Analyzing unstructured data to unmask cognitive impairment

What if clinicians could use technology to find early indicators of cognitive impairment hidden in the medical record? In this episode, Dr. Gordon Moore speaks with Andrea Gilmore-Bykovskyi, PhD, RN, who leads an interdisciplinary research program at the University of Wisconsin Madison that is analyzing free text in the EMR to identify patients with early signs of Alzheimer’s and other dementias. Using machine learning algorithms, Dr. Gilmore-Bykovskyi’s research has uncovered terms that relate to a decline in cognition, paving the way for more accurate and timely diagnoses. 

Read the podcast episode transcript.

My interest in dementia is in trying to understand how we can support clinicians and patients by bringing to bear information that's already been collected...so we improve recognition of dementia and also provide more continuity as people move across our very fragmented system of care.
— Dr. Gilmore-Bykovskyi

Resources

Gilmore-Bykovskyi, Andrea et al. The consequences of poor communication during transitions from hospital to skilled nursing facility; a qualitative study. 

Gilmore-Bykovskyi, Andrea et al. Missing Warfarin discharge communication and risk of 30-day rehospitalization and/or death: Retrospective cohort study.

Gilmore-Bykovskyi, Andrea et al. Unstructured clinical documentation reflecting cognitive and behavioral dysfunction: toward an EHR-based phenotype for cognitive impairment