Christopher Gallo | Engineering Papers

NASA NTRSDate not supplied

Artificial Neural Networks to Predict Cognitive Impairment of Rodents Subjected to Space Radiation

INTRODUCTION We use artificial neural networks (ANNs) as an example machine learning (ML) tool to predict the cognitive performance impairment of rats induced by irradiation. The experimental data in the analyses is attentional set-shifting (ATSET) test scores from a rodent model exposed to ≤15 cGy of individual galactic cosmic radiation (GCR) ions: 4He, 28Si, or 56Fe, expected for a Lunar or Mars mission [1]. This work investigates rats at a subject-based level and uses applied dose and performance scores taken before irradiation to predict whether a rat will be impaired when irradiated. The results of this study are significant to crewed space missions as they support the potential of predicting an astronaut’s impairment in a specific task before spaceflight through the implementation of appropriately trained ML tools. METHODS Data used in this work are scores from the ATSET, a multi-stage constrained cognitive flexibility test [2]. Our computational model utilizes the number of attempts to reach the criterion to pass a stage as a behavioral performance measure for rats. We use the post-irradiation scores, generate thresholds from cumulative distribution plots of non-irradiated rats, and calculate the percent of irradiated rats whose scores fall below the threshold to infer how each radiation type/dose affects a population. Rats scoring above the threshold are labeled impaired while the others are non-impaired. We then employ ANNs as a typical ML technique, and use each subject’s individual scores taken before radiation along with the applied dose, to predict their personal susceptibility to cognitive impairment due to space radiation exposure. RESULTS AND CONCLUSION A significant finding is the exhibition of a dose-dependent increasing probability of impairment for 1 to 10 cGy of 28Si or 56Fe in the simple discrimination (SD) stage of the ATSET, and for 1 to 10 cGy of 56Fe in the compound discrimination (CD) stage. On a subject-based level, implementing ML classifiers such as ANNs identifies rats that have a higher tendency for impairment after GCR exposure [1]. The receiver operating characteristic (ROC) and the precision-recall (PR) curves of the ML models show a better prediction of impairment when 56Fe is the ion in question in both SD (Figure 1) and CD stages. They, however, do not depict impairment due to 4He in SD (Figure 1) and 28Si in CD, suggesting no dose-dependent impairment response in these cases. In this work, “good” prediction pertains to “better-than-random-chance”, due to the limited sample size and the high inter- and intra-individual variabilities in response to brain stimulation paradigms, as applicable to both animals and humans. More behavioral tests and biomarkers should be investigated on the same subjects, to be fed to the ML models to capture the agents responsible for performance alterations of some individuals versus others.

machine learning↗

NASA NTRSDate not supplied

Machine Learning Models to Predict Cognitive Impairment of Rodents Subjected to Space Radiation

INTRODUCTION We use artificial neural networks (ANNs) as an example machine learning (ML) tool to predict the cognitive performance impairment of rats induced by irradiation. The experimental data in the analyses is attentional set-shifting (ATSET) test scores from a rodent model exposed to ≤15 cGy of individual galactic cosmic radiation (GCR) ions: 4He, 28Si, or 56Fe, expected for a Lunar or Mars mission [1]. This work investigates rats at a subject-based level and uses applied dose and performance scores taken before irradiation to predict whether a rat will be impaired when irradiated. The results of this study are significant to crewed space missions as they support the potential of predicting an astronaut’s impairment in a specific task before spaceflight through the implementation of appropriately trained ML tools. METHODS Data used in this work are scores from the ATSET, a multi-stage constrained cognitive flexibility test [2]. Our computational model utilizes the number of attempts to reach the criterion to pass a stage as a behavioral performance measure for rats. We use the post-irradiation scores, generate thresholds from cumulative distribution plots of non-irradiated rats, and calculate the percent of irradiated rats whose scores fall below the threshold to infer how each radiation type/dose affects a population. Rats scoring above the threshold are labeled impaired while the others are non-impaired. We then employ ANNs as a typical ML technique, and use each subject’s individual scores taken before radiation along with the applied dose, to predict their personal susceptibility to cognitive impairment due to space radiation exposure. RESULTS AND CONCLUSION A significant finding is the exhibition of a dose-dependent increasing probability of impairment for 1 to 10 cGy of 28Si or 56Fe in the simple discrimination (SD) stage of the ATSET, and for 1 to 10 cGy of 56Fe in the compound discrimination (CD) stage. On a subject-based level, implementing ML classifiers such as ANNs identifies rats that have a higher tendency for impairment after GCR exposure [1]. The receiver operating characteristic (ROC) and the precision-recall (PR) curves of the ML models show a better prediction of impairment when 56Fe is the ion in question in both SD (Figure 1) and CD stages. They, however, do not depict impairment due to 4He in SD (Figure 1) and 28Si in CD, suggesting no dose-dependent impairment response in these cases. In this work, “good” prediction pertains to “better-than-random-chance”, due to the limited sample size and the high inter- and intra-individual variabilities in response to brain stimulation paradigms, as applicable to both animals and humans. More behavioral tests and biomarkers should be investigated on the same subjects, to be fed to the ML models to capture the agents responsible for performance alterations of some individuals versus others.

machine learning↗

NASA NTRSDate not supplied

Assessing Long-Term Health (LTH) Outcomes in Astronauts

NASA has linked potential Long-Term Health (LTH) risks of astronauts to their spaceflight experience, including the effects of space radiation, microgravity, and stressors such as isolation, sleep deprivation, and stress. Therefore, assessing the LTH outcomes for astronauts exposed to these hazards becomes a critical investigation in ensuring any deleterious effects are minimal. Prior research conducted into LTH outcomes of astronauts has not revealed increases in non-accidental mortality rates when compared to the U.S. general population, analog healthy cohorts, and professional athletes. One study did report an increased incidence of non-fatal cardiovascular disease events (Hazard Ratio=2.41, 95% Confidence Interval 1.26 to 4.63). Additional evidence indicates that astronauts may have increases in incidences of melanoma, prostate, and hematologic cancers, though the former is consistent with rates observed among aircraft pilots and the latter two are likely caused by detection bias due to increased screening on former astronauts.

long term health↗

NASA NTRSDate not supplied

Proof-of-Concept for a Long-Term Health Metric to Quantify End of Mission Health Status in Astronauts

NASA has long used Probabilistic Risk Assessment (PRA) when high-stakes decisions need to be made about complex systems. For spaceflight medical risk, the Human Research Program’s Medical Extensible Dynamic Probabilistic Risk Assessment Tool (MEDPRAT) is a significant step towards robustly quantifying the risk to crew health during exploration missions. However, there remains a significant gap in the ability to comprehensively characterize and assess risk across the disparate functionalities and capabilities which comprise the entire Crew Health and Performance (CHP) system. To fill this gap, the Crew Health and Performance – Probabilistic Risk Assessment (CHP-PRA) project aims to perform risk characterization for the CHP system by assessing performance risk in addition to medical risk. This effort also includes quantifying Long-Term Health (LTH) risk in addition to in-mission risk outcomes within the CHP-PRA results. LTH risk encompasses the timeframe from immediately post-flight, through the rest of an astronaut’s career, through retirement, and until death. A proof-of-concept LTH risk metric is based on medical condition end-state, as defined by the Evidence Library, capturing the spaceflight specific medical impacts persisting into post-flight[1]. Condition outcomes in the Evidence Library progress through three Clinical Phases (CP): the diagnostic phase (CP1), the treatment/convalescent phase (CP2), and the end-state phase (CP3) which represents the detrimental effects of the condition after the crew member has recovered to the maximal extent. Each CP has an associated Task Impairment (TI), defined as the degree of crew incapacity due to experiencing the condition, and is quantified with a 0-1 range. Conditions with an associated CP3 (e.g. Sepsis, Traumatic Hypovolemic Shock, Sudden Cardiac Arrest, etc.) typically have serious consequences that can cause an astronaut to be fully or partially debilitated throughout the remainder of the mission. Consequently, the Cumulative CP3 TI End-of-Mission Health Status Metric is developed by CHP-PRA to quantify the cumulative effects of all conditions which progressed to the CP3 state throughout the entirety of the mission. Hence, this End-of-Mission Health Status Metric attempts to serve as an indicator of an astronaut’s health state at the time of landing. The severity of the lingering effects of in-mission medical events are dependent on mission activities and the level of available in-mission medical care. This allows the associated cumulative TI metric to be used in comparison with the crew’s end of mission health status for different levels of in-mission resources. This presentation provides the strategy for using CP3 as an LTH metric component, as well as a proof-of-concept demonstration of LTH risk characterization using this component.

long term health↗

NASA NTRSDate not supplied

Performance Risk Model (PRisM) Proof-of-Concept: An Operational Decision Support Tool to Predict Crew Performance in Space from Available Performance Tests

The Crew Health and Performance-Probabilistic Risk Assessment (CHP-PRA) team at NASA Glenn Research Center has developed a range of tools to evaluate astronaut health during spaceflight and to optimize the medical set required for missions. Among these, the Performance Risk Model (PRisM) represents a novel advancement, extending CHP-PRA’s focus beyond medical systems into the domain of human performance. Such tool could be pivotal in optimizing astronaut capabilities during space travel, thereby enhancing overall mission success. PRisM leverages data from well-established performance assessments conducted during previous crewed space missions and Earth analogs to strategically predict outcomes for planned tasks, even when direct performance testing has not been conducted for those specific tasks. To evaluate performance, PRisM references the various metrics outlined in NASA-STD-3001 and employs a methodology to integrate different performance scales. This framework analyzes the contribution of various human system task categories (HSTCs) to task execution and compares these contributions to the HSTCs’ involvement in other known tests. The model further employs a Monte Carlo simulation to sample performance scores from their distribution in operationally relevant tests such as those in Mulavara et al. (2018) and, by leveraging similarities in HSTC involvement, transfers this knowledge to actual mission tasks, such as those outlined in the "Human Exploration of Mars: Preliminary List of Crew Tasks”. The current PRisM proof-of-concept includes analyses of the impact of exercise and specific medications on astronauts’ performance, with provisions to incorporate additional countermeasures as data becomes available. Furthermore, the tool is customizable to include any system necessary to fully encompass the domain of human systems and can be adapted to evaluate performance for any spaceflight activities as requested by operational stakeholders. PRisM has the potential to assist the Human Research Program in exploring the capabilities trade space for optimized crew performance.

performance modeling↗

NASA NTRSDate not supplied

Performance Risk Model Validation with Operationally Relevant Tasks

Human Research Program aims to develop methods to support astronauts’ health and productivity during spaceflight. The Crew Health and Performance Probabilistic Risk Assessment (CHP-PRA) team uses powerful computational methods to predict mission risk in both domains: medical and performance. Here, we show how CHP-PRA uses the Performance Risk Model (PRisM) to quantify the performance risk and show an application of the model on operationally relevant tasks. There are various metrics adopted across performance researchers that PRisM can accommodate. For data analysis, interpretation, and integration, we use a method of unifying data from multiple sources by converting each to a single metric. We consult subject matter experts prior to integrating the converted data into PRisM. The method we use is inspired by the Cooper-Harper rating scale [1]. Using this unified metric, we can easily combine data from various tests and lab groups. We explain our conversion method in detail and show how it pertains to the process of testing and validation of PRisM on operational tasks. We conducted an initial validation in collaboration with the Behavioral Health and Performance (BHP) lab. We test PRisM using data on their operationally relevant task ROBoT-r, a track-and-capture task for grappling incoming resupply vehicles [2]. Several other labs at NASA Johnson Space Center worked together to design 7 Functional Task Tests (FTTs) in pursuit of simulating the tasks required after landing on a planetary surface and after return to Earth [3]. Here we use the results from both ROBoT-r and the 7 FTTs and compare their experiment data to PRisM’s computational output to demonstrate how PRisM can support operations by predicting crew performance on future missions.

performance modeling↗