Dynamic Digital Biomarkers of Motor and Cognitive Function in Parkinson’s Disease

This protocol can help us stratify the heterogeneous Parkinson’s disease by cognitive and motor severity. Using wearable sensors, we can digitize clinical tasks that are commonly used by physicians. By co-registering activities across different functional levels of the nervous systems, we identify highly informative parameters of cognition, motion, and emotion, to produce dynamic digital biomarkers of Parkinson’s disease.

This work offers a unifying framework to characterize and track the signatures of neurological disorders in a personalized manner, along with the evolution of an entire cohort receiving a given treatment or therapy. This technique provides provides a unifying statistical platform to dynamically track changes in the signatures of statistical variability produced by self-generated biological activities spanning from molecules, to cells, to behavior. Streaming multiple software requires a lot of computer memory, so we would advise using computers with high computation capacity, otherwise systems may crash while they are recording.

We use multiple devices to record the brain, body, and the heart, all in tandem, during natural movements, so a dynamic medium, such as video, would best illustrate this. Demonstrating the procedures from my lab will be the graduate students Jihye Ryu and Joseph Vero. After obtaining informed consents from the participant, measure the participant’s body dimensions to allow creation of their body avatar in the motion-capture system.

Next, set up the motion-capture system, including the seventeen wireless motion-tracking sensors, and the motion-tracking software. Place sensors on all of the body parts, as indicated, and secure the sensors with strap bands, to allow for unimpeded movement. When all of the sensors have been placed, calibrate the participant’s position to create the avatar.

To set up the EEG device and recording software, position thirty-one channel sensors across the scalp, and place the recording device on the back of the head. Attach the last channel sensor to a connector to measure the heart signal, and position the sensor on the left side of the participant’s stomach. Then, attach to reference channel sensors behind the participant’s left ear, and use a syringe to insert electrode gel into the sensors on the EEG cap.

To capture the participant’s voice during the evaluations, place a microphone in front of the participant, and connect the microphone to the computer on which the lab stream layer will be running. Next, to set up the lab stream layers system to synchronize the streams of EEG, motion, audio, and mouse-click timestamps, open the lab recorder app, open and link the lab stream layer apps for mouse and audio capture win app, and then open an in-house built xsense synchronizing app. In the record from stream section of the lab recorder app, press Update’to get a full list of streaming items, and check the boxes for AudioCaptureWin’EEG’MouseButtons’Position’and TrackerKinematics’to link the audio, EEG, mouse clicks, and motion through the lab stream layer system.

Set up the recording for capturing the pen movement, including the pen tablet and the movement analysis software, and place the drawing tablet and tablet pen in front of the participant. Connect the tablet to the computer on which the movement analysis software will be recorded, and tape a piece of white paper onto the tablet. Then, press record in the lab stream layer, the motion-capture software, and the EEG recording software.

At the start and end of each task, click the time stamping button on the motion-capture software to timestamp the task. Don’t forget to write down the actual time when you click the time stamping button as a backup, so when you look at the data later, you can identify the task from the timestamps. For an immediate Benson complex figure copy, instruct the participant to copy the Benson figure on the paper, and to remember the design, because they will be asked to draw the design again from memory later.

For part A of a trail-making test, instruct the participant to draw a line between circles that are numbered in ascending order. To conduct part B of the trail-making test, instruct to draw a line between circles that contain either numbers or letters in ascending order, while alternating between the numbers and letters. For a clock-drawing task, instruct the participant to draw an analog clock with the numbers one through twelve, and to set the time to ten past eleven.

To conduct a delayed Benson complex figure copy, instruct the participant to draw the Benson complex figure from memory on a blank piece of paper. To conduct a forward numbers fan test, instruct the participant to repeat the numbers that the experimenter reads out loud in the same order. For a backward numbers fan test, instruct the participant to repeat the numbers that the experimenter reads out loud in the reverse order.

For a pointing task, position a target in front the participant to point and touch, and instruct the participant to point at the target forty times in a self-paced manner with the dominant hand. For a metronome pointing task, instruct the participant to point at the target forty times at a self-paced manner, while setting a metronome to beat at thirty-five beats per minute in the background, but do not instruct anything about the metronome beat. For a paced pointing task, instruct the participant to point at the target forty times following the pace of the metronome beat set at thirty-five beats per minute.

To conduct a walking task, first instruct the participant to walk naturally around the room for five minutes. Next, instruct the participant to walk naturally around the room while setting the metronome to beat in the background for twelve beats per minute. Then, instruct the participant to walk naturally around the room while pacing their breathing rate to the metronome beat set at twelve beats per minute.

For a face video, first, instruct the participant to sit comfortably and set up a camera in front of the participant. For the control assessment, instruct the participant to stare at a space without any stimuli for five minutes. For the smile assessment, instruct the participant to watch a funny video for five minutes.

In each of the drawing tasks, the patients in this representative study stratified apart from the controls, differentiating their individual stochastic signatures of motor variability according to the Movement Disorder Society unified Parkinson’s disease ratings scale medium ranked scores. A task of pointing to a spacial target is performed to assess the different levels of volitional control, and this figure illustrates the degradation of the center of mass trajectory as the participant severity of the disorder increases. Using this assay, it is possible to distinguish each subtype of patient with Parkinson’s disease, and to track changes in the patient’s stochastic signatures from baseline to metronome-associated pointing tasks of both spontaneous and uninstructed and deliberate and instructed cases.

In an automatic walking evaluation, the center of mass trajectories for controls and patients with Parkinson’s disease, as medium-ranked by the Movement Disorders Society unified Parkinson’s disease ratings scale scores can be obtained. Stochastic analysis of the walking task can then be performed, for which differentiations can be made between patients and controls, and among patients with different severities of the disease. Open pose software can be used to ascertain areas of the face that are most active during a given task, and to probe the emotional content by ascertaining area transitions across emotions during tasks.

Integrating digital biophysical signals for the EEG, magnotometer motion, and EKG, by using informational theoretical measurements and network connectivity analyses, it is possible to differentiate between patients and controls, and among patients with different severity, by examining the density of network. Using the digitized data of the brain a body signals, we can apply a variety of analytics, including cross-correlation and cross-coherency to characterize the interactivity between different modes of biosignals. Our lab has extended these methods to characterize other neurological and behavioral disorders, such as pain, and to study embodied cognition, and all of these address different questions.