Health Technologies

Interview: the future of healthcare and inclusive AI

AI is rapidly innovating the field of healthcare, however, ensuring diversity and inclusivity is key to accuracy and usefulness of data in order to truly transform quality of life and treatment outcomes.

Good AI requires models trained on data that accurately reflect the diverse patient population they are designed to serve, Dr. Georgi Chaltikyan, program director master of digital health at the Deggendorf Institute of Technology, tells Health Tech World.

Biases can occur in AI when datasets overrepresent certain populations, and currently, underrepresented groups often lack sufficient representative digital health data, particularly in low- and middle-income countries due to limited or insufficient digital infrastructure, says Chaltikyan.

As factors such as genetics, treatment responses and disease progression can vary greatly across different populations, making data inclusive is crucial in achieving fairness and unbiased outcomes.

“It’s also about equity,” says Chaltikyan. “One thing that we definitely want to avoid with digital health, and especially with AI, is a digital divide. It’s not such an easy task to avoid, because the nature of digital health predisposes us to this divide, and we need to make significant efforts to avoid it.

“The divide can be related to several factors. For example, it can be a generational divide. The technology is now developing so fast that people of younger age already are naturally used to these technologies and are digitally literate compared to the senior generation.”

The main goal of the digitalisation of healthcare is making the person the centre of their care, Chaltikyan explains, noting that it is not just specialists who must become users of digital health, but also the citizens, who must make a combined effort to implement and use innovative technologies.

“The vision of the digitalisation of healthcare is essentially to make the person in the centre of their care, to be a co-pilot of their wellness and health,” said Chaltikyan.

In order for the digitalisation of healthcare to be successful, it must be approached from three angles, says Chaltikyan – from developers, healthcare workers and patients.

“Only that will allow wide adoption of digital health tools such as telehealth or digital health applications, sensor monitoring, and so forth,” he said.

Chaltikyan explains that part of the digitalisation vision includes the vision of 10P Health, a concept that builds on earlier work by Dr. Leroy Hood, one of the pioneers in personalised medicine, who first coined the ‘4P’ model of participatory, predictive, preventive, and personalised medicine.

The 10 P’s include: Predictive, Preventive, Personalised, Precision, Participatory, Pertinent, Proactive, Pervasive, Permanent, and Platform-based Health and Wellness.

“The term ’10P Health’ originates from a concept often associated with digital health: precision or personalised health,” said Chaltikyan.

“Historically, medicine was highly standardised, designed around evidence-based guidelines to harmonise treatments and ensure consistency. This approach was crucial when treatments were varied and unregulated.

“However, while standardisation ensures consistency, it often overlooks individual differences. For example, two people undergoing a cardiac evaluation at different clinics might receive identical protocols and treatments, even though individual needs may vary. Precision health aims to address these differences, tailoring care to the unique biology and circumstances of each person.”

Chaltikyan explains that one of the most important aspects of personalised medicine is that, based on this approach, massive amounts of data will be coming from multiple sources and will be analysed and processed in real time by AI.

“I often compare AI to an autonomous car’s onboard computer,” said Chaltikyan.

“Imagine your body is the car, and the digital wellness platform is the onboard computer. You wake up in the morning and engage in a dialogue with your AI co-pilot or doctor – the platform. It might inform you, for example, that last night’s choices, like eating a fatty steak, have raised certain levels in your body. It then provides tailored recommendations to address these issues.

“The platform can flag potential risks, predicting health concerns months or even years in advance. You are able to predict, and also conduct preventive care, improving the chances of preventing a lot of conditions.

“All of this, of course, requires high quality models, and high quality models require data, and data, ideally, must be inclusive.”

You may also like

Health Technologies

Accelerating Strategies Around Internet of Medical Things Devices

  • December 22, 2022
IoMT Device Integration with the Electronic Health Record Is Growing By their nature, IoMT devices are integrated into healthcare organizations’
Health Technologies

3 Health Tech Trends to Watch in 2023

Highmark Health also uses network access control technology to ensure computers are registered and allowed to join the network. The