POLAND Trends and Developments Contributed by: Barbara Kiełtyka, Jakub Gładkowski and Małgorzata Kiełtyka, Kieltyka Gladkowski KG Legal
eration of false or misleading scenarios by AI systems can undermine trust in the doctor-patient relationship, which in the long run may negatively impact the effec - tiveness and quality of treatment. Important Issues for Entities Using AI in Medicine Entities implementing AI systems that impact the health and lives of patients have a primary obligation to ensure transparency in the operation of these systems. In medicine, transparency should involve informing phy - sicians and patients about the use of AI systems in the treatment process. Furthermore, appropriate technical and organisational measures should be implemented to ensure the systems are fit for purpose. The AI Act further introduces the requirement to establish an appropriate human oversight system, meaning that entities creating such a system should designate appropriately qualified individuals respon - sible for such oversight. The user is also responsible for ensuring the appro - priateness of input data, ie, ensuring that the data feeding the system is relevant, current, accurate, and fit for its intended purpose. A hospital using an X-ray image analysis algorithm will therefore be responsi - ble for ensuring its high quality and ensuring that key information is not omitted. It is also essential to continuously monitor the opera - tion of the AI system throughout its use, so that if a fault is detected, the system’s creator can be immedi - ately notified. Furthermore, the entity using the system is obligated to retain logs (event records) generated by the AI system for at least six months. Further, a requirement to register the system in a special database dedicated to high-risk AI systems is provided (Article 71 of the AI Act). This ensures that untested systems do not enter the market. Digitalisation in Healthcare – Frequent Investments in AI AI-based solutions are increasingly being considered for use in healthcare, particularly given staffing short - ages. AI can play a significant role in both record - ing and automating bureaucratic processes and pro - vide real support for doctors at every stage – from
diagnosis, through treatment, to recovery. A growing percentage of Poland’s funds allocated to the digi - talisation of the healthcare system is being directed towards AI-based projects. Among the innovative solutions streamlining the healthcare system, many focus on supporting diagnostics, health monitoring and preventive measures. An example is the use of AI-assisted chest CT scan analysis, which allows for the identification of lesions and the determination of their size, extent and location. Other solutions include the use of pseudonymised medical data and AI algorithms to detect individuals potentially at risk of specific diseases – including rare and ultra-rare diseases – and to create lists of patients requiring further diagnosis. AI is also intended to sup - port infertility treatment, including in vitro fertilisation. Although questions may arise in the future about the impact of automation on physician autonomy and ethical issues, currently AI systems only play an aux - iliary role – they support physicians, but do not replace them in making medical decisions. NIL Innovation Network An important “space on the map” of Polish healthcare innovations is the NIL Innovation Network (Physician Innovators Network at the Supreme Medical Cham - ber), an initiative aimed at supporting and promoting innovation in medicine. It also aims to integrate physi - cians from various fields interested in new technolo - gies, treatments, and approaches to healthcare. The NIL Innovation Network comprises several groups. The largest is the Working Group on Artificial Intelli - gence (WGAI). Its goal is to monitor, develop and sum - marise the implementation of AI-based technologies in the Polish healthcare sector. WGAI operates by collect - ing information on the performance of such technolo - gies from providers, healthcare recipients and patients who have had experience with such solutions. The NIL Innovation Network also has groups on inno - vation in hospitals, outpatient care, medical technol - ogy, health and well-being, medical data, and medi - cal workflow and culture. Each of these areas may be expanded to include AI-based innovations in the future.
114 CHAMBERS.COM
Powered by FlippingBook