Basics of Artificial Intelligence!
We’ve heard a lot about COVID-19 in the news, it dominates our news cycles. And its impacted almost everyone on this planet in some way, shape or form. We have covered the COVID-19 pandemic in detail in our blogs, and we thought it might be refreshing to have a break, and look at advancements in medicine.
One area of innovation that interests me greatly is the use of AI in medicine. AI, or artificial intelligence, is a long sought after holy grail of technology, the ability to make an artificial sentient construct with the intellect and creativity of a human. Whilst that dream has not been realised yet, we have made huge advancements in computers emulating aspects of human behaviour and decision making.
In this blog series I will try to go over
· Basics of AI
· How AI can impact different aspects of medicine
· Recent advancements in AI
· Legal hurdles with AI in Medicine
To get started lets briefly cover some areas where artificial intelligence could help.
DETECTION
There are entire specialties of medicine dedicated to detecting patterns of disease. This is seen in radiology, where radiologists look at x-rays, CT scans, MRI scans and more to try and detect tell-tale signs of disease. Patchyness in the lung fields could suggest an infection, gas seen under the diaphragm a worrying sign of bowel perforation. But these can be, and sometimes are, missed in the hospital setting, even by highly experienced specialists utilising thorough diagnostic frameworks.
A trained AI programme is not liable to fatigue, and in some studies have been shown to be more accurate at finding pathologies in scans than consultant radiologists. Of course the key issue is training these programmes adequately to recognise the vast number of subtle signs for an ever expanding number of diseases.
DIAGNOSIS
Doctors will use a variety of frameworks to make a diagnosis. The majority of the diagnosis lies in the history, and it is the ability to recognise the key components in a history that makes specialists so valued. This skill is learnt in medical school but sharpened throughout a doctor’s training over many years, as they gain exposure to vast number of clinical cases for different diseases, and unusual presentations of the same disease.
How can AI help? Whilst AI programmes still have difficulty in parsing through non-categorised qualitative data that is present in natural speech, they are unmatched in analysing categorised data and performing relevant statistical analyses. This is applicable for quantifiable data such as heart rate, blood pressure blood results and so on. Quickly analysing relationships between these data sets can help diagnose a patient with a certain disease, determine if they are responding to treatments instigated by the medical team and highlight deteriorating patients to clinicians. Whilst doctors can diagnose and monitor patients depending on whether certain parameters increase or decrease in relation to other parameters, AI programmes will be able to analyse far more parameters instantaneously, whilst also running far more complex calculations that would be unfeasible for doctors to perform in a ward setting.
RESEARCH
There are also key gains to be made in a research setting. Currently the research involved in producing a drug to target a known disease is very laborious, time intensive and expensive. It can cost a company $359 million to develop a new drug from the research lab to a patient, and out of 5,000 drugs being researched, only 1 may see approval for human use.
Use of AI processes may help streamline drug discovery, by looking at and comparing drug targets (such as cell receptors| to known chemical compounds, or by repurposing existing approved medication to be used against new, novel targets. AI systems will be able to perform this task far more efficiently and across a much larger data set than human researchers, thus cutting costs and saving valuable time in finding new therapies.
In our next blog we will delve into the basics of AI, and the complex theories that sit behind the buzzwords thrown about in news and headlines.
Second Medic Inc
Dr Rajan Choudhary
Comments
Post a Comment