Research
My work falls into two broad areas of work:
- Applying speech and language technologies to problems relating to communication disorders, both in terms of assessment as well as Augmentative and Alternative Communication applications
- Automated analysis of scientific literature and electronic medical record data
Research Interests
From a clinical perspective, my research interests include:
- Automated approaches to screening for and otherwise assessing language and communication disorders, both in pediatric and adult populations
- Language deficits in post-stroke aphasia
- Confrontation naming tests
- Autism Spectrum Disorder
- Augmentative and alternative communication
- Automation to support systematic reviews and evidence-based medicine
- Patient cohort discovery from electronic medical records
- Information extraction and data mining in free-text clinical notes
- Brain-computer interfaces for communication
From a computational perspective, my research interests include:
- Evaluation of machine learning algorithms
- Information retrieval
- Text normalization
- Computational analysis of dialogues & discourse
- Vector-space models of lexical and document semantics
- Hierarchical and temporal document representations
- Human-computer interaction
- Language modeling (including both classical finite-state approaches as well as modern neural methods)
- CS curriculum development
From both perspectives, I also have a strong research interest in the societal and ethical implications that arise from speech and language technology, particularly regarding healthcare.
Assorted Ongoing Projects
I have a spectacular group of colleagues and collaborators, both at OHSU and at other partner institutions. Below are links to some of my ongoing or recently-wrapped projects.
- TREC-COVID
- Algorithmic classification of paraphasias (NIDCD, 5R01DC015999, Bedrick & Fergadiotis)
- Identifying and understanding drivers of selection bias and information bias in clinical COVID-19 data (NLM, 1R21LM013645, Weiskopf)
- Using Machine Learning to find a life saving needle in a haystack of children’s emergencies (NHLBI, 1R01HL161385, Guise)
- Voice as a Biomarker of Health: Building an ethically sourced, bioaccoustic database to understand disease like never before (NIH-OD, 1OT2OD032720, Bensoussan)
- Automated measurement of language outcomes for neurodevelopmental disorders (NIDCD, 2R01DC012033, Fombonne)
- Automatic voice-based assessment of language abilities (NIDCD, 5R01DC013996, Dolata)
- Assessing usability of Mhealth Autism screeners in the medically underserved (NIMH, 1R21MH120349, Zuckerman)
- Clinic interactions of a brain-computer interface for communication (NIDCD, 2R01DC009834, Fried-Oken)
- Semi-structured information retrieval in clinical text for cohort identification (NLM, 1R01LM011934, Liu)