Anticipating acute kidney problems that concerns to a large number of the world's population, Artificial Intelligence is implemented to maximizes prevention processes up to a couple of days before collapse.
The company deep mind, dedicated to the development of Artificial Intelligence (AI), is working with 26 physicians from the British National Health Service (NHS) to combat kidney problems. For this purpose, Streamsis used, an application that collects relevant medical information, processes it to detect some anomaly and prioritizes to act without delay in order to anticipate future diseases.
The algorithms working in this system yield the results for the corresponding clinics to develop a pre-diagnosis thanks to access to important data such as vital signs and medical history of patients.
The high-prestige journal, Nature, indicates that Streams is 55.8% effective and is able to predict kidney injuries 2 days before they appear.
Thanks to this AI, preventive work can be done, an also reducing the treatment costs in 17%.
Samples from 703,782 adults were taken for the study; with a database of more than 600,000 different indicators such as:
- Blood tests
- Authorized treatments
- Clinical history
The algorithm works based on the data that was introduced, it measures its danger and also makes a prediction of future drawbacks in the kidney; specialists could then anticipate up to two days before the disease appears and start using specific treatments such as certain intravenous and diuretic fluids.
The kidney is an indispensable organ that works on blood cleansing and, with help as the application of Artificial Intelligence of Streams, conventional treatments – such as dialysis – are expected to be reduced in use and cost. The system is also expected to be improved as it is at an early stage and there are still areas of intervention to be addressed in the future.
Some of these fields are the prediction of other diseases such as sepsis, liver problems, diabetes and perfecting tests to avoid false results. This will refine details in the algorithms and other technical configurations that will provide better tracking to user cases.
At the beginning of any artificial intelligence project, it is very important to have wide ranges so that false positives can be generated, but avoid at all costs cases where there is no detection of deficiency in the patient and their health is put at risk.
MOBILE HEALTH NEWS
SCIENTIFIC AMERICAN
https://www.scientificamerican.com/article/ai-can-predict-kidney-failure-days-in-advance/
NEXT WEB
