Projects

admin — Mon, 22 Jun 2015 17:05:18 +0000

In previous projects we have developed a combination of rule-based and machine-learning methods to identify diseses that a patient has or does not have (“disease status”), including identification of various co-morbidities, problems, tests and treatments. We also work on the extraction of medication-related information (such as medication name, dosage, reason for taking, frequency, duration etc.) and clinical temporal text mining. These tasks were assessed as part of an international text mining challenge in the clinical/healthcare domain:

2012: Temporal mining of clinical narratives (ranked shared 1st for the temporal expression extraction task)
2011: Sentiment analysis of suicide notes (ranked 8th/26 teams, invited talk)
2009: Medication extraction from clinical notes (ranked shared 2nd-3rd/19 teams, invited talk)
2008: Extraction of obesity and co-morbidity status from hospital discharge summaries (ranked 1st for the explicit extraction, invited talk)

For more details, see Natural Language Processing for Clinical Data: Continuous Success at i2b2 Challenges.

Tools and Data

admin — Mon, 22 Jun 2015 17:03:41 +0000

Toolsets and resources developed as part of the i2b2 2014 challenge:

Task 1: De-identification of Clinical Narratives
Task 2: Heart disease risk factor identification in EHRs

Toolset and resources developed as part of the i2b2 2012 Temporal Relation challenge:

Clinical NorMA
TERN
TEid
CliNER
Clinical Departments dictionary

Related publications:

Kovačević A, Dehghan A, Filannino M, Keane JA, Nenadic, G: Combining rules and machine-learning for extraction of temporal expressions and events from clinical narratives. Journal of the American Medical Informatics Association. (submitted)

Clinical NorMA: temporal expression normaliser

Clinical NorMA is a rule-based temporal expression normaliser explicitly designed for clinical data. It is open-source (GNU licence) and is written in Python. It has been used by our team to participate to i2b2 2012 challenge.

People involved: Michele Filannino.
Source code: https://github.com/filannim/clinical-norma.

TERN: TEmporal expressions Recognizer and Normalizer

TERN is a rule-based temporal expressions identification and normalisation software; designed for clinical data. The identification (TEid) and normalization component (Clinical NorMA) were used and evaluated as part of the Temporal Relation challenge (i2b2 2012). This software provides various features for easy integration into existing pipelines (e.g., server mode); see code repository / README file.

People involved: Azad Dehghan
Source code: https://sourceforge.net/projects/temporal-x/.

TEid: TEmporal expressions IDentifier

TEid is a rule-based temporal expressions identification software; designed for clinical data. This software was used and evaluated as part of the Temporal Relation challenge (i2b2 2012). This software provides various features for easy integration into existing pipelines (e.g., server mode); see code repository / README file.

People involved: Azad Dehghan
Source code: https://sourceforge.net/projects/temporal-id/.

CliNER

CliNER is a command line tool for tagging of clinically relevant events and recognition and normalisation (using Clinical NorMA) of temporal expressions. The tool is a part of the set of tools developed for the i2b2 2012 challenge. CliNER also provides integration with the rest of the tools developed during the challenge.

People involved: Aleksandar Kovačević.
Download: CliNER.tgz.

Clinical Departments Dictionary

A clinical departments dictionaty/gazetteer; semi-automatically curated from i2b2 2010 and i2b2 2012 datasets.

People involved: Azad Dehghan
Resource: Email Me.