Early neurological deterioriation

- working preprocessing pipeline - working dataloader - trial of gridsearch
JulianKlug · Jun 23, 2024 · 7a9d41c · 7a9d41c
1 parent ee85170
commit 7a9d41c
Show file tree

Hide file tree

Showing 24 changed files with 2,147 additions and 8 deletions.
diff --git a/...va_stroke_unit_patient_characteristics.py → ...va_stroke_unit_patient_characteristics.py b/...va_stroke_unit_patient_characteristics.py → ...va_stroke_unit_patient_characteristics.py
diff --git a/meta_data/global_outcomes.ipynb → ...a/long_term_outcome/global_outcomes.ipynb b/meta_data/global_outcomes.ipynb → ...a/long_term_outcome/global_outcomes.ipynb
diff --git a/meta_data/imaging_meta_data.ipynb → ...long_term_outcome/imaging_meta_data.ipynb b/meta_data/imaging_meta_data.ipynb → ...long_term_outcome/imaging_meta_data.ipynb
diff --git a/meta_data/length_of_intermediate_care.ipynb → ...outcome/length_of_intermediate_care.ipynb b/meta_data/length_of_intermediate_care.ipynb → ...outcome/length_of_intermediate_care.ipynb
diff --git a/meta_data/mimic_patient_characteristics.py → ..._outcome/mimic_patient_characteristics.py b/meta_data/mimic_patient_characteristics.py → ..._outcome/mimic_patient_characteristics.py
diff --git a/meta_data/missingness_table.ipynb → ...long_term_outcome/missingness_table.ipynb b/meta_data/missingness_table.ipynb → ...long_term_outcome/missingness_table.ipynb
diff --git a/meta_data/number_of_datapoints.ipynb → ...g_term_outcome/number_of_datapoints.ipynb b/meta_data/number_of_datapoints.ipynb → ...g_term_outcome/number_of_datapoints.ipynb
diff --git a/..._data/patient_characteristics_table.ipynb → ...tcome/patient_characteristics_table.ipynb b/..._data/patient_characteristics_table.ipynb → ...tcome/patient_characteristics_table.ipynb
diff --git a/meta_data/patients_per_split.ipynb → ...ong_term_outcome/patients_per_split.ipynb b/meta_data/patients_per_split.ipynb → ...ong_term_outcome/patients_per_split.ipynb
diff --git a/meta_data/screened_population_stats.ipynb → ...m_outcome/screened_population_stats.ipynb b/meta_data/screened_population_stats.ipynb → ...m_outcome/screened_population_stats.ipynb
diff --git a/meta_data/other/ICH_meta.ipynb b/meta_data/other/ICH_meta.ipynb
@@ -0,0 +1,345 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "initial_id",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:31:38.271391Z",
+     "start_time": "2024-06-08T09:31:38.269256Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "import pandas as pd"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "849a22126f7029aa",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:31:43.483108Z",
+     "start_time": "2024-06-08T09:31:43.480718Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "path = '/Users/jk1/Library/CloudStorage/OneDrive-unige.ch/stroke_research/geneva_stroke_unit_dataset/data/stroke_registry/post_hoc_modified/stroke_registry_post_hoc_modified.xlsx'"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "4b4eb9f80c2168da",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:31:51.642087Z",
+     "start_time": "2024-06-08T09:31:45.739450Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "df = pd.read_excel(path)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "545fc6a9ebaebd43",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:31:51.673951Z",
+     "start_time": "2024-06-08T09:31:51.644093Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "df.head()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8d41b4ef12dd5e77",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:32:49.358260Z",
+     "start_time": "2024-06-08T09:32:49.353614Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "ich_df = df[df['Type of event'] == 'Intracerebral hemorrhage']"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6e54c2946daff8b9",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:37:33.621971Z",
+     "start_time": "2024-06-08T09:37:33.618054Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df = ich_df[(ich_df['GCS on admission'] < 15) & (ich_df['GCS on admission'] > 4) & (ich_df['NIH on admission'] > 5)]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "27ced1a91acbbdbe",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:37:37.180639Z",
+     "start_time": "2024-06-08T09:37:37.176622Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df.shape"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9e9d1929af575eff",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:39:58.297972Z",
+     "start_time": "2024-06-08T09:39:58.277700Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df.head()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "c376e1959b791708",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:43:39.071640Z",
+     "start_time": "2024-06-08T09:43:39.045643Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "cardiovascular_disease = ['MedHist Hypertension', 'MedHist Atrial Fibr.', 'MedHist CHD', 'MedHist Low ejection fract.', 'MedHist PAD']\n",
+    "# if any is yes then 1\n",
+    "restricted_ich_df['Cardiovascular disease'] = restricted_ich_df[cardiovascular_disease].apply(lambda x: 1 if 'yes' in x.str.lower().values else 0, axis=1)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "1ff6a75647795a2e",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:43:47.655733Z",
+     "start_time": "2024-06-08T09:43:47.650299Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['Cardiovascular disease'].value_counts(normalize=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "8d28806a743c04",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:44:23.163803Z",
+     "start_time": "2024-06-08T09:44:23.158016Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['MedHist Smoking'].value_counts(normalize=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6d2369e02258a2d8",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:39:15.940176Z",
+     "start_time": "2024-06-08T09:39:15.935400Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "# now with percenteges\n",
+    "restricted_ich_df['Prestroke disability (Rankin)'].value_counts(normalize=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "43ed70a1cae265ff",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:49:56.235548Z",
+     "start_time": "2024-06-08T09:49:56.200849Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['onset_date_time'] = pd.to_datetime(restricted_ich_df['Onset date'],\n",
+    "                                                     format='%Y%m%d').dt.strftime('%d.%m.%Y') + ' ' + \\\n",
+    "                                        pd.to_datetime(restricted_ich_df['Onset time'], format='%H:%M',\n",
+    "                                                       infer_datetime_format=True).dt.strftime('%H:%M')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "e3adb50cb03e64df",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:50:38.570257Z",
+     "start_time": "2024-06-08T09:50:38.560646Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['image_date_time'] = pd.to_datetime(restricted_ich_df['1st brain imaging date'],\n",
+    "                                                     format='%Y%m%d').dt.strftime('%d.%m.%Y') + ' ' + \\\n",
+    "                                        pd.to_datetime(restricted_ich_df['1st brain imaging time'], format='%H:%M',\n",
+    "                                                       infer_datetime_format=True).dt.strftime('%H:%M')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "996b3fdf5bc7229e",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:50:55.662336Z",
+     "start_time": "2024-06-08T09:50:55.655262Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['onest_to_image'] = pd.to_datetime(restricted_ich_df['image_date_time'], format='%d.%m.%Y %H:%M') - \\\n",
+    "                                     pd.to_datetime(restricted_ich_df['onset_date_time'], format='%d.%m.%Y %H:%M')"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "9e9a7b1c5349f7e9",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:50:59.509073Z",
+     "start_time": "2024-06-08T09:50:59.500323Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['onest_to_image'].describe()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "e2e74ce32473c31e",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:54:06.861381Z",
+     "start_time": "2024-06-08T09:54:06.855845Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "# restricted_ich_df['GCS on admission'] categories into 4-8 and 9-14\n",
+    "restricted_ich_df['GCS_4_8'] = restricted_ich_df['GCS on admission'].apply(lambda x: 1 if 4 <= x <= 8 else 0)\n",
+    "restricted_ich_df['GCS_9_14'] = restricted_ich_df['GCS on admission'].apply(lambda x: 1 if 9 <= x <= 14 else 0)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "451f84bb2fe9f6e4",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:54:12.085391Z",
+     "start_time": "2024-06-08T09:54:12.080386Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['GCS_4_8'].value_counts(normalize=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "54122428c7f5937b",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:54:17.381812Z",
+     "start_time": "2024-06-08T09:54:17.376348Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['GCS_9_14'].value_counts(normalize=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "cd12e7be02bb5b0",
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2024-06-08T09:55:30.013696Z",
+     "start_time": "2024-06-08T09:55:30.006481Z"
+    }
+   },
+   "outputs": [],
+   "source": [
+    "restricted_ich_df['NIH on admission'].describe()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "b257516d677011ec",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 2
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython2",
+   "version": "2.7.6"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}