INS NYC 2024 Program

Poster	Poster Session 04 Program Schedule 02/15/2024 12:00 pm - 01:15 pm Room: Shubert Complex (Posters 1-60) Poster Session 04: Neuroimaging \| Neurostimulation/Neuromodulation \| Teleneuropsychology/Technology Final Abstract #54 Using In-Home MetaSensors to Sufficiently Track Daily Activities and Behaviors: A Case Study Regan Jenkins, Washington State University, Pullman, United States Maureen Schmitter-Edgecombe, Washington State University, Pullman, United States Category: Aging Keyword 1: activities of daily living Keyword 2: aging (normal) Keyword 3: technology Objective: Assessment of ability to perform daily tasks and live independently is often made through self-and caregiver-report, a process requiring extended time, effort, and insight into one’s own or another's daily routine. This case study examines whether MetaSensors placed on items that are part of an individual’s everyday routine (e.g., pill-holder) are effective for tracking everyday daily activities and detecting task completion. We hypothesized that sensor data would strongly associate with the self-report of the participant and would reveal inconsistencies in daily routine. Participants and Methods: Following a clinical interview about her daily routine, a 70+ year-old woman had item sensors placed on standard items related to basic- (toothpaste, hairbrush) and Instrumental- (pill organizer, coffee pot lid) activities of daily living, and general daily tasks the individual engaged in (meditation journal). One week of data collection occurred between two neurocognitive testing sessions. Each time a sensor detected motion (i.e., item being used), it registered a time-stamped data point. Data was uploaded to a secure testbed and then used to construct the participant’s morning, medication, and nighttime routines. The participant’s self-report of these routines was used as ground truth. Each of the 5 sensors was assigned a numerical value (1-5) and then placed ordinally for each day of the week. A value of 6 was used when no task was completed. The percentage of each event occurring throughout the week in all ordinal spots was calculated and used to order the routine. The routine was then correlated with the participant report. We also examined if important aspects of daily routine (e.g., medication) appeared to be missed. Results: According to the sensor-derived ordinal percentages, the average morning routine was taking medications (100% in first event), brewing coffee (71.43% in second event), daily meditation (42.86% in third event), and no task completed in steps 4 and 5 (57.14% and 71.43%, respectively). Compared to the participant’s reported routine of taking medications, brewing coffee, meditation, brushing teeth, and combing hair, we found a correlation of r=.96. Examination of sensor data to include the morning and afternoon, for a constructed average day matched exactly with the morning self-report. Her nighttime routine was less consistent because it appeared that she may not have always brushed her teeth at night (28.57%). The constructed medication routine was a 100% match to the self-report. Conclusions: Although future research is needed, the data suggest that MetaSensors can be used as a stand-alone tool to measure and analyze daily patterns and events within the real-world environment. Unlike Mem caps, the MetaSensors allowed for tagging of her entire pill organizer. Data were limited as this was a case study, certain movements may not have been consistently detected (e.g., tooth brushing), and some aspects of her routine could not be captured with sensors (e.g., length of time brushing teeth due to the sensor on toothpaste instead of toothbrush, actual consumption of coffee). These sensors appear to be a promising method for future assessment and observation within the real-world environment.

Poster

02/15/2024
12:00 pm - 01:15 pm
Room: Shubert Complex (Posters 1-60)

Poster Session 04: Neuroimaging | Neurostimulation/Neuromodulation | Teleneuropsychology/Technology

Final Abstract #54

Using In-Home MetaSensors to Sufficiently Track Daily Activities and Behaviors: A Case Study

Regan Jenkins, Washington State University, Pullman, United States
Maureen Schmitter-Edgecombe, Washington State University, Pullman, United States

Category: Aging

Keyword 1: activities of daily living
Keyword 2: aging (normal)
Keyword 3: technology

Objective:

Assessment of ability to perform daily tasks and live independently is often made through self-and caregiver-report, a process requiring extended time, effort, and insight into one’s own or another's daily routine. This case study examines whether MetaSensors placed on items that are part of an individual’s everyday routine (e.g., pill-holder) are effective for tracking everyday daily activities and detecting task completion. We hypothesized that sensor data would strongly associate with the self-report of the participant and would reveal inconsistencies in daily routine.

Participants and Methods:

Following a clinical interview about her daily routine, a 70+ year-old woman had item sensors placed on standard items related to basic- (toothpaste, hairbrush) and Instrumental- (pill organizer, coffee pot lid) activities of daily living, and general daily tasks the individual engaged in (meditation journal). One week of data collection occurred between two neurocognitive testing sessions. Each time a sensor detected motion (i.e., item being used), it registered a time-stamped data point. Data was uploaded to a secure testbed and then used to construct the participant’s morning, medication, and nighttime routines. The participant’s self-report of these routines was used as ground truth. Each of the 5 sensors was assigned a numerical value (1-5) and then placed ordinally for each day of the week. A value of 6 was used when no task was completed. The percentage of each event occurring throughout the week in all ordinal spots was calculated and used to order the routine. The routine was then correlated with the participant report. We also examined if important aspects of daily routine (e.g., medication) appeared to be missed.

Results:

According to the sensor-derived ordinal percentages, the average morning routine was taking medications (100% in first event), brewing coffee (71.43% in second event), daily meditation (42.86% in third event), and no task completed in steps 4 and 5 (57.14% and 71.43%, respectively). Compared to the participant’s reported routine of taking medications, brewing coffee, meditation, brushing teeth, and combing hair, we found a correlation of r=.96. Examination of sensor data to include the morning and afternoon, for a constructed average day matched exactly with the morning self-report. Her nighttime routine was less consistent because it appeared that she may not have always brushed her teeth at night (28.57%). The constructed medication routine was a 100% match to the self-report.

Conclusions:

Although future research is needed, the data suggest that MetaSensors can be used as a stand-alone tool to measure and analyze daily patterns and events within the real-world environment. Unlike Mem caps, the MetaSensors allowed for tagging of her entire pill organizer. Data were limited as this was a case study, certain movements may not have been consistently detected (e.g., tooth brushing), and some aspects of her routine could not be captured with sensors (e.g., length of time brushing teeth due to the sensor on toothpaste instead of toothbrush, actual consumption of coffee). These sensors appear to be a promising method for future assessment and observation within the real-world environment.