Naslov
Application of arteficial intelligence in personalized treatment of oncology patients

Autori
Kelemenić-Dražin, Renata ; Budisavljević, Anuška

Izvornik
Libri oncologici : Croatian journal of oncology (0300-8142) Vol. XLIX (2021), Supplement 2; 37-38

Vrsta, podvrsta i kategorija rada
Radovi u časopisima, kratko priopcenje, znanstveni

Ključne riječi
artificial intelligence, oncology, personalized medicine

Sažetak
Objective: Oncology is one of the most dynamic branches of medicine. As a result of numerous oncology studies, there has been a significant increase in scientific and clinical data that the human brain cannot store. Advances in artificial intelligence (AI) technology have led to its rapid clinical application. In this paper, we wanted to see the role of the use of artificial intelligence (AI) in oncology. Methods: We conducted a comprehensive database search (Pub Med, MEDLINE, and Google Scholar) using the keywords: artificial intelligence, deep learning, machine learning, oncology, personalized medicine. We supplemented the database search with reference list checks. The primary objectives were clarity and innovation regarding the use of artificial intelligence in oncology and the ability to apply the results in everyday clinical work. Exclusion criteria were: not about cancer and AI, lack of relevance after reviewing the title and abstract, papers not written in English, publications before 2015, studies not reviewed, not applicable in everyday clinical work. From a large number of articles available to us, we have selected review articles and results of clinical trials according to their clarity and innovation in terms of the use of artificial intelligence in oncology. Results: The possibilities of using artificial intelligence in oncology are innumerable. It can be used for diagnostic purposes (screening programs, histopathology, and molecular diagnostics), therapeutic purposes (personalized treatment, anticipation of treatment side effects and response to therapy, treatment decisions), as well as for prognostic purposes (risk stratification, survival prediction, monitoring). For example, the use of stored mammograms can help us identify breast cancer, the use of stored digital pathology images can help us diagnose cancer and the analysis of stored dermoscopy images can identify skin lesions, such as melanoma. In general, we can say that AI is already used to perform various tasks in oncology at a level equal to or sometimes greater than the level of clinicians. The goal of modern oncology is personalized treatment. Given the large amount of data that oncologists are daily exposed if we want to personalize oncology, we need the help of AI. Namely, the prerequisite for personalized treatment is knowledge of the genomic data of the tumour, i.e. whether there are possible genomic mutations in the tumour as target points of oncological treatment. Hundreds of thousands of articles are published annually on genomic mutations and cancer. Therefore, databases are being created that aim to help clinicians (e.g. COSMIC, ExPecto, The Cancer Imaging Archive [http://www.cancerimagingarchive.net], and Genomic Data Commons Data Portal [https://portal.gdc. rak.gov]). Genomic tests in oncology also play an important role. Thus, only in 2017., the FDA approved several such tests (Oncotype Dx, Praxis Extended RAS Panel, MSK- IMPACT, and FoundationOne CDx). They help us predict the prognosis of cancer so we can avoid over-treating cancer patients. Although the application of AI in oncology seems very complex and still inaccessible to all oncology centres, given the required infrastructure and skills of clinicians, some segments of AI can already be used in everyday clinical practice. Thus, AI can help us make decisions about cancer treatment. Namely, decisions on oncology treatment are based on the assessment of the patient’s clinical condition (PS = performance status). It is a subjective assessment made by the clinician by observing the patient’s condition and based on data obtained in conversation with the patient. An objective assessment of the patient’s clinical condition (PS) is difficult because patients spend most of their time outside the hospital. But objective real-time activity data that we can collect with physical activity tracking devices, such as smartphones or smartwatches, can help. A prospective study by Gresham et al (2018) showed us that monitoring patient activity (steps, distance, stairs) can help not only in assessing the clinical condition of patients with metastatic cancer but also in assessing clinical outcomes (side effects, hospitalization, survival). These findings should be confirmed in larger, randomized trials. Conclusions: The application of AI in clinical practice presents new challenges for clinicians. Namely, in the era of evidence-based and patient-centered medicine, they will have to master statistical as well as computer skills, in addition to clinical ones. Therefore, it is necessary to start educating future doctors about the importance of artificial intelligence as soon as possible.

Izvorni jezik
Engleski

Znanstvena područja
Kliničke medicinske znanosti

POVEZANOST RADA