Artificial intelligence and big data: a paradigm shift in healthcare

Artificial intelligence (AI) and big data are frequently mentioned in many domains, often hyped as universal solutions to problems, in particular in healthcare where they already have a significant impactb

AI and big data will be central to the future of medical care to prevent and treat a myriad of issues, by assisting physicians and treatment. According to Allied Market Research, the global big data analytics in healthcare market size is projected to reach USD 67.82 billion by 2025, growing at a CAGR of 19.1% from 2018 to 2025.

Since the use of AI in healthcare means access to a large volume of data about individuals, the data must remain private and be protected from cyber threats

Key to healthcare future?

International standards for AI and big data are developed by ISO/IEC JTC 1/SC42, the Subcommittee on Artificial intelligence, of the ISO/IEC Joint Technical Committee on Information Technology. These standards will support the work of IEC TC 62: Electrical equipment in medical practice and that of other IEC TCs.

Healthcare challenges

Most regions face significant healthcare challenges. These include aging populations, lack of physicians or inadequate healthcare infrastructure.

AI can provide important benefits in some domains, such as research, diagnosis, finding the most effective treatments, telemedicine, or developing entirely new and better drugs, to achieve what is known now as personalized or precision medicine.

To be of any use in any domain, AI needs to process and interpret extensive data sets to draw relevant conclusions. Agreements and safeguards permitting the exchange of healthcare information at country and international levels, provide an unequalled volume of data.

Big Data analytics

According to the ISO/IEC 20546:2019 standard, the term Big Data implies datasets that are so extensive in volume, velocity, variety and/or variability that they can no longer be handled using existing data processing systems.

Big data for healthcare covers many domains and a wide range of categories. These include data from imaging technologies, patient electronic health records (EHR) information, from research on genome sequencing, epidemiology, cancer, aging and, increasingly, data generated by wearable or implanted sensors or medical devices.

Machine learning (ML) and "deep learning" are central to the processing and interpretation of all these datasets to develop AI tools appropriate for each category and, within these, applicable to individual cases.

Imaging technologies provide incomparable information

Medical imaging has been used to visualize the interior of patients’ bodies ever since the application of X-rays was discovered in the late 19th century. It is an indispensable tool that allows doctors to diagnose and treat internal illnesses and traumatic injuries.

The diagnostic imaging domain has made dramatic advances with the introduction of new technologies, which allowed it to expand well beyond its initial limitations.

IEC work central to medical imaging technologies

Medical imaging technologies rely essentially on IEC International Standards developed by IEC SC 62B: Diagnostic imaging equipment. Some examples include:

  • X-ray imaging equipment
  • computed tomography (CT), which uses X-ray images processed by computer to produce tomographic images or “slices” of internal organs
  • magnetic resonance imaging (MRI) scanners and related associated equipment and accessories

International standards for ultrasound imaging, used widely to monitor and diagnose the condition of certain internal organs, such as liver, kidneys, gallbladder and even the heart, are prepared by IEC TC 87: Ultrasonics.

Magnetic resonance imaging (MRI) scanners use superconducting magnets and radio waves to produce images of the inside of the body. International standards for superconducting materials used in MRI scanners are developed by IEC TC 90: Superconductivity.

AI and ML are needed to interpret huge volume of data

Imaging technologies produce vast amounts of data, according to a recent paper published in a science journal; brain imaging alone “is currently producing more than 10 petabytes [10 000 terabytes] of data every year with a staggering nine fold increase in data complexity (i.e. data acquisition modalities) over the last three decades.”

In addition to imaging, other sources, such as electrocardiography and electroencephalography (ECG, EEG) and HER, generate large volumes of data.

A meaningful use of these for diagnosis and treatment requires AI, which, in turn, entails the interpretation of high-quality comprehensive data using ML.

Developing drugs

Another very promising AI healthcare application is the development of new, targeted drugs. Major pharma companies started working with hospitals, researchers and others using AI to analyze large volumes of data and then make estimates or recommendations. Several start-ups have stepped in to occupy niche opportunities.

They compile everything the AI software needs to “learn” before it analyzes a patient’s condition, according to a news agency report. This includes information on disease causes, symptoms and progression, many past patients’ test results, doctor reports and scanned images. Given the time saved and the reduced number of clinical trials, the ROI can be very significant.

Cautionary, protective measures and shortcomings

Since the use of AI in healthcare means access to a large volume of data about individuals, the data must remain private and be protected from cyber threats.

This can be achieved, to a certain extent, though the anonymization of data, making it impossible to trace back information to individuals or clusters of individuals.

This is first and foremost the responsibility of healthcare professionals and service providers using best practices and, among others, standards developed by ISO/IEC JTC 1/SC 27: Information security, cyber security and privacy protection.

Another less obvious but important issue, linked to the previous two points, is the possible use of collected healthcare data by healthcare insurance providers, in certain countries, to reduce coverage or raise premiums.