When: Thursday 24th October 2019
Time: 3:00pm BST; 4:00pm CEST; 11:00am EDT; 8:00am PDT
Duration: 60 minutes.
"The Use of Natural Language Processing to Improve Phenotype Extraction for Precision Medicine - University of Iowa NLP-Based Phenome Extraction from the EHR"
Precision medicine focuses on disease treatment and prevention, taking into account the variability in genes, environment, and lifestyle between individual patients. In order to understand the best treatment pathway for a particular patient or group of patients, it is important to be able to access and analyze detailed information from the medical records of patients, and ideally broader aspects beyond their medical history.
Clinical genomic testing (e.g. chromosome microarray, gene panel testing, exome/genome sequencing) is performed on thousands of patients annually and relies heavily on manual chart review by a healthcare professional to identify the patient's clinical phenotype. Natural language processing (NLP) is being used to obtain high-quality comprehensive phenotype information from the electronic medical record for patients who have undergone clinical genomic testing. While this project is still on-going, the work has the potential to aid in the interpretation of genetic test results, and to directly improve the diagnosis and clinical care of patients seen at the University of Iowa Stead Family Children's Hospital.
80% of EHR data is unstructured but has many insights locked in clinical notes. Researchers and clinicians need a way to rapidly search and extract key information to use in clinical research and improved care. Learn how NLP is being used to transform unstructured source data into clinical and research decision support insights:
- Phenotypic characteristics are often extracted by exhaustive manual chart review, find out how NLP can reduce this time significantly
- Understand how EHR data can be exported and used in NLP systems, what issues to expect
- Explore the training and evaluation of NLP results to achieve high levels of accuracy