Artificial intelligence (AI) is the tech of the future and it is substantially contributing to significant developments in technology and business, be it Tesla’s autonomous vehicles or IBM Watson, moving AI from theory to commercial applications. AI has come a long way since it was first established as a field in 1956. AI in healthcare has the potential to transform the industry and the change is already underway. Accenture said that growth in the AI healthcare market is expected to reach $6.6 billion by 2021. Similarly, according to a recent report “Artificial Intelligence (AI) in Healthcare Market 2020-2026”, the global market for AI in healthcare is projected to have a CAGR of around 51.5 percent during the mentioned period.
The Covid-19 pandemic has also expedited the use of AI in the healthcare sector. An increased emphasis is now being given to the use of modern technologies such as the use of brain-computer interfaces (BCIs), arterial spin labeling (ASL) imaging, biomarkers and natural language processing (NLP). Healthcare organisations across continents are ramping up the use of technology in healthcare and AI is at the forefront. It enables analysing complex medical data by utilising software and algorithms and achieve results.
Since the pandemic, we have seen that many healthcare startups, digital healthcare service providers and even universities are emphasizing AI and collaborating its integrating in the healthcare industry. Demand for healthcare is increasing every year and we are noticing emerging trends of manpower shortages. AI can solve this problem. The advancement of technologies in other fields also especially telecommunication is also significant. Since the pandemic, telehealth as a sector has boomed. In the near future, AI could deliver healthcare services through a mobile phone.
Role of AI in healthcare
Today, there are numerous applications of AI on the market when it comes to healthcare. But the primary role of AI in healthcare is to increase efficiency, to help deliver healthcare services in a much better way and to help save lives. AI helps perform sophisticated and complex task at a much faster pace, in an efficient manner and at lower cost. It is changing how we deliver healthcare. From chronic diseases, cancer to radiology and risk assessment, AI helps healthcare workers or medical practitioners better understand the day-to-day patterns and needs of their patients. AI can help study complex medical data in an efficient manner and this allows the medical practitioners to provide their services in a much better way, offer solid feedback as well as guidance and support.
PwC said in a report that AI is already being used to detect diseases, such as cancer, more accurately and in their early stages. According to the American Cancer Society, a high proportion of mammograms yield false results, leading to 1 in 2 healthy women being told they have cancer. The use of AI is enabling the review and translation of mammograms 30 times faster with 99 percent accuracy, reducing the need for unnecessary biopsies.
At present, algorithms do a far better job than radiologists at detecting malignant tumours in patients. According to PwC, AI can help clinicians take a more comprehensive approach for disease management, better coordinate care plans and help patients to better manage and comply with their long-term treatment programmes. AI is also used for minimising errors and control disease progression.
We are living in the age of information and data is vital. Tech giants such as Google or social media giants such as Facebook use advanced AI to collect and store consumer data. Similarly, healthcare data is also crucial. In the age of digitalisation, electronic health record (EHR) developers are now using AI to create more intuitive interfaces to store health records of patients. These interfaces present medical records of the patient, from various providers, test labs, imaging labs, pharmacies in real-time which can be used by the patient’s doctor as per requirement. A sophisticated AI-based EHR software can also draw conclusive analysis, identify risks, evaluate health conditions and even auto-book you for an appointment.
Mads Jarner Brevadt, CEO and co-founder of Radiobotics told International Finance,” We have only touched the tip of the iceberg in terms of the impact AI has brought to healthcare.” His startup uses AI to help 3 main groups of clinicians – Radiologists, Orthopaedic Surgeons and Emergency Care staff.
AI is becoming common within healthcare and leveling up in terms of its ability with each year that passes. AI is being used to make the administrative side of healthcare more efficient and intelligent. “This has both saved money and had an impact on hospitals’ waiting times. Yet, the impact AI can bring towards clinical healthcare delivery, enabling us to diagnose and treat at a pace and of a quality that would have been unthinkable just a few years ago. As AI is built using data, it truly allows us to live up to the ideals of evidence-based medicine that we have long preached about,” he added.
Implementing AI in healthcare, what are the hurdles?
Even though we are at a primitive stage of AI development, it has been fast-paced.. As we move ahead and break AI barriers, the question of AI singularity will get asked more often. Singularity by definition is a hypothetical point in time at which technological growth becomes uncontrollable and irreversible. In short, does it mean AI will replace humans? While we may not have a definite answer to that question, there are chances that we may not hit singularity at all.
When it comes to AI in healthcare, there are a few challenges that the industry needs to overcome to make AI more mainstream in healthcare. The biggest concern that arises at this point in time is privacy. Data privacy is a growing concern in the modern-day and for AI to be efficient, it needs to collect a lot of data. So, amid privacy concerns, how willing will a patient be to share his or her medical data? Also, for years, many have raised concerns about the ethical implications of healthcare data storage and data security.
Along with privacy, there is also the issue of regulation. While regulations differ from jurisdiction to jurisdiction, regulations related to the use of AI are not clear or strong enough. Recently, the European Union (EU) introduced its ‘Proposal for a Regulation on a European approach for Artificial Intelligence,’ which stresses on the importance and creation of the first-ever legal framework on AI.
Another very important challenge that needs to be addressed is the black box problem. AI interprets the data available and offers a conclusion, however, we may not know how AI reached that conclusion. While addressing the needs of his patients, a doctor should be able to understand and explain to his patients why a certain procedure that was recommended by an algorithm will help them overcome their medical issues. Understanding how the AI algorithm works is key.
From an industry point of view, Mads said, “I think the answer to this question is two-folded. On the production side, access to data is a barrier we need to overcome. Data is truly the lifeblood of innovation in AI and the bedrock on which all technology is built upon. We will only unleash the true potential of AI when we have access to high-quality data with which we can have the best-performing AI. This will ultimately benefit us all as citizens and patients.
“The other challenge which I see is the change readiness of healthcare staff. We expect a lot from this group of people who are often overworked and suffer from burnout. Investment in change always has a human capital investment alongside to make sure the change realises its maximum benefit. As healthcare systems, we need to give these staff allocated time in which to invest in change programs, both from when we are providing our healthcare staff their education and when they are practising clinically.”
How AI is fighting Covid-19?
Given the highly infectious nature of the Covid-19 virus, AI can help reduce transmissions by reducing human contact. AI is not only playing a crucial role in protecting frontline healthcare workers, but it is also helping to identify high-risk patients at an earlier stage. Today, we are witnessing multiple AI-powered projects with technologies such as machine learning and big data are being constantly used across a broad range of fields to manage the different scenarios caused by the pandemic.
Interestingly, a team of leading scientists at the University of Liverpool, UK, has used machine learning to predict where the next novel coronavirus could emerge. According to them, this could help them predict or even prevent a future pandemic. AI is also being used extensively by healthcare workers to predict the transmission rate as well as track the spread of the virus across the globe.
In Singapore, amid a manpower crisis, NCS helped deploy AI-powered thermal cameras to carry out manual, one-to-one temperature measurements with handheld scanners. This drastically reduced the need for manpower. Similarly, NCS also co-implemented the Robotic Process Automation (RPA) solution to automate the admission, discharge and transfer of patients in and out of the Community Care Facility (CCF), again solving the problem of manpower crisis. In France, startup Clevy is using augmented assistance to help diagnose Covid-19 symptoms without having to leave the comfort of their premises.
MIT-IBM Watson AI lab is involved in multiple projects to help fight the spread of the Covid-19 pandemic. According to IBM’s website, it is pre-loaded to understand and respond to common questions about public services, including voting, government assistant, unemployment, Covid-19.
One of these projects includes identifying sepsis in Covid-19 patients as it can prove to be life-threatening for Covid-19 patients. This can be done by analysing white blood cells (WBC) of Covid-19 patients by using machine learning. This early diagnosis will offer doctors valuable time while treating the patients.
Tech giants such as Google, Microsoft and Apple are also involved in initiatives including contact tracing, drug development, remote communications between patients and clinicians among others. AI also played a big role in the development of Covid-19 vaccines as well. MIT’s website says a machine learning model developed jointly by Janssen and MIT data scientists played a key role in the clinical trial process for the Johnson & Johnson Covid-19 vaccine.
One thing is very clear. The pandemic in many ways has been a teleporter to the future. Many of the barriers which were long-standing in healthcare preventing large-scale digital transformation began to be reduced as the world came to terms with the crisis that we faced. “This increased the appetite for digital innovation in general and accelerated progress in AI in healthcare to a large extent,” Mads said.
“We are now facing another crisis in healthcare, that is how to treat all of the patients whose conditions have not been treated during the pandemic. I think this is another crisis where AI can help to equip clinicians with the tools to treat these patients more effectively. This is another opportunity for AI in healthcare to shine and once again prove that this is not only an area of ‘hype’ but can deliver real tangible impact,” he added.
The future of AI in healthcare
AI indeed has an important role to play to define how we deliver healthcare services in the future. We can see right now that the potential of AI is growing exponentially. Mads said, “This is great news for all industries but I think healthcare is uniquely positioned to feel the positive impact that AI can have. If you look at where Radiobotics helps most, Radiology departments, we are seeing right now a drastic imbalance between the number of Radiologists available to report, and the number of reports in which to be completed.
“This of course has a negative impact on our Radiologists as well as patients. AI tools deployed in this environment can help correct for this imbalance and ultimately increase patient outcomes. In the next five years, I would love to see AI deployed and doing repetitive rule-based tasks for which it is best at, and clinicians gifted with more time to treat our patients. Looking even further into the future, I would like AI to be deployed to increase preventative medicine, so figuring out how we best keep patients well rather than treating them when they are sick. “
While it is without doubt, we can claim that AI will achieve great heights when it comes to healthcare, the greatest challenge for the industry is to ensure AI can be successfully integrated to deliver sophisticated healthcare services at scale. For widespread adoption to take place, the challenges such as privacy, regulation and data collection must be addressed.
With regard to AI technology making humans obsolete, we can be quite confident that it won’t happen. Reality is over time; human efforts will still be required to deliver top class healthcare services but not at the same capacity. Moreover, human interaction with regard to healthcare would shift to delivering more human skills such as empathy, persuasion and big-picture integration.
As a global community, the sheer number of patients who require support is increasing year on year. “We do not have the manpower to deliver this care that is required. We are fortunate that the amount of medical innovation that is coming to the market each year is helping to address this shortfall. I am confident that this level of medical innovation is not slowly down any time soon, that we can meet the healthcare challenges of tomorrow and the future is bright for the healthcare sector,” Mads added.
So far, AI-powered solutions have only taken small steps in the healthcare industry. For AI to have a largescale impact on the global healthcare industry, we will have to wait for a couple of more years.