TY - GEN
T1 - Role-specific Language Models for Processing Recorded Neuropsychological Exams
AU - Hanai, Tuka
AU - Au, Rhoda
AU - Glass, James
N1 - Funding Information:
The authors thank David Stuck, Maggie San-doval, Alessio Signorini, and Christine Lemke from Evidation Health Inc., and Ida Xu, Brynna Wasserman, Maulika Kohli, Nancy Heard-Costa, Yulin Liu, Karen Mutalik, Mia Lavallee, Christina Nowak, Alvin Ang, and Spencer Hardy from Boston University and The Framingham Heart Study. Data curation was sponsored by DARPA and NIH grants AG016495, AG008122, AG033040. T. Alhanai thanks the Abu Dhabi Education Council for sponsoring her studies.
Publisher Copyright:
© 2018 Association for Computational Linguistics.
PY - 2018/6/1
Y1 - 2018/6/1
N2 - Neuropsychological examinations are an important screening tool for the presence of cognitive conditions (e.g. Alzheimer’s, Parkinson’s Disease), and require a trained tester to conduct the exam through spoken interactions with the subject. While audio is relatively easy to record, it remains a challenge to automatically diarize (who spoke when?), decode (what did they say?), and assess a subject’s cognitive health. This paper demonstrates a method to determine the cognitive health (impaired or not) of 92 subjects, from audio that was diarized using an automatic speech recognition system trained on TED talks and on the structured language used by testers and subjects. Using leave-one-out cross validation and logistic regression modeling we show that even with noisily decoded data (81% WER) we can still perform accurate enough diarization (0.02% confusion rate) to determine the cognitive state of a subject (0.76 AUC).
AB - Neuropsychological examinations are an important screening tool for the presence of cognitive conditions (e.g. Alzheimer’s, Parkinson’s Disease), and require a trained tester to conduct the exam through spoken interactions with the subject. While audio is relatively easy to record, it remains a challenge to automatically diarize (who spoke when?), decode (what did they say?), and assess a subject’s cognitive health. This paper demonstrates a method to determine the cognitive health (impaired or not) of 92 subjects, from audio that was diarized using an automatic speech recognition system trained on TED talks and on the structured language used by testers and subjects. Using leave-one-out cross validation and logistic regression modeling we show that even with noisily decoded data (81% WER) we can still perform accurate enough diarization (0.02% confusion rate) to determine the cognitive state of a subject (0.76 AUC).
UR - http://www.scopus.com/inward/record.url?scp=85083509969&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85083509969&partnerID=8YFLogxK
U2 - 10.18653/v1/N18-2117
DO - 10.18653/v1/N18-2117
M3 - Conference contribution
T3 - NAACL HLT 2018 - 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies - Proceedings of the Conference
SP - 746
EP - 752
BT - Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 2 (Short Papers)
PB - Association for Computational Linguistics (ACL)
ER -