An analysis carried out last year showed that Estonians are living longer today than ever, but as much as a third of this time can be spent battling health issues. This is clearly below the European average. “The number of years we spend living healthily in Estonia is an area that needs to be addressed,” Fridolin says. This concerns people’s physical and mental well-being alike.

One of the main goals of health technology research is to find ways of detecting changes in physiological systems and in their functions at an early stage, before the development of disease symptoms and pathological changes.

“Information technology, especially signal-processing, has a significant part to play in our research and is needed in order for us to develop new methods and devices,” Fridolin explains, adding that the widespread use of activity monitors and sports watches shows that people like to be constantly informed about their health. At the same time, the world is moving towards personal medicine and a people-centric approach.

The key element is mental well-being

The number of people suffering from depression has increased 40-fold in the last decade and now affects more than 6% of the population. Being impaired by depression is mainly caused by years of living with ill health.

The group researching the brain’s bioelectrical signals at the Department of Health Technologies at TalTech (Professor Maie Bachmann, senior researchers Hiie Hinrikus and Jaanus Lass, researcher Laura Päeske and doctoral students Toomas Põld and Tuuli Uudeberg), are working to develop an objective method for assessing the mental state of the brain that allows for the detection of mental disorders before the onset of subjective symptoms and that can be used in regular medical examinations among workers with high levels of responsibility (police, rescue workers and military personnel) and the general population.

These signals are measured from the scalp. It is like taking measurements from the shell of a computer to find out how the computer is performing. Electroencephalography (EEG) provides an overview of the bioelectrical processes in the brain without interfering with those processes, which would be impossible in an invasive study. The disadvantage of EEG is its low spatial resolution, as the signal at the surface electrodes reflects the fields generated by many neurons. At the same time, the great advantage of EEG is its temporal resolution, which enables the monitoring of ongoing bioelectrical processes to an accuracy of more than a millisecond.

Mental disorders, including depression, are mostly diagnosed at present based on subjective methods (tests and questionnaires). “One of the main goals of the research group that’s being led by Professor Bachmann is to find an objective method, based on the assessment of brain signals, for the early detection of mental disorders,” Fridolin explains.

To characterise the EEG power balance, the working group introduced the SASI spectral asymmetry index, which characterises the power balance above and below the average frequencies and is calculated as the relative difference between these powers.

“Professor Bachmann’s team made an interesting discovery that the power balance in the frequency bands selected above and below the maximum frequency changes in patients diagnosed with depression,” Fridolin says. “SASI was positive in the group of depressed patients, while in most healthy subjects it was negative. Laura Päeske defended her doctoral dissertation in March on the basis of her research in this very field.”

Päeske’s dissertation, which was partly based on the collaboration that took place within the EXCITE framework between the research group into the brain’s bioelectrical signals and the Department of Computer Systems (under Professor Jaan Raik) at TalTech, concluded that a signal from a single EEG channel provides comparable results to combining multiple channels, and it was suggested that the brain compensates for a poor functional network with stronger connections. In addition, EXCITE cooperation is ongoing with TalTech’s Department of Software Science (under Professor Yuri Belikov), where doctoral student Tuuli Uudeberg combines traditional EEG signal-processing methods with those used in data science to develop personal brain condition markers.

Fatigue and mental disorders go hand in hand. Fatigued workers are at a much higher risk of becoming depressed. One of the outputs of the research collaboration between the research groups of the EXCITE Department of Computer Systems (Professor Gert Jervan and senior researcher Mairo Leier) and the Department of Health Technologies (Professor Ivo Fridolin, senior researcher Kristjan Pilt, researcher Moonika Viigimäe and doctoral student Ardo Allik) is the automatic measurement of worker’s physical fatigue and stress, for which the first steps have been taken with the development of algorithms based on artificial intelligence (AI) machine-learning methods. The initial focus will be on the assessment of physical fatigue using non-invasive real-time recordings of physiological signals. A doctoral thesis on the topic, entitled ‘Assessment of physical fatigue based on portable devices’, is currently being written. In the future, one of the outputs of the EXCITE research could be boosting work safety and comfort through the development of systems hidden in workwear and the algorithms that operate within them. This kind of smart workwear would give workers the opportunity to be aware of their work and movement habits, which could be used as input for planning a more efficient time management system and safer work environment, or for automatically identifying patterns that could prevent accidents among workers in hazardous conditions.

A major killer can be identified early

Cardiovascular diseases such as atherosclerosis are among the leading causes of death around the world. Early detection of atherosclerosis and proper treatment can prevent the progression of the disease. Atherosclerosis and the formation of plaque on the walls of blood vessels are associated with the ageing of the arteries. In various diseases, including diabetes, the ageing process of the arteries is accelerated. Vascular calcification increases the stiffness of the blood vessels, which causes strain on the heart due to the greater resistance exerted by the blood vessels, making it more difficult to pump blood through them. This leads to cardiovascular diseases such as hypertension and coronary heart disease.

“A person’s vascular condition can be assessed non-invasively using pulse wave analysis,” Fridolin explains. “The pulse wave, which results from the wave of pressure changes caused by the pumping of blood by the heart, travels along the aorta to smaller arteries, eventually reaching the peripheral blood vessels. The speed and shape of the pulse wave depend on blood pressure and the elasticity of the blood vessel, therefore providing useful information about ageing and the condition of the person’s vascular system. As the elasticity of the blood vessel decreases, the velocity of the pulse wave increases and its shape changes.”

It is important to choose a non-invasive method to assess the parameters of the vascular system that does not affect the blood vessels or blood flow. A good option is the photoplethysmography (PPG), which can be used to record volumetric changes in blood. The method has already been integrated into many activity monitors and smart watches, which are finding increasing use.

Senior researcher Kristjan Pilt has developed a novel algorithm to analyse the PPG signal using the pulse waveform index PPGAI, which provides reliable information about the condition of the blood vessels. This algorithm was compared with the methods used in clinical practice in cooperation with North Estonia Medical Centre cardiologist and TalTech Professor Dr Margus Viigimaa, and the results showed the advantage of the new method in distinguishing subjects with normal arteries from subjects with increased arterial stiffness.

“Kristjan’s algorithm also has the advantage of suppressing noise in PPG signals by using adaptive filtering, and linking pulse waveform analysis to heart rate,” says Fridolin. Mairo Leier, a senior researcher at the Department of Computer Systems, has worked with EXCITE to develop an optical sensor for recording the PPG signal from arteries at different depths in the body. This enables the signal measurement locations on the body to be expanded and the accuracy of the method to be enhanced.

This novel approach, combined with an optical pulse wave recording system, is a promising method for use in assessing the elasticity and age of arteries in the early diagnosis of cardiovascular diseases.

Chronic diseases are a major concern

If preventive measures are delayed, a disease can turn into a chronic illness. The growing spread of chronic diseases is an inevitable process in an ageing society. The spread of chronic illnesses is approaching epidemic proportions, especially in more developed countries. Approximately 70% of public health funding is spent on combating chronic diseases in the European Union. Epidemiological studies have shown that 14% of people aged 65 or older have six or more chronic diseases. In addition, COVID-19 poses a serious challenge to people who are in risk groups due to the chronic illnesses from which they suffer.

“Final-stage kidney disease is one example of a serious chronic disease,” says Fridolin. “There are about 850 million people in the world suffering from various types of kidney disease, which is why it’s estimated that chronic kidney disease will be the fifth most common cause of shortened life expectancy by 2040.”

The kidneys cleanse the blood by filtering out metabolic residue, including toxic residue like uremic toxins. In the final stages of chronic kidney disease, the only option is to use renal replacement therapy: a so-called artificial kidney, which filters harmful substances out of the blood during haemodialysis. Renal replacement therapy is a vital but time-consuming and uncomfortable procedure for the patient. The treatment is very expensive, one of the most expensive among all chronic disease, but without it the patient would die. As such, it is very important to ensure the effectiveness of haemodialysis. Patients with severe COVID-19 are also at high risk of experiencing irreversible renal damage and therefore becoming dialysis patients.

The quality and effectiveness of renal replacement therapy have traditionally been monitored chemically, using periodic blood tests. The disadvantages here are the loss of valuable blood for the patient, the use of chemical reagents and the complicated measurement procedure.

The group researching biofluid optics includes Professor Ivo Fridolin, senior researchers Jürgen Arund, Jana Holmar and Risto Tanner, researcher Sigrid Kalle, doctoral student Joosep Paats and engineers Rain Kattai and Deniss Karai. The group has proposed a fundamentally new approach to monitoring haemodialysis and providing quality control in collaboration with North Estonia Medical Centre nephrologists Dr Merike Luman, Dr Annika Adoberg and Dr Liisi Leis and biochemistry laboratory quality manager Kai Lauri (SYNLAB Eesti OÜ). This approach uses optical radiation, in particular the absorption and fluorescence properties of the respective marker molecules. By measuring the absorption and fluorescence spectra of various molecules in the effluent dialysate using this optical method, it is possible to estimate their concentration and thus monitor the renal replacement therapy in real time.

Effluent dialysate is a fluid which carries away the waste products filtered out of the blood, usually into the sewer. The advantage of the optical method is that the measurement takes place on the effluent dialysate, thus providing the easiest way to connect the monitor without disturbing the haemodialysis process in any way. There is no need for blood tests or the use of chemical agents. By varying the optical frequencies and the absorption and fluorescence spectra, it is possible to record molecules both large and small. As a result of this research project, a real-time measurement sensor – the so-called Multicomponent Monitoring (MCM) concept – is being developed. It uses an optical method developed by the research group to monitor the removal of toxic residue from the blood. Cooperation with Professor Artur Jutman from Testonica Lab OÜ to test the hardware and software reliability of the MCM sensor developed for the implementation of On-Chip Health Technology at EXCITE is a continuation of previous joint projects with Professor Raimund Ubar’s research group.

More high-quality information for the digital health system

The Estonian Health Insurance Fund will soon find itself in a situation where there is not enough money for everyone. Accessing medical care is becoming more difficult. People need to take into account that if they do not look after their own health, their quality of life will decline prematurely.

“This is something that people are afraid of more than anything else,” says Fridolin. “That you would otherwise be able to enjoy life but you can’t because of poor health. It’s a great motivator to get people to change their behaviour.” This leads to the question of whether society should pay for the treatment of people who, despite simple, widely known health facts, do not follow them and refuse to take care of their health. Would having sufficient personalised health information help in such situations?

Fridolin is convinced that the e-health systems in Estonia should feature more high-quality information. People should be assessed and then informed of the results. Take, for example, classic parameters such as blood pressure or blood markers: indicators that are related to some of our biggest health risks. There are normative values available based on gender and age, and you can check how you compare to them at any time.

“If the values are in the red zone, it’s important to step in and intervene,” Fridolin advises. “We should be creating a timeline to gain an overview of trends. When people are young, before they become adults, all key personal indicators should be measured so that trends can be projected.”

Fridolin concedes that our e-health system is no longer a success story and has not been for many years now. In particular, this is because there is no simple, user-friendly way to enter data in a high-quality manner that enables doctors to listen to their patients. Additionally, there are technological compatibility issues between systems. Partly because of this, doctors are unable to obtain an overview of a patient’s vital indicators and risk factors and of the results of analysis on a timeline that would provide a snapshot of the patient’s condition. Our data are fragmented and unsystematic, and there are no big data, which medicine desperately needs. Moreover, machine-learning and artificial intelligence are of no use in such a situation. “If the information that’s entered is poor or incomplete, the results will be the same,” Fridolin remarks. Incorporating doctors in the process as a whole is also very important.

Finally, it is worth noting that the development of health technologies is not leading us closer to immortality, but instead to the knowledge that death is inevitable. This is why it is especially important that the years we are alive are lived as healthily as possible. For this to happen, people need to do a lot of work themselves – precisely because the closest sensors are located in and on ourselves.

The article was first published in the Estonian Centre of Excellence in ICT Research (EXCITE) magazine in 2021. 

Find out more about the EXCITE Center of Excellence: www.excite.it.ee

Photos: Hendrik Osula, Arno Mikkor

EXCITE is a project funded by the European Regional Development Fund.