LMM


Our Large Medical Model (LMM) is a generative healthcare AI that uses deep learning to predict future outcomes for individuals and populations, pre-trained on over a million patients' claims and clinical FHIR data.


Simulate patient futures

LMM (patent pending) can generate future paths for a patient, condition, procedure, or their combinations. Aggregated paths can probabilistically shed light on a patient or population's journey.

cohort definition (patient history includes)

female

male

patients

30

50

70

year old
with

stroke

colon cancer

heart failure

undergoing a

physical

heart bypass

knee replacement

LMM (frequency of predictions over next year)


Risk calculation

LMM can calculate the risk for how quickly a disease progresses for a patient or population.

cohort definition (patient history includes)

Male patients who are 67 years old with obesity with a treatment plan starting with

metformin

gastric bypass surgery

liraglutide (semaglutide)

diet / exercise (counseling)

LMM (frequency of predictions over next year)

⬤ All other events⬤ Diabetic or Kidney events



Cost of care

LMM can identify futures impacted by medical decisions. The monetary impact of future events can be aggregated to understand the economics of a care decision.

cohort definition (patient history includes)

59 year old female with osteoarthritis for the hip who is

going in for a

physical

hip replacement

LMM (frequency of predictions over next year)

Quality measurement and readmissions

LMM can can calculate the quality of care via readmissions or other events indicative of positive healthcare outcomes.

cohort definition (patient history includes)

71 year old female patient with congestive heart failure and a treatment plan beginning with the following:

a primary care appointment

furosemide

angiotensin receptor blocker (losartan)

implantable cardiac defibrillator

LMM (frequency of predictions over next year)

⬤ All other events⬤ Reoccurrence of Heart Failure