Skip to main content

Sample Size at Different Check-Up Time Points in Digital Data​

note

Below is a dynamically loaded bubble chart. Click each circle to view the summary statistics, and hover your mouse over the bar plot to see the numbers.

Summary​

The FHS Brain Aging Program has and continues to collect multiple digital data phenotypes that include:​

  • Brain MRI scans in their native form (e.g., DICOM images), collected since 1999
  • Digital voice recordings of the neuropsychological test protocol, collected since 2005
  • Digital pen recordings of written responses to the neuropsychological test protocol

Separately, through the NHLBI funded contract, FHS-BAP has assisted in the collection and post-collection processing of additional raw digital data files that include:​

  • Digital voice recordings of the cognitive screening tests (e.g., CERAD List Learning, Victoria Stroop) for the Gen 3/OmniGen 2/NOS participants who attended Health Exams 2, 3
  • Digital pen recordings of written responses from the cognitive screening test (e.g., MMSE) for the Gen 2/OmniGen 1 participants who attended Health Exam 10.
  • Digital pen recordings of written responses from the cognitive screening test (e.g., MoCA) are currently being collected as part of the Gen 3/OmniGen 2/NOS Health Exam 4

From other ancillary studies:​

  • Digital ECG data from SleepImage wearable, collected between 2016-2018
  • Cardiac MRI scans, collected from 2002-2006
  • Digital smartphone applications that include screen-based time-stamped features and digital voice recordings are currently being collected as an extension of the Gen 2/OmniGen 1 Health Exam 10 and also as part of the Gen 3/OmniGen 2/NOS Health Exam 4.

Combining these different studies under FHS has contributed to FHS standing unique from virtually any other large scale epidemiologic study in its longitudinal digital data resource. FHS has had a long history of collecting blood (as well as urine and brain tissue) and banking it for current and future use. Digital data adds to this, with the most important difference between blood and other biospecimens and digital data being that digital data is a non-diluting resource. This means it can be used and reused without reducing its size or quantity, and new insight can be gained from data collected decades ago. As such, analysis of digital data can embrace high amounts of processing that would inherently contain a large level of risk if conducted finite biosamples. This data is documented on privately and publicly available FHS resources, including this website.