Hey, play improvement neighborhood,
I’m engaged on a medical care simulation sport that integrates AI in actual time to offer dynamic responses from the affected person primarily based on person entry. The target is to create a sexy and academic expertise the place digital sufferers pushed by AI realistically react to completely different therapies and interventions.
Key challenges: OPTIMIZATION OF IA for giant -scale simulations: What’s one of the best ways to deal with a number of entities of AI promoted with out overloading CPU/GPU assets? Automated studying for the adaptive habits of the affected person: Has anybody applied reinforcement studying or predictive evaluation in NPC’s habits to simulate actual world medical eventualities? Discount of latency in actual -time determination making: How can I be certain that determination -making promoted by AI stays speedy and receptive, particularly in multiplayer or cloud -based simulations?
Integration with medical care information (FHIR/HL7 requirements): Are there any really helpful methodology to combine actual affected person (anonymity) information into simulations to enhance realism? Right here is an instance of how I’m contemplating optimizing the decision-making of unity and ml-agents:
of mlagents_envs.surroundings import unityenvironment
ENV = UNITENVIRONMENT (file_name = “Healthcaresim.exe”)
Env.reset ()
For _ in vary (1000):
Decision_Step, Terminal_Step = Env.get_steps (“Healthcareagent”)
ACTIONS = MODEL.PREDICT (DECISION_STEP.OBS) # ai the mannequin predicts the following step
SEV
MV
env.shut ()
I might like to hearken to any concept of ​​builders who’ve labored in AI pushed simulations, particularly within the well being area. As well as, if you’re searching for knowledgeable help within the improvement of medical care functions with AI meals, see Well being Software program Improvement Providers for options appropriate with the business.
We count on your ideas and proposals!
