_path('d_labitems'))
def read_diagnoses_icd(self):
return self.read_data(self.join_path('diagnoses_icd'))
def read_drgcodes(self):
return self.read_data(self.join_path('drgcodes'))
def read_emar(self):
return self.read_data(self.join_path('emar'))
def read_emar_detail(self):
return self.read_data(self.join_path('emar_detail'))
def read_hcpcsevents(self):
return self.read_data(self.join_path('hcpcsevents'))
def read_labevents(self):
return self.read_data(self.join_path('labevents'))
def read_microbiologyevents(self):
return self.read_data(self.join_path('microbiologyevents'))
def read_pharmacy(self):
return self.read_data(self.join_path('pharmacy'))
def read_poe(self):
return self.read_data(self.join_path('poe'))
def read_poe_detail(self):
return self.read_data(self.join_path('poe_detail'))
def read_prescriptions(self):
return self.read_data(self.join_path('prescriptions'))
def read_procedures_icd(self):
return self.read_data(self.join_path('procedures_icd'))
def read_services(self):
return self.read_data(self.join_path('services'))
def read_chartevents(self):
return self.read_data(self.join_path('chartevents'))
def read_d_items(self):
return self.read_data(self.join_path('d_items'))
def read_datetimeevents(self):
return self.read_data(self.join_path('datetimeevents'))
def read_icustays(self):
return self.read_data(self.join_path('icustays'))
def read_inputevents(self):
return self.read_data(self.join_path('inputevents'))
def read_outputevents(self):
return self.read_data(self.join_path('outputevents'))
def read_procedureevents(self):
return self.read_data(self.join_path('procedureevents'))
class MIMICManipulations:
def __init__(self, dir_path):
self._mimic_instance = MIMIC(dir_path)
self._util_instance = generic_utils()
def filter_admission(self, df):
df_admission = df.drop(['hadm_id', 'edregtime', 'edouttime', 'deathtime'], axis=1)
df_admission['admittime'] = dd.to_datetime(df_admission['admittime'])
df_admission['dischtime'] = dd.to_datetime(df_admission['dischtime'])
df_admission['los_admission'] = (df_admission['dischtime'] - df_admission['admittime']).dt.total_seconds()/86400
df_admission = df_admission.drop(['admittime', 'dischtime'], axis=1)
df_admission = df_admission[df_admission['los_admission'] > 0]
df_admission = self._util_instance.remove_duplicates_and_re_index(df_admission, 'subject_id')
return df_admission
def filter_patients(self, df):
df_patient = df.drop(['anchor_year', 'anchor_year_group', 'dod'], axis=1)
df_patient = self._util_instance.remove_duplicates_and_re_index(df_patient, 'subject_id')
return df_patient
def filter_transfers(self, df):
df_transfer = df.drop(['hadm_id', 'transfer_id', 'intime', 'outtime'], axis=1)
df_transfer = self._util_instance.remove_duplicates_and_re_index(df_transfer, 'subject_id')
return df_transfer
def merge_core_tables(self):
df_admission = self.filter_admission(self._mimic_instance.read_admissions())
df_patient = self.filter_patients(self._mimic_instance.read_patients())
df_transfers = self.filter_transfers(self._mimic_instance.read_transfers())
df_core_merge = df_admission.merge(df_patient, how='left', on=['subject_id'])\
.merge(df_transfers, how='outer', on=['subject_id', 'hadm_id'])
return df_core_merge
def filter_diagnoses_icd(self, df):
df_diagnoses_icd = df.drop(['hadm_id', 'seq_num', 'icd_version'], axis=1)
df_diagnoses_icd = df_diagnoses_icd.rename(columns = {'icd_code':'diagnosis_icd_code'})
df_diagnoses_icd = self._util_instance.remove_duplicates_and_re_index(df_diagnoses_icd, 'subject_id')
return df_diagnoses_icd
def filter_procedures_icd(self, df):
df_procedures_icd = df.drop(['hadm_id', 'seq_num', 'chartdate', 'icd_version'], axis=1)
df_procedures_icd = df_procedures_icd.rename(columns = {'icd_code':'procedures_icd_code'})
df_procedures_icd = self._util_instance.remove_duplicates_and_re_index(df_procedures_icd, 'subject_id')
return df_procedures_icd
def filter_lab_events(self, df):
df_lab_events = df.drop(['labevent_id', 'specimen_id', 'itemid', 'charttime', 'storetime', 'valueuom',
'ref_range_lower', 'ref_range_upper', 'comments', 'valuenum', 'hadm_id'], axis=1)
df_lab_events = self._util_instance.remove_duplicates_and_re_index(df_lab_events, 'subject_id')
return df_lab_events
def merge_lab_events_d_labitems(self):
df_lab_events = self.filter_lab_events(self._mimic_instance.read_labevents())
df_d_lab_items = self._mimic_instance.read_d_labitems()
df_lab_events_merge = df_lab_events.merge(df_d_lab_items, how='left', on=['itemid'])
return df_lab_events_merge
def filter_drgcodes(self, df):
df_drgcodes = df.drop(['hadm_id', 'description', 'drg_severity', 'drg_mortality'], axis=1)
df_drgcodes = self._util_instance.remove_duplicates_and_re_index(df_drgcodes, 'subject_id')
return df_drgcodes
def filter_emar(self, df):
df_emar = df[['subject_id', 'medication', 'event_txt']]
df_emar = self._util_instance.remove_duplicates_and_re_index(df_emar, 'subject_id')
return df_emar
def filter_poe(self, df):
df_poe = df[['subject_id', 'order_type', 'transaction_type']]
df_poe = self._util_instance.remove_duplicates_and_re_index(df_poe, 'subject_id')
return df_poe
def merge_emar_poe(self):
df_emar = self.filter_emar(self._mimic_instance.read_emar())
df_poe = self.filter_poe(self._mimic_instance.read_poe())
df_merge_emar_poe = df_emar.merge(df_poe, how='outer', on=['poe_id', 'subject_id', 'hadm_id'])
df_merge_emar_poe.drop(['emar_id', 'poe_id'], axis=1)
return df_merge_emar_poe
def filter_microbiologyevents(self, df):
df_microbiologyevents = df[['subject_id', 'org_name', 'test_name', 'quantity', 'ab_name']]
df_microbiologyevents = self._util_instance.remove_duplicates_and_re_index(df_microbiologyevents, 'subject_id')
return df_microbiologyevents
def filter_prescriptions(self, df):
df_prescriptions = df[['subject_id', 'drug', 'route']]
df_prescriptions = self._util_instance.remove_duplicates_and_re_index(df_prescriptions, 'subject_id')
return df_prescriptions
def filter_service(self, df):
df_service = df[['subject_id', 'curr_service']]
df_service = self._util_instance.remove_duplicates_and_re_index(df_service, 'subject_id')
return df_service
def merge_core_hosp_tables(self):
df_core_merged = self.merge_tables()
df_hosp_lab_events_merge = self.merge_lab_events_d_labitems()
df_hosp_emar_poe_merge = self.merge_emar_poe()
df_core_diagnoses_merge = df_core_merged.merge(self.filter_diagnoses_icd(self._mimic_instance.read_diagnoses_icd()), how='outer', on=['subject_id', 'hadm_id'])\
.merge(self.filter_procedures_icd(self._mimic_instance.read_procedures_icd()), how='outer', on=['subject_id', 'hadm_id', 'icd_code', 'icd_version'])
df_core_hosp_merge = df_core_diagnoses_merge.merge(self.filter_drgcodes(self._mimic_instance.read_drgcodes()), how='outer', on=['subject_id', 'hadm_id'])
df_core_hosp_merge = df_core_hosp_merge.merge(df_hosp_lab_events_merge, how='outer', on=['subject_id', 'hadm_id'])
df_core_hosp_merge = df_core_hosp_merge.merge(df_hosp_emar_poe_merge, how='outer', on=['subject_id', 'hadm_id'])\
.merge(self.filter_prescriptions(self._mimic_instance.read_prescriptions()), how='outer', on=['subject_id', 'hadm_id'])\
.merge(self.filter_microbiologyevents(self._mimic_instance.read_microbiologyevents()), how='outer', on=['subject_id', 'hadm_id'])\
.merge(self.filter_service(self._mimic_instance.read_services()), how='outer', on=['subject_id', 'hadm_id'])
return df_core_hosp_merge
def filter_icustays(self, df):
df_icustays = df.drop(['hadm_id', 'stay_id', 'intime', 'outtime'], axis=1)
df_icustays = self._util_instance.remove_duplicates_and_re_index(df_icustays, 'subject_id')
return df_icustays
def filter_chartevents(self, df):
df_chartevents = df.drop(['hadm_id', 'stay_id', 'charttime', 'storetime', 'value', 'valuenum', 'valueuom', 'warning'], axis=1)
df_chartevents = self._util_instance.remove_duplicates_and_re_index(df_chartevents, 'subject_id')
return df_chartevents
def merge_core_hosp_icu_tables(self):
df_core_hosp_merged = self.merge_core_hosp_tables()
df_core_hosp_icu_merged = df_core_hosp_merged.merge(self.filter_icustays(self._mimic_instance.read_icustays()), how='outer', on=['subject_id', 'hadm_id', ])\
.merge(self.filter_chartevents(self._mimic_instance.read_chartevents()), how='outer', on=['subject_id', 'hadm_id', 'stay_id'])
df_core_hosp_icu_merged.drop(['stay_id'], axis=1)
return df_core_hosp_icu_merged