Artificial Intelligence Defining the Future of Healthcare
Healthcare has been at the top of the priority list for countries & corporations for a long time. But what changed in the last decade? More importantly, what changed in the last year?
The answer to these questions is ‘Artificial Intelligence’. AI has seen increased involvement in healthcare for the past three decades. But let us consider the developments in the last few years only.
The basic function of AI is to duplicate human cognitive functions. Heavy investment in Artificial Intelligence has become the norm of the Venture Capital domain and has amassed significant interest from the likes of Google, Apple, IBM & Amazon.
Google has acquired 30 startups totalling close to $4 billion in valuation since the 2008 economic depression, whereas Apple has 18 to its name, totalling close to $900 million in the same duration. Google’s Deep Mind also joined the Google Health family in 2019 with an aim to improve patient care.
The advantages of AI have been a subject of extensive discussion in the medical communities & literature. Essentially AI uses algorithms that are sophisticated enough to learn from volumes of healthcare data from networks and platforms that it has access to. In fact, the biggest enabler of AI- the internet- was ARPANET, a US Department of Defence project meant to help scientists share their research & data on a network of computers. AI can use sophisticated algorithms to ‘learn’ features from a large volume of healthcare data from a network of information sources, and then use the obtained insights to assist clinical practice. IBM's Watson has been one of the biggest AI-related innovations in the healthcare space. Its nongovernment repository of healthcare data (largest in the world in this category) consists of 600 petabytes of information covering 300 million patients.
Here’s an interesting fact: An AI system can assist physicians by providing up-to-date medical information from journals, textbooks and clinical practices to inform proper patient care. In addition, an AI system can help to reduce diagnostic and therapeutic errors that are inevitable in human clinical practice.
AI: A point of logical action in Healthcare
Traditionally, AI in the healthcare domain refers to a deep neural network that allows a computer to perform logical actions based on data from the network. The problem with such a system is that the conclusions derived by computers can be monumentally incorrect for data that hasn’t been seen before. But, if such data was used to assist healthcare professionals, doctors & medical researchers by making their work easy, wouldn’t that be something!
AI in Patient Care & Safety
From a safety perspective, AI in its current state has the ability to reduce variations in patient care. Standardised tests, prescriptions and even procedures can help doctors stick to safe & proven methods of success, rather than taking unnecessary risks.
From the standpoint of the crucial belief that algorithms can provide unsafe suggestions, AI is only used as a medical assistant for practitioners.
AI as a Global Health Shield- ML & NLP
On December 31, 2019, a Canadian health monitoring platform, BlueDot sent word to its customers informing them of a ‘flu-like’ outbreak emerging from Wuhan, China. This was almost a week before the WHO & CDC (US) made an official statement about the same.
Now, we all know how arduous traditional methods of information dissemination & authentication are. BlueDot’s AI automated this. It gathers information from hundreds of thousands of sources (in 65 languages) and draws not only conclusions but also accurate predictions. The company’s AI picks up indications & murmurs from specialised forums, and wait for it; accesses global airline ticketing data to gather where and when people from affected areas are headed. The technology essentially uses Natural Language Processing and Machine Learning to this effect.
For those wondering how accurate & effective this is, once the automated data is sifted through by the NLP & ML technology, Epidemiologists check the conclusions to confirm it from a scientific perspective.
As a warning system, that is pretty effective!
AI in Drug Discovery & Treatment Methods for Healthcare
Earlier, we spoke of Artificial Intelligence using data to help the medical professionals. Well, it is not just limited to patient care and outbreak prediction. Typically, it takes about £1.94 billion and almost a decade of R&D etc. for a drug to reach the patients. This is majorly because of the amount of human labour that is required in largely manual clinical research. Research that essentially involves solving complex clinical problems.
This presents the challenge of acquiring, scrutinising, and applying that knowledge to solve these complex clinical problems. AI is a technology-based system involving various advanced tools and networks that can mimic human intelligence to perform the same tasks; at the end of which, human scientists can evaluate conclusions from a scientific standpoint.
To avoid the idea of an oversimplified process, here is some insight into how the current development of AI/Artificial Intelligence works to this end:
AI in drug discovery uses methods like reasoning, knowledge representation, solution search and Machine Learning (ML). ML uses algorithms that can recognize patterns within a set of data that has been further classified. Deep learning (DL), a part of ML engages artificial neural networks (ANNs) to mimic the impulses of the human brain to suggest conclusions & actions to healthcare researchers, professionals & practitioners. The aforementioned Watson of IBM uses a similar approach to suggest the right treatment methods for cancer patients.
Patient Discovery for Healthcare Via AI
Recently, there has been some success in optical imaging cameras helping discover patients with symptoms such as fever. Cameras using AI-based multisensory technology can detect people with fever, whether or not they’re wearing facial masks, and track their movements.
Although this does not guarantee immediate action from a patient care standpoint, such information can be used to curb the spread of contagious diseases.
Artificial Intelligence/AI has been around the block since the 1950s, and it has emerged from simply being an information-gathering tool to a crucial assistant in important fields such as healthcare. As it stands, Machine Learning & Natural Language Programming are moving leaps and bounds in improving & standardising simple and complex solutions such as patient care, drug discovery, R&D & information dissemination. Current issues with global healthcare pose a serious threat in terms of delayed measures- As such, Artificial Intelligence/AI can be a useful aide to the healthcare industry in ways more than one.