Challenge
Develop a computational diagnostic clinical decision support system that can autonomously evaluate medical data from crew and consult medical guidelines to provide medical decision support without a connection to NASA. Coverage of the diagnoses outlined in the Medical Conditions list is required, as is other common medical conditions that may occur and require treatment during long-duration space flights.
TRISH is soliciting proposals for the rapid development of commercially available or viable computer-based algorithms that facilitate diagnosis of medical conditions on NASA’s Medical Conditions List in addition to other medical conditions that may occur and require treatment during long-duration space flights as well as for terrestrial applications.
Background
While astronauts are generally healthy, medical issues can occur during long-duration space flights as on Earth, and space flight increases the risks for certain medical complications, including muscle atrophy, bone demineralization, cardiovascular deconditioning, motion sickness, visual problems (perhaps related to elevated intracranial pressure), kidney stones (resulting in part from bone demineralization), back pain, urinary retention, toxic exposure, decompression, ear infection, and corneal abrasions (due to an excess of floating objects in-flight).
Definitive medical care is not feasible on spaceflight missions. In-flight medical expertise and supplies will be limited during long-duration space flights, and therefore definitive care may be weeks to months away. Additionally, long-duration missions will be limited in telemedicine support. Therefore, on-board medical diagnostic systems need to either provide user-guidance capabilities or minimize training time by being exceptionally intuitive and easy to use.
Limitations and related desirable attributes/specifications for use during long-duration space flights for an algorithmic tool for generating a differential diagnosis are shown in the following table:
Spaceflight Limitation |
Device Design Attributes/Specifications |
|
Multiuse devices, or devices integrated into the vehicle Components robust to failure in space environment |
Communication delays |
Easy-to-use devices Built-in self-guidance or just-in-time training systems Automated data management |
Limited power generation capabilities |
Minimize power demands |
Busy astronaut schedules |
Minimize time required for training, setup, use, and data management |
It can be assumed that the astronaut’s complete medical history will be available, as will vital signs, EKG, output from regular medical exams, possibly ultrasound imaging, and other physical signs as obtained onboard by a physician-astronaut or other crewmembers with less medical expertise and experience. In addition, the computational system should be able to refine the differential diagnosis based on the results of a complete blood count (CBC) with differential, basic metabolic panel, and liver function tests. No other medical diagnostic modality (such as X-rays or MRI scans) will be assumed to be available.
To accelerate clinical adoption of these technologies for use during long-duration space flights as well as terrestrial clinical settings, the system must be comprehensive, accurate, easy to use with minimal potential for operator error, and should facilitate therapeutic decisions, store results for later downloading and analysis, and connect and integrate with electronic medical record systems.