Artificial intelligence to help in contacts with dementia patients

Credit: Adobe Stock
Credit: Adobe Stock

A Polish nurse has developed an artificial intelligence (AI) solution for learning to communicate with dementia patients. For this idea, Adrian Nowakowski has been awarded a prestigious distinction by the Queen of Sweden - the Queen Silvia Nursing Award.

The award will be presented to the winner in September.

Adrian Nowakowski works in the Department of Neurology of the Central Clinical Hospital of the University Clinical Centre of the Medical University of Warsaw, and studies computer science. He decided to combine the two fields to create a tool that 'could help health care workers and patients'.

'I addressed the communication challenges faced by medical staff, especially nurses caring for patients with dementia. I have developed an AI simulator that helps them to learn and practice effective communication', he told PAP.

He explains that a person with dementia often does not remember what happened recently, but may clearly recall events from years or decades ago.

'In the clinic, I work with people with communication difficulties every day. The first days of work were quite a challenge for me because I had to learn to communicate with them. I was not prepared for this after graduation', says Adrian Nowakowski.

His aim is primarily intended to benefit nursing students. The idea is that the student turns on the phantom and talks to artificial intelligence that 'pretends' to be a patient. The future nurse tries to persuade AI, for example, to take medication. The AI behaves like elderly people with dementia sometimes do, for example it refuses to swallow pills. The student must find a way to convince the machine to change its mind.

'You need to develop certain techniques, understand this person, their point of view - so that you can enter into a conflict-free dialogue. If we don't know that the patient suffers from dementia and we start talking as we would to a healthy person, we will encounter really big difficulties', Nowakowski points out.

He says that while working on the project, he combined several currently available tools. 'I created appropriate prompts, i.e. instructions, and introduced them to various artificial intelligence models responsible for +sight+ or text processing', he explains.

The phantom prepared in this way can recognize the image of the environment, e.g. objects, shapes, people, and recognizes speech. It can process information to provide answers. Appropriate prompting makes the machine 'perceive' reality and react like a patient suffering from dementia.

'I had to write a characterization of such a patient so that the language model would know what to do. The patient has flashes of mental clarity, which means that sometimes the conversation is similar to one with a person without such disorders. After a while, however, the patient may not remember what just happened, and become upset or irritated by this. Such patients tend to be lost, they may not recognize their surroundings and familiar people. I included all this in the prompt', says Nowakowski.

He recalls that, as a novice nurse, the biggest problems for him were contacts with patients who did not understand that their actions could be harmful to themselves.

'For example, the danger arises when a stroke patient tries to stand up despite not being able to maintain balance. It is often difficult to convince such a patient to stay in bed', Nowakowski continues.

He emphasises that the way to get such a person to stay in bed is to have a 'very long and polite conversation'. 'In such situations, what matters most is patience', he explains and adds that this is what artificial intelligence teaches future nurses.

After completing each exercise, the AI analyses what happened and summarises the students' behaviour towards the virtual patient. It also suggests phrases that would be more effective in a conversation with a person suffering from dementia and reports at what point the future nurse made a mistake.

'If you use the word +tachycardia+ when talking to a patient, at the end of the simulation the artificial intelligence will point out to us that this professional term was most likely incomprehensible to the patient. It will advise that in such a case it is better to say +heart problems+', says Nowakowski.

He adds that his solution can be used not only in medical communication exercises, but in many areas such as basic nursing classes.

'Throughout the first year, students have classes in laboratories and learn on phantoms. Artificial intelligence could be used there and thus offer an additional year of communication training', he says.

In addition, better communication with a patient with dementia means more effective care and lower stress levels for the patient and medical staff.

Nowakowski believes that artificial intelligence could also help patients' relatives - family members could exercise with the machine when the elderly person shows the first symptoms of dementia. 'Advanced dementia does not appear suddenly, it is a long process. Therefore, it is worth using all methods to prepare for contact with a person whose disease is progressing', he says.

On the question of whether advanced robots may one day replace medical staff, Nowakowski thinks it is unlikely.

'If we were to think about science fiction, robots could one day support families of patients or medical caregivers, for example by doing shopping or relieving them in the kitchen. However, I am convinced that machines cannot replace humans in patient care. Firstly, a robot may cause anxiety in a person with dementia. Secondly, in the case of patients suffering from dementia, contact with another person is extremely important, it can even delay the progression of the disease. After all, what counts in medical care is empathy and human care', he says.

The Queen Silvia Nursing Award is a scholarship established by Swedish Care International in 2013 for nursing students and nurses. The winners receive EUR 6,000 and the opportunity to complete an internship. The organisation adds that the award aims to 'foster positive change, growth, innovation and excellence in the field of nursing, especially geriatric nursing’. The scholarship is awarded in Sweden, Finland, Poland, Germany, Lithuania and at the University of Washington (USA).

The organizer of the Queen Silvia Nursing Award in Poland is the Medicover Foundation. In 2023, the tenth edition in history and the eighth in Poland was held. Nearly 170 Polish nurses and nursing students competed in this edition.

The Queen of Sweden will present the awards to the winners on September 12 during the Award Grand Ceremony at the Royal Palace in Stockholm.

Data from the Central Statistical Office show that in 2050, people aged 60+ will constitute 40% of the Polish society - there will be almost 14 million of them. In turn, the Public Information Bulletin of the Commissioner for Human Rights contains information that there are currently over 500,000 people with dementia living in Poland, many of them without a diagnosis. Among them, 310,000 suffer from Alzheimer's disease but are also mostly undiagnosed. (PAP)

PAP - Science in Poland, Anna Bugajska

abu/ bar/ amac/ kap/

tr. RL

The PAP Foundation allows free reprinting of articles from the Nauka w Polsce portal provided that we are notified once a month by e-mail about the fact of using the portal and that the source of the article is indicated. On the websites and Internet portals, please provide the following address: Source:, while in journals – the annotation: Source: Nauka w Polsce - In case of social networking websites, please provide only the title and the lead of our agency dispatch with the link directing to the article text on our web page, as it is on our Facebook profile.

More on this topic

  • Credit: Adobe Stock

    Improved air quality saved at least 25,000 Polish lives, new study finds

  • Fot. Adobe Stock

    Mind wandering supports creativity, says scientist

Before adding a comment, please read the Terms and Conditions of the Science in Poland forum.