AI In Healthcare: Artificial Intelligence Can Heal The Healthcare Industry

 

This article was written by Julia Masch, a Financial Analyst at I Know First.

 

Summary

  • How Artificial Intelligence Can Be Used In Healthcare
  • Different Implementations of AI
  • I Know First Forays Into Predictive Healthcare

The popularity of artificial intelligence (AI) technology has been growing exponentially as more uses are explored and AI becomes a typical part of our lives, showcasing itself in both expected and unexpected places, such as fintech, smartphones, self-driving cars, and even the healthcare industry. There are multiple ways AI can be incorporated into the healthcare realm with potentially life-saving impacts.

AI is capable of converting massive amount of data that humans cannot productively parse through into useful information. There are many ways that this is already being utilized. For example, DNA is made up of around 3 billion combinations of 4 nucleotides compiled in 23 pairs of chromosomes. Cancer is most often a result of a mutation in a particular gene/chromosome that reproduces and spreads the mutation. While it is possible for a human for use a list of base pairs to find the cancerous mutation, it is a tedious process that takes hours upon hours of work and may still reap not be successful.

However, thanks to improved processing power, AI can take in a data set of patients with a particular type of cancer and those without it and compare their DNA sequences in order to identify potential cause of the cancer. This information can then be integrated into better healthcare solutions in one of two ways: drugs can be created that pinpoint this mutation and the cells that have it or people can be tested in advance to see if they have a particular mutation and eventually methods to combat reproduction of mutation can be created before cancer starts to spread.

The Institute of Cancer Research is already implementing AI into their research. Their canSAR database digests genetic and clinical data from patience and uses AI to target new cancer drugs. Other companies like Foundation Medicine Inc. have created products using AI such as the FoundationOne CDx which assesses genomic alterations known to instigate cancer growth. Additionally, FoundationOne can also be used as a broad treatment companion in particular cancers to identify targeted therapies and relevant clinical trials. Given enough data, AI can also build relationships between the various components of the body, cancer, and various treatment options and then assess a patient’s best treatment option.

This is an extremely important development for cancer patients as one needs to swiftly decide the best course of action once a diagnosis is given and patients are often overwhelmed by the myriad options presented. While an oncologist can provide their opinion and help to advise the patient, there so many miniscule factor that may affect how a person will react to each particular treatment that a doctor cannot predict. This implementation of AI should help to decrease the mortality rate of cancer and save numerous lives as there were over 8 million cancer related deaths worldwide in 2012 alone.

Cancer is not the only disease that afflicts humans that AI can help predict and combat. I Know First, a startup that uses a self-learning algorithm to predict the stock market, has recently partnered with KST Medical Group in order to adapt their algorithm to predicting cardiovascular disease in patients given their medical history, which will allow those susceptible to take more precautions. Eventually the algorithm will be able to predict the likelihood of cardiovascular events such as heart attacks.

There are so many more ways the future of healthcare will be impacted by the advancement of technology. Too many people know the horror of inputting their symptoms on WebMD, self-diagnosing themselves with some fatal disease, and then freaking out about it until they can see a doctor. However, AI in healthcare will be able to provide telemedicine – the use of technology to give clinical healthcare from any distance. AI will be able to examine users health and give almost immediate doctorate level recommendations via the internet. For example, Babylon Health is working on implementing this type of program and now has success rate 9% higher for passing the general practitioner exam compared to human doctors. This will make healthcare much more accessible to those cannot afford a doctor’s visit and can speed up the diagnosis process.

The presence of AI in the healthcare industry does not mean that doctors will be phased out completely. The growing healthcare partnership between machine and man is meant to assist the medical field in their work and make their job more effective. Researchers at Thomas Jefferson University Hospital have trained AI to detect tuberculosis in patients by examining over 1000 X-Rays of patients with and without tuberculosis. What is unique about machine-learning algorithms is that they are continuously adapting, thus as diseases evolve so does the output.

Self-learning algorithms incorporate past predictions into future predictions with a constantly running cycle. If it makes a mistake, it corrects itself and improves the current relationships to reflect the most current information. This is why more accessible data leads to greater accuracy in predictions.

Even large tech giants like Google are joining the movement to integrate AI in healthcare. Google’s DeepMind Health strives to help doctors, nurses, and patients. IBM is using its Jeopardy-winning Watson in a new program entitled WatsonPaths. This technology examines information from medical records and uses that to help physicians make more accurate decisions.

One of the primary expenses of pharmaceutical companies is research and development of new treatments and drugs. With the ability of AI to collect vast amounts of data to find patterns and make forecasts, the cost of medical research will significantly drop thus making pharmaceutical companies more profitable and increasing shareholder value. The worldwide pharmaceutical market revenue was over a trillion dollars in 2016; if expenses can be diminished, this number will only increase. Additionally, lower drug development costs can also lead to lower prices on the final product which will make medicine more accessible to the masses.

Clearly, there are numerous ways to apply artificial intelligence to the healthcare industry. The future of healthcare is not only to react once a diagnosis has been given, but to progress towards prediction and eventually prevention of disease. The most amazing thing is that as the technology continues to develop, new possibilities that we couldn’t even dream of before will become plausible and further advance the industry.