EHRKit Documentation

Directory Structure

• EHRKit
• ehrkit
  • ehrkit.py
  • solr_lib.py
  • classes.py
• external
  • icd9
  • status: incorporated but doesn't seem useful
  • patient2vec
  • status: not yet incorporated
  • could be useful:
  • phrase-at-scale
  • status: incorporated, but only a crude version
  • problem: does not return the phrases but outputs it to files
• scripts
  • helper scripts for handling data
  • see README_scripts for more descriptions
• tests
  • see README_tests for more descriptions

ehrkit.py functions

• helper functions:
• def start_session(db_user, db_pass): -> dict:
• def createPatient(data):
• def flatten(lst):
• def numbered_print(lst):
• def init_embedding_model():
• def get_abbs_sent_ids(text):
• def post_single_dict_to_solr(d: dict, core: str)-> None
• def abbs_disambiguate(ABB):

• def get_documents_solr(query):

• class ehr_db

• attributes:
  • cnx -- MySQL connection object
  • cur -- MySQL cursor
  • patients = {} -- Patient (from classes.py) dictionary

• methods:

  • def close_session(self):
  • def get_patients(self, n):
  • def count_patients(self):
  • def count_gender(self, gender):
  • def count_docs(self, query, getAll = False, inverted = False):
  • def get_note_events(self):
  • def longest_NE(self):
  • def get_document(self, id):
  • def get_all_patient_document_ids(self, patientID):
  • def list_all_patient_ids(self):
  • def list_all_document_ids(self):
  • def get_document_sents(self, docID):
  • def get_abbreviations(self, doc_id):
  • def get_abbreviation_sent_ids(self, doc_id):
  • def get_documents_d(self, date):
  • def get_documents_q(self, query, n = -1):
  • def get_documents_icd9_alt(self,query):
  • def get_documents_icd9(self,code):
  • def get_prescription(self):
  • def extract_key_words(self, text):
  • def count_all_prescriptions(self):
  • def get_diagnoses(self):
  • def get_procedures(self):
  • def extract_patient_words(self, patientID):
  • def output_note_events_file_by_patients(self, directory):
  • def output_note_events_discharge_summary(self, directory):

• yet to refactor:

  • def extract_key_words(self, text):
  • def get_abbreviations(self, doc_id):
  • def get_abbreviation_sent_ids(self, doc_id):
• unfinished:
  • def docs_with_phrase(self, phrase):
  • def outputAbbreviation(self, directory):
  • def extract_phrases(self, docID):