Hospitals began to apply the AI system of IBM revealing risks of sepsis at patients

In the middle of September, 2019 the network medical organizations Geisinger IBM was also announced start new predictive AI systems which will help doctors to estimate risk of sepsis on the basis of data of the electronic medical record.

Sepsis is potentially life-threatening status, but difficult for diagnostics on initial stages as its symptoms are too not specific and can point to other disease. Sepsis becomes the reason of more than 250,000 fatal cases annually. The new algorithm created using IBM Data Science Elite will help to create Geisinger companies the personalized plans of the clinical help for patients with sepsis and also will help the persons performing leaving to pay attention to the key factors connected with risk of death at this status.

Geisinger and IBM announced start of a new predictive AI system which will help doctors to estimate risk of sepsis on the basis of data of the electronic medical record

Researchers of IBM Data Science and AI Elite gathered a team of six people who developed model for forecasting of death from sepsis and also the tool allowing a command to be aware of the last researches in this area. Scientists used technology open source from IBM Watson for creation of predictive model which analyzed clinical data of several thousand patients with sepsis.

The new algorithm helped researchers to reveal the clinical biomarkers connected with the highest rates of death and to use them for the forecast of survival of patients during hospitalization and within 90 days after the statement. Besides, within the project descriptive and clinical features which can indicate the increased risk of development of sepsis, such as age, the previous diagnosis of cancer, a lowering of arterial pressure, the number of hospitalization, start time of action of vazopressor and even type of a pathogen were revealed.^[1]

Notes