Dementia Diagnosis Classification Using Deep Learning Neural Networks Based on Clock Drawing Test (CDT) and Medicare claims

Mengyao Hu
Adjunct Assistant Research Scientist, Survey Research Center, Institute for Social Research

Co-Investigators

  • Yi Lu Murphey, University of Michigan-Dearborn

Abstract

This pilot study will develop advanced deep learning neural networks to analyze Clock-Drawing Test images to predict dementia diagnosis. The pilot will draw upon Medicare claims linked with a large, publicly available repository of clock images from the 2011-2019 National Health and Aging Trends Study, a panel study of Medicare beneficiaries ages 65 and older. 

Outcomes

  • Hu, M., Murphey, Y.L., Wang, S., Qin T., Zhao Z., Gonzalez, R., Freedman V. A. & Zahodne L. (July 2023) Exploring the Use of Deep Learning Neural Networks to Improve Dementia Detection: Automating Coding of the Clock-Drawing Test. 2023 European Survey Research Conference, Milan, Italy.
  • Hu, M., Murphey, Y.L., Wang, S., Qin T., Zhao Z., Gonzalez, R., Freedman V. A. & Zahodne L. (May 2023) Exploring the Use of Machine Learning to Automate the Coding of the Clock-Drawing Test. 78th American Association for Public Opinion Research Annual Conference. Philadelphia.
  • Hu, M., Murphey, Y.L., Wang, S., Qin T., Zhao Z., Gonzalez, R., Freedman V. A. & Zahodne L. (Nov 2022) Exploring the Use of Deep Learning Neural Networks to Improve Dementia Detection: Automating Coding of the Clock-Drawing Test. Gerontological Society of America (GSA) Annual Scientific Meetings 2022, Indianapolis.
  • Hu, Mengyao; Murphey, Yi Lu; Wang, Song Qin Tian; Zhao Zixuan; Gonzalez, Richard; Freedman Vicki A.; Zahodne Laura (2022) Exploring the Use of Deep Learning Neural Networks to Improve Dementia Detection: Automating Coding of the Clock-Drawing Test. NHATS/NSOC Research in Progress Seminar, Ann Arbor, Michigan.