Most of us thought that we were safe and socially secure, until one small particle came to destroy the world we knew. Even though it's not the first epidemic challenge in the last 15 years, this one has hit us the hardest.
If we put the grim question of the lives lost aside, try to be more positive and learn the lessons – what would those lessons be?
Almost none of the world's healthcare systems managed to pass this stress test.
From the US to Europe and Asia, they failed to pass the crush test. Even though we have a lot of data and modern equipment, mortality rates were still too high, and hospital logistics were far from being ideal, especially in the area of supply chains.
Tired doctors obviously made mistakes, as we all are human beings and can't keep the same level of focus during such a time.
How can AI help to improve the situation and eliminate the gaps we all found?
Healthcare is probably the largest sector in generating data, but are we really capable of using it for the better if, based on a PwC report, over 80% of this data remain unstructured? In reality, when facing a specific case with a patient, a doctor can't check even 1% of existing data relevant for the case. The only option is to follow standard protocols, which are not always ideal for specific cases. If we manage to include AI applications into daily workflow, we'll get the significant advantage from available data.
One of the cases we are researching now is an MRI analytics tool for diagnostics of stroke in the acute phase. Simple benchmark shows that that system is capable of analyzing an MRI in 20 seconds against 15 minutes for human-operated analysis. At the same time, the trained neural network and powerful analytics tool is capable of giving treatment recommendations, tips and projections of a patient's condition in another couple of minutes. If we consider the matter of time, which is critical for patients in the acute phase of stroke, then we know how important a 10 times better diagnostic window would be.
The same could be applied to CT and MRI monitoring of lungs, and simulations for learning, assessment and research purposes. This means, in the end, quicker and more valuable in-patient care with lower costs and better health outcomes.
If you would like to know more about our healthcare expertise, just drop us a line.