Artificial intelligence is revolutionising healthcare

Artificial intelligence in medicine – the da Vinci robot

Da Vinci robots support the work of specialists in many fields of medicine – urology, surgery, gynaecology, cardiology or earyngology (ENT). In 2017, nearly 900,000 surgeries using this system were performed in Poland. The first such surgery took place in 2010 at the Regional Specialist Hospital in Wrocław. It was performed by prof. Wojciech Witkiewicz on a patient with colorectal cancer [1]. This was a true symbol of the technological change in Polish hospital care. Currently, our health care is increasingly using this modern technology. Its widespread use has been influenced by the positive opinions and recommendations of the Agency for Health Technology Assessment. From 1 August 2023, patients with endometrial and colorectal cancer can benefit from surgical robot support as part of treatment under the National Health Fund. The procedures are performed in 22 facilities that meet the requirements. As participants in the debate on challenges in colorectal cancer diagnosis and treatment commented – its “future are robots” [2].

Artificial intelligence offers an opportunity to advance many areas of life, including medicine.

Progressive computerisation in healthcare

Technology is also making its way into Polish hospitals and medical facilities through digitisation and digital transformation. A breakthrough in this area was the “Act on the amendment of certain acts in connection with the implementation of solutions in the area of eHealth”, adopted in 2019. The development of services such as, among others: the Patient’s Internet Account (IKP) or the streamlining of documentation workflow through electronic medical record (EDM) are its practical aftermath. Progressive computerisation in healthcare helps to reduce the time it takes to complete documents, as well as enables and improves the coordination of patient care [3]. Digitisation makes it possible to issue prescriptions and make recommendations for the patient in an instant, and this improves the work of the staff.

Artificial intelligence as a milestone in healthcare

Another milestone in healthcare is the use of artificial intelligence. Machine learning algorithms analyse medical imaging to help radiologists identify abnormalities better (more accurately and efficiently). Natural language processing (NLP) algorithms sift through vast amounts of medical literature to aid evidence-based practice and treatment recommendations. In addition, AI-based predictive models have proven effective in predicting disease outbreaks and optimising resource allocation in healthcare systems.

Polish contribution to the development of AI technologies

Poland also has its part in the development of such technologies. Future Processing, a technology company from Gliwice, implemented the ECONiB project [4] in collaboration with the Gliwice Oncology Centre. Its aim was to increase the diagnostic power of dynamic imaging after contrast enhancement in personalised oncology by extracting new and improved biomarkers. The key of the project was to analyse the pixels of medical images in search of relevant diagnostic information. The findings made it possible to increase the diagnostic efficiency of MRI and enabled faster detection of malignant lesions. Reviewing a large number of images by a human, who has to maintain a high level of concentration, is less efficient than when technology does it.

Gliwice scientists work on technologies to support medicine

Gliwice is also home to the European HealthTech Innovation Centre [5]. It is a modern scientific and research centre of the Silesian University of Technology supporting the processes of generating, developing and implementing new technologies applicable in medicine, rehabilitation, and sport [5]. The centre’s mission and goals are clearly defined. The idea is to create a platform combining the research and development capabilities of ultra-modern laboratories with the potential of the scientific team of engineers at the Faculty of Biomedical Engineering of the Silesian University of Technology and business entities as well as other research and development units operating in the health sector.

– We develop and implement projects that will, on the one hand, relieve the burden on the health service and, on the other, make a real contribution to better treatment of patients. It is here that both Polish and foreign companies find research facilities for the development of their ideas in the field of medicine or, more broadly, health, declares the centre’s director, prof. dr hab. inż. Marek Gzik [6].

Artificial intelligence as support for medical diagnostics.

One of the most recent examples of success in the field of using artificial intelligence to support diagnostics is the StethoMe project. “A stethoscope the size of half an orange which contains artificial intelligence that listens to check whether everything is OK in your lungs. And it does so even better than doctors. It’s not science fiction, it’s StethoMe, a device created by Polish scientists,” writes
Gazeta Wyborcza newspaper about the project whose originator is Dr Honorata Hafke-Dys [7]. As the author explains, in the device, instead of the doctor, the user’s breath is monitored by an app installed on the user’s phone, which guides the user step by step through the examination process.

Artificial intelligence as support for medical diagnostics.

Artificial intelligence – is it completely reliable?

One of the most pressing questions about artificial intelligence in medicine is the one concerning its reliability. Although artificial intelligence shows remarkable potential and performance, will we ever be able to say that it is completely reliable? Healthcare professionals verify the insights gained through artificial intelligence to ensure patient safety and effective care. The responsibility lies with them and this situation, in all likelihood, will not change any time soon. Even if, and such opinions are being voiced, AI becomes “smarter than humans”.

Artificial intelligence in legal terms

Another issue is legislation. 2022 saw the adoption of the “White Paper” developed by the Polish Hospital Federation, the AI in Health Coalition and the AI Working Group (GRAI) [8]. It includes, among other things, guidance on the use of AI in accordance with the rules of medical professions. The document emphasises that AI cannot replace the medical professional, make decisions for them, or make independent judgements about the medical condition of a patient. AI is intended to support therapeutic, diagnostic, scientific and research, organisational and management processes and to be a tool in the hands of the medical professional.

Data processing

There are legal acts in force in the European Union that regulate these issues. One of these is the General Data Protection Regulation (GDPR). It sets out the key and overarching legal framework governing the processing, storage, and handling of personal data, including health data. And the development of algorithms is closely linked to data processing. They are simply taken as examples and a kind of training material. By comparing millions of records, the algorithm “learns” to recognise the results it has been programmed to look for.

Artificial intelligence in the EU legal framework

The “AI Act” of the European Union, adopted in May this year, has been widely commented. It is a regulation designed to create a framework to safeguard the interests of citizens and provide a space for digital business to thrive [9]. According to its provisions, AI systems are divided according to four risk categories (unacceptable, high, limited (risk of manipulation) and minimal). The use of systems characterised by unacceptable risk is to be completely prohibited. The creation and use of systems with a high degree of risk, on the other hand, is to be subject to a number of restrictions and obligations. For systems with minimal risk, it will be the responsibility of developers and administrators to inform users that they are encountering artificial intelligence when using a particular product or service.

Ethics

A separate and also extremely important aspect is the ethics relating to this issue. Although guidelines in this area are not legally binding, they are certainly also crucial when using artificial intelligence, including in healthcare. The World Health Organisation (WHO) in May this year called for caution regarding the use of Artificial Intelligence. “While WHO is enthusiastic about the appropriate use of technologies, including LLMs, to support health-care professionals, patients, researchers and scientists, there is concern that caution that would normally be exercised for any new technology is not being exercised consistently with LLMs*. This includes widespread adherence to key values of transparency, inclusion, public engagement, expert supervision, and rigorous evaluation” [10].

Human collaboration with innovative technologies

The integration of artificial intelligence in medicine is a milestone in improving healthcare practices. It comes with high hopes for diagnostics, treatment and operational efficiency. However, achieving the absolute reliability of artificial intelligence is still a vision for the distant future, and this highlights the need for continuous improvement and collaboration between human knowledge and technological innovation.

Sources:

[1] https://www.zwrotnikraka.pl/robot-da-vinci-w-polsce-zastosowanie/ (dostęp 02.01.2024)

[2] https://www.isbzdrowie.pl/2023/11/przyszloscia-leczenia-raka-jelita-grubego-sa-roboty/ (dostęp 02.01.2024)

[3] https://www.pzh.gov.pl/informatyzacja-w-ochronie-zdrowia/ (dostęp 02.01.2024)

[4] https://better.future-processing.com/projects/econib (dostęp 02.01.2024)

[5] Strona główna – EHTIC (dostęp 02.01.2024)

[6] https://ehtic.eu/ (dostęp 02.01.2024)

[7] https://wyborcza.biz/biznes/7,177150,30324693,stethome-polka-zbudowala-ai-do-leczenia-astmy-i-chorob-oddechu.html?disableRedirects=true (dostęp 02.01.2024)

[8] https://www.gov.pl/web/ai/sztuczna-inteligencja-w-medycynie (dostęp 02.01.2024)

[9] https://cyberdefence24.pl/polityka-i-prawo/co-zmienia-unijne-regulacje-sztucznej-inteligencji-ai-act (dostęp 02.01.2024)

[10] https://www.who.int/news/item/16-05-2023-who-calls-for-safe-and-ethical-ai-for-health (dostęp 02.01.2024)

* LLM (Large Language Model) to wyspecjalizowany rodzaj sztucznej inteligencji (AI), który został przeszkolony na ogromnych ilościach tekstu w celu zrozumienia istniejącej treści i wygenerowania oryginalnej treści.