The goal of this project is to assess operationally-relevant behavioral performance over 45-day isolation and confinement periods in the Human Exploration Research Analog (HERA), as well as associated neurophysiological status during this period. Operational performance is being evaluated using the ROBoT-r task—an operationally used track-and-capture task for grappling incoming resupply vehicles using Canadarm2. This task was modified for research use as part of the separate Behavioral Core Measures project. Neurophysiological assessments include resting-state connectivity and functional brain activation during the ROBoT-r task trials using our near-infrared spectroscopy and imaging (NIRS/NIRI) based NINscan devices.
In Year 3 of this project, the following tasks have been completed.
ROBoT-r v6.3 Software Deployment: While ROBoT-r v6.2 remained deployed in HERA for Campaign 4, v6.3 was deployed for Campaign 5. This upgrade included an important upgrade that prevents any accidental bumping of the hand controllers between trials from causing a trial abort, as well as a few minor modifications to the feedback displays provided after each trial. These display modifications were made based on discussions with ROBoT-r trainers to ensure consistency of feedback between research and operations use of the ROBoT-r task. In addition, at the start of HERA C5M2, a patch was deployed to speed up the recycling process between ROBoT-r runs to help maintain the task timing and schedule for ROBoT-r performance.
HERA Data Collection: In the past year, we completed data collection for HERA Campaign 4 Mission 5, HERA Campaign 5 Mission 1, and initiated data collection for HERA Campaign 5 Mission 2. In each case, Dr. Ivkovic traveled to Houston to confirm appropriate setup of the system and conducted all crew familiarization (with both ROBoT-r and NINscan-SE), all crew training, and all baseline data collection. ROBoT-r and NINscan-SE data was also validated (to confirm the appropriate data was being collected) prior to hatch-closing for each mission. Data collection included the ROBoT-r behavioral performance data and multi-modal NINscan brain and physiological data.
In addition, following the finalization of the HERA C4M5 crew, we began recruiting control subjects at Massachusetts General Hospital for HERA Campaign 4. To date, we have recruited n=10 control subjects, and completed running n=7, with no off-nominal data collection issues.
HERA Data Analysis: Analyses to date remain preliminary, given the limited number of control subjects. However, a number of features have been clearly identified.
• Weighted scores increase steadily and significantly throughout the ~60-day pre-/during-/post-HERA periods, representing improved accuracy at point of contact between Canadarm2 and the HTV-II vehicle. The proportion of successful captures also increases over this period.
• Duration to completing vehicle capture decreases steadily and significantly over this same period. Increased speed combined with the improved performance is a hallmark of learning, which appears to continue throughout the 60 day missions (which represent ~10-12 hours of hands-on ROBoT-r performance).
• Performance is significantly affected by run difficulty, with the highest difficulty resulting in significantly poorer and slower performance.
• There were notable and significant differences in overall performance across crews.
• In preliminary comparisons between HERA subjects and controls, we found controls performed slightly but significantly worse than HERA participants overall. Interestingly, controls performed better but slower in the high difficulty levels, suggesting a somewhat different strategy (slow and careful) compared to HERA subjects (faster but less accurate).
• Physiological data from NINscan demonstrated significant differences between HERA crews and controls in heart rate (HERA>Controls). Both groups exhibited changes in heart rate, as well as frontal pole and dorsolateral prefrontal brain activation within runs, suggesting progressive “activation” as the more challenging end of the run approached.
Given the limited number of controls, the above findings are preliminary and we cannot state any firm conclusions regarding the differences between HERA crewmembers and controls, for either behavioral performance or for physiological variables.
Dissemination: The results to date of ROBoT-r data collection were presented at the Human Research Program Investigator Workshop (HRP IWS) conference in Galveston, TX in late Jan 2019, and portions of this effort were included in the Master’s Thesis of Dr. Stijn Thoolen at Kings College, London.
Remainder of Year 3: In the remaining 2 months of grant year 3 we anticipate completing the following activities:
Data Collection: Data collection and support for C5M2 is underway and will continue through mid-July 2019. We will also continue to actively recruit additional control subjects, and steadily run those through this year and next.
Data Quality Control: We will conduct a detailed data quality-control assessment for all C4 datasets as well as for C5M1. This will include identifying more subtle anomalies (dropout, noise, erroneous values, etc.) in all datasets so that analysis programs can be robust to such data features. C5M2 will undergo the same process at the beginning of the next grant year.
Data Analysis: We will work to complete the interim analysis of our NINscan data during the remainder of this grant year, with analyses designed to address our Specific Aims.
|