logo

Technological revolution for a preventive healthcare system

Artificial intelligence (AI) now characterises many aspects of our modern existence. But what about healthcare? Can the technology help prevent or detect disease early and be a helping hand for healthcare workers?

Doctor looking at virtual screen
When we achieve a balance between privacy and access to data, we can help create a sustainable health service for the future. (Illustration: colourbox.no)

The answer is a resounding yes!

One of the most exciting technologies that is the talk of the town this year is ChatGPT, an advanced AI-based language model. ChatGPT can perform various language tasks and is particularly well suited to natural language processing (NLP), a field of study that focuses on teaching computers to understand and process human language.

Technologies that analyse text in patient records could have a huge impact on healthcare. ChatGPT could potentially be the everyday assistant of the future for healthcare workers in terms of disease prevention and early detection. But how far from reality is this new technology when it comes to everyday service delivery?

Taking notes on paper towels

While those of us working with AI in healthcare share enthusiastic visions of how it can improve healthcare, it can seem far from the everyday lives of nurses. They tell stories about writing notes by hand on paper towels, which is far from the idea of using advanced technological solutions.

For nurses to benefit from AI-based tools in disease prevention, it is crucial that governments invest in raising awareness and taking action to ensure successful implementation of technology in healthcare.

Identifying at-risk patients earlier

Let's take postoperative delirium as an example. This is a complication that can impair a patient's consciousness and prolong their hospitalisation. In many cases, it is detected far too late and can have tragic consequences.

But what if we could use technology to identify at-risk patients even before surgery? By analysing text in patient records, decision-making tools can alert doctors to a patient's risk of developing postoperative delirium and prioritise preventive measures.

Technology can help provide information to healthcare professionals, both for prevention and to streamline their work. The burden on nurses, other staff, relatives and the patient themselves could be reduced if the tool existed.

How can such technology support the health service of the future?

Patient records contain a lot of data. Much of this is in text form. It is physically impossible for healthcare professionals to go through all the information in a patient record to gain a better understanding of a patient's condition and compare it with data from similar patients.

AI is able to read text from large volumes of patient records, learn from all available data and detect patterns in condition development in the context of patient characteristics to provide decision support that can improve patient care and, in some cases, save lives.

Privacy must be safeguarded

In order to provide accurate treatment prognoses and decision support, it is essential to have access to large amounts of data representing the patient group. This means that we must be able to retrieve patient data from different healthcare institutions, sometimes even across national borders, for example through an upcoming European health data area.

But we must be aware that health data is sensitive personal data, and privacy must be safeguarded. This can be done using privacy-preserving techniques, such as data anonymisation and federated learning that allow data analysis to be conducted without the sensitive data leaving healthcare institutions.

When we achieve a balance between privacy and access to data, we can help create a sustainable healthcare service for the future.

Solutions must be validated before use

The ongoing technological revolution has the potential to transform healthcare, but to achieve success, we need to take some important steps. A key factor is ensuring that the technological solutions work optimally for the patients they are intended to help, i.e. the solutions need to be validated before they are used in practice.

Unfortunately, this process can be complicated and unclear for many healthcare organisations. Therefore, we need a national ICT infrastructure and guidance for the validation of AI solutions in healthcare at a national level.

Is ChatGPT going to save the world?

When we ask ChatGPT if it could be the saviour of the future of healthcare, we get an interesting answer:

"Artificial intelligence can be of great help in healthcare in terms of patient monitoring, diagnosis, forecasting disease progression and optimising treatment plans."

At the same time, ChatGPT emphasises that "technology should not replace real healthcare professionals. Medical expertise and human care are still needed".

We couldn't agree more!

We look forward to a future where preventive tools can simplify the work of healthcare professionals, patients and relatives, and give us hope for a better healthcare service for all.