LATEST UPDATES » Vol 26, Nos. 11 & 12, November & December 2022 – Worlds Within Worlds – Viruses, Humanity, and the Environment       » Pinpointing How This Key Protein Facilitates Viral Transmission From Insects to Plants       » A New Approach to Treating Organic Wastewater       » Using Old Plants for New Tricks?       » Using Gas Bubbles as Lenses to View Tissues More Deeply       » Seawater as a Renewable Energy Source       » Generating Oxygen Within Cells
Vol 22, No. 09, September 2018   |   Issue PDF view/purchase
Welcome to the healthcare revolution
Dr Mark Burby tells us more about the potential of artificial intelligence in healthcare, from treatment and diagnostics, to predictive medicine and wellness.

In your own words, how would you define artificial intelligence in healthcare?

The application of artificial intelligence (AI) in healthcare is very broad but essentially refers to the use of technology to analyse data to obtain meaningful insights—similar to, or better than, a medical professional.

These insights can take many shapes and forms. For instance, the interpretation of large clinical data sets such as medical records or genomic data for correlations, detecting lesions on a radiology image, converting speech or free text in clinical notes to structured data (NLP), enhancing image quality to enable more accurate diagnosis, or predicting clinical outcomes.

How important is AI in healthcare?

If we look at the current state of affairs the world over, public healthcare is being challenged in many ways. Issues ranging from shortage of caregivers and other resources, increase in cancer and chronic diseases to rising costs, are responsible for these challenges. Additionally, spending on healthcare is expected to keep increasing substantially driven by factors like rising income levels, greater health awareness, improved access to insurance and an increased precedence of lifestyle diseases. But at the same time, in many countries, government spending on healthcare remains woefully inadequate, or is actively being controlled. All these factors present an ideal opportunity for AI to address medical situations where technology can fill in for the lack of staff or manual calculation or analysis.

For instance, AI could be used to address shortages of doctors wherein machines are able to manage common symptoms and diagnose diseases. Another example is where AI systems can actively assist clinicians to help them work more efficiently, enabling faster diagnosis and help manage finances better.

Outside the healthcare provider space, AI systems are being used to accelerate drug discovery that can subsequently improve access to medicines and save lives lost to chronic diseases.

Compared to other industries, how does the healthcare industry fare when it comes to utilising AI?

The healthcare industry is one of the leading industries in the use of AI. Large scale investments – both public and private – are being made in AI across the globe. The underlying reason for this is the sheer potential of benefits that AI promises. Many industries are greatly benefitting from the use of AI— from oil and gas exploration, and defect detection in manufacturing, to autonomous vehicles. It would be hard to say where the healthcare industry really is, compared to others, given the explosion of the activity in all the various fields.

Mainstream adoption of AI, however, still faces roadblocks as patients, healthcare professionals and regulators require more confidence in the technology, particularly in the areas of accountability and potential for errors. Some forms of AI, especially Deep Learning, make it difficult for us to understand how insights are gathered or decisions are made. Over time, these challenges will be addressed and AI will become integral to how we experience care and live our lives.

In Asia-Pacific, which countries are leading in AI-healthcare and which countries are seeing potential?

In terms of leadership, China is a global leader in AI - both in healthcare and other industries. According to local sources, the total available market for AI medical devices in China reached US$1.95 billion in 2017 and is projected to hit US$2.93 billion in 2018.

South Korea and Japan are also making great progress —with South Korea taking steps to develop homegrown medical AI systems through investments of up to US$33.4 million over the next three years. The Japanese government is expected to invest more than US$100 million to set up 10 AI enhanced hospitals by 2022.

We are also seeing a lot of activity in terms of start-ups and innovation in other countries like Singapore, India and Australia.

It is all about personalised healthcare now. Tell us more about this.

Huge amounts of healthcare data are being generated, creating a system of record, such as electronic medical records, wearables generated and genomic data. These data sets collectively generate what is frequently referred to as Big Data. AI offers us the ability to not just manage this deluge of data and make sense of it, but also generate a system of insights. Additionally, AI has the ability to handle unstructured or non-standardised data as well as manage disparate data types simultaneously.

When we take data that is personal to a patient, such as their genome, demographics, radiology images or medical history, we can analyse this against the population or research data.

To look at a few examples of how this can help, think about how we can use this approach. We can find out which medication is the most effective for that patient’s specific genetic profile, what is the optimal dose, what are the expected outcomes – all tailored to that specific patient. This is in stark contrast to the established population health system which aims to standardise patient care.

How has AI helped medical researchers in tackling chronic diseases?

Medical research, particularly in the pharmacological field involves high risks and high costs. Patients suffering from chronic diseases can often benefit from remote monitoring. Many patients are part of clinical studies or trials that generate huge volumes of data, which needs to be analysed by researchers. Advanced machine learning can not only analyse this data, but also translate the results quickly to the researchers managing the trial. This, in turn, can help reduce trial costs, improve data quality, and decrease time to market.

What is the outlook for the adoption of AI in hospitals in the near future?

Looking back three or four years ago, only a few hospitals were looking at AI, mainly out of curiosity and an interest in understanding the potential of AI. Today, we see widespread interest across most hospitals, and AI is almost always a discussion point when one is talking about technology to these hospitals.

Most healthcare IT events have AI as a theme and most solution providers and device manufacturers are incorporating AI into their products. Intel revealed findings from a new study about AI in Healthcare conducted in partnership with Convergys Analytics. For the study they surveyed 200 U.S. healthcare decision-makers about their attitudes about AI and the perceived barriers to adoption in the industry. Some highlights of the data from that study include the following points:

  • 37 percent already use AI today, though most to a limited extent;
  • 54 percent expect widespread adoption of AI within the next five years;
  • Respondents see a lack of trust in AI among patients (36 percent) as well as clinicians (30 percent) as barrier to adoption

What do you think are the top three AI applications with the most value potential in healthcare?

The value potential is immense but a few examples of the uses that AI can be leveraged on today, include the following:

  1. Imaging analytics
    1. Radiology (x-ray, CT, MRI, USS etc.) e.g. detection of possible tumours on mammograms
    2. Scopes e.g. fundoscopy to detect diabetic retinopathy
    3. Smartphones e.g. identifying possible skin melanoma
  2. Clinical decision support
    1. Virtual triage through chatbots or NLP (voice)
    2. Assisting doctors in diagnosis and identifying optimal care pathways (based on personalised medicine)
    3. Sepsis prediction – identifying patients at higher risk and intervening earlier to prevent deterioration
  3. Laboratory testing
    1. Improving accuracy of genomic sequencing and analysis
    2. High content screening for drug discovery
    3. Analysis of lab results together with patient data to get more accurate insights faster, leading to quicker decisions and better outcomes

What do you wish to see in the future of AI in healthcare?

AI is going to be at the heart of the advancement in nearly every industry, particularly healthcare, transforming the way we experience medical care. AI will take over the areas of diagnosis and treatment given that it is expected to become much faster and more accurate than human doctors. We will see this first in fields like radiology and dermatology, which are very image-focused. These are the fields that are witnessing and experiencing most of the advancements taking place in AI today. These developments will empower clinicians to become advisors or counsellors who will be able to help patients interact with the system and support them with decisions and education.

Another development will be around predictive medicine – this is an exciting area where diseases and critical events are pre-empted and prevented. Imagine wearing a device which detects your vital signs that sends your data to an AI system in the cloud, in real-time. Combining this data with your existing medical record, your genomic data, environmental data, and daily activity data, an AI system could potentially identify patients that are suddenly at a high risk of having a heart attack, stroke, or seizure. Workflow could be set up to either alert you to go to a nearest hospital for preventative treatment, send an autonomous ambulance to drive you to the nearest hospital, or send a drone with medication to you. All examples of AI in action!

There will also be greater emphasis on wellness – AI will help people make better decisions about their lifestyle and give people insights into the consequences of their actions. For example – how going for a run will benefit me, how drinking an entire bottle of wine in one evening will affect my system, what is the impact of losing 5kg on my risk of diabetes or cardiovascular disease?

The overall impact of all this is that we will see a reduction in the number of doctors, hospital beds, clinics as compared to today. Insurance premiums will go down and will be tightly linked to behaviour. This will have a big socio-economic impact too. Even things like heart attacks or strokes will become a thing of the past. The possibilities of how AI can transform healthcare are endless.

Dr Mark Burby is the health and life sciences sales director at Intel, Asia Pacific Japan Territory

news analytica Vietnam Exhibition Returns to Reunite the Industry After Its 4-Year Hiatus
news 2022 PDA Aseptic Processing of Biopharmaceuticals Conference
news Thailand LAB INTERNATIONAL, Bio Asia Pacific, and FutureCHEM INTERNATIONAL are ready to offer the Science and Technology Industry complete solutions this September!
news Better together: registration opens for Vitafoods Asia 2022 co-located with Fi Asia in October

About Us
Available issues
Editorial Board
Letters to Editor
Contribute to APBN
Advertise with Us
World Scientific Publishing Co. Pte. Ltd.
5 Toh Tuck Link, Singapore 596224
Tel: 65-6466-5775
Fax: 65-6467-7667
» For Editorial Enquiries:
   [email protected] or Ms Carmen Chan
» For Subscriptions, Advertisements &
   Media Partnerships Enquiries:
   [email protected]
Copyright© 2022 World Scientific Publishing Co Pte Ltd  •  Privacy Policy