Uporaba umetne inteligence v nuklearni medicini - sistematičen pregled literature : diplomsko delo

Hrovat, Polona (author), Žibert, Janez (mentor), Matjašič, Alenka (comentor), Rep, Sebastijan (reviewer)

Uvod: Umetna inteligenca v nuklearni medicini pripomore k optimizaciji načrtovanja in pripravi posegov, pridobivanju slik, analizi nastalih slik ter k hitrejši avtomatizaciji izvidov. Z večjo učinkovitostjo vseh korakov procesa se lahko zmanjša število potrebnih slikanj in odmerek sevanja za pacienta. Namen: Namen diplomskega dela je bil sistematično predstaviti vlogo uporabe umetne inteligence v nuklearni medicini. Želeli smo ugotoviti, na kakšne načine in v kolikšni meri je umetna inteligenca prisotna v nuklearnomedicinskem delovnem procesu radiološkega inženirja ter specialista nuklearne medicine. Metode dela: V diplomskem delu smo uporabili deskriptivno metodo dela s sistematičnim pregledom literature. Strokovne članke na temo uporabe umetne inteligence v nuklearni medicini smo iskali s pomočjo podatkovnih baz PubMed, ScienceDirect, Journal of Nuclear Medicine in Digitalna knjižnica Univerze v Ljubljani. Literaturo smo iskali s pomočjo vključitvenih in izključitvenih faktorjev ter po ključnih besedah v angleškem jeziku. Rezultati: V rezultatih smo pregledali 13 člankov. V njih so predstavljeni različni načini uporabe umetne inteligence v nuklearni medicini. Nekateri članki zajemajo celoten delovni proces nuklearnomedicinskega slikanja, spet drugi se osredotočajo na posamezno področje. Natančneje predelana področja so izboljšanje načrtovanja nuklearnomedicinskih preiskav, avtomatska interpretacija slik in izboljšanje kakovosti slik. Razprava in zaključek: Ugotovili smo, da je tematika teoretično že zelo dobro predelana in ponuja obetavne rešitve za dolgoletne težave, ki se pojavljajo v nuklearni medicini. Obenem pa nam pregledana literatura ne ponuja veliko podatkov o tem, v kolikšni meri so ti načini že prisotni v praktičnem okolju. Algoritmi umetne inteligence ponujajo rešitve že pri samem načrtovanju preiskave. Prav tako omogočajo samodejno izbiro protokola slikanja. Zelo obetavno je področje avtomatizirane interpretacije slik, ki ponuja vrsto različnih funkcij. Veliko pa se dela tudi na področju izboljšanja kakovosti slik. Pomembno je, da se radiološki inženirji in specialisti nuklearne medicine, ki delajo z umetno inteligenco, na tem področju ustrezno izobrazijo in tako zagotovijo varnost.

• Language: Slovenian
• Keywords: diplomska dela, radiološka tehnologija, umetna inteligenca, nuklearna medicina, PET/CT, izboljšanje načrtovanja nuklearnomedicinskih preiskav, avtomatska interpretacija slik
• Work type: Bachelor thesis/paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: ZF - Faculty of Health Sciences
• Place of publishing: Ljubljana
• Publisher: [P. Hrovat]
• Year: 2024
• Number of pages: 44 str.
• PID: 20.500.12556/RUL-162494
• UDC: 616-07
• COBISS.SI-ID: 208920323
• Publication date in RUL: 24.09.2024

Secondary language

• Language: English
• Title: Applications of artificial intelligence in nuclear medicine - a systematic literature review : diploma work

Abstract

Introduction: Artificial intelligence in nuclear medicine contributes to the optimization of planning and preparation of procedures, to the acquisition of images, the analysis of acquired images and faster automation of reports. The number of necessary scans for the patient as well as the radiation dose can be reduced by making all steps of the process more efficient. Purpose: The purpose of this thesis was to present the role of artificial intelligence in nuclear medicine systematically. We wanted to determine in what ways and to what extent artificial intelligence is present in the nuclear medicine working process of a radiographer and a specialist of nuclear medicine. Methods: In the thesis, we used a descriptive method with a systematic review of literature. We used the following databases to search for professional articles on the usage of artificial intelligence in nuclear medicine: PubMed, ScienceDirect, Journal of Nuclear Medicine and Digital Library of the University of Ljubljana. Literature was searched using inclusion and exclusion criteria and keywords in the English language. Results: We reviewed 13 articles in the results section. The articles present various ways of using artificial intelligence in nuclear medicine. Some articles cover the entire working process of nuclear medicine imaging, while others focus on specific areas. Specifically addressed areas include the improvement of the planning of nuclear medicine examinations, automatic image interpretation and image quality enhancement. Discussion and conclusion: We have found that the topic has been theoretically discussed in detail and it offers promising solutions for long-standing issues which appear in nuclear medicine. The algorithms used by artificial intelligence offer solutions already in the planning phase of examinations. They also enable automatic selection of imaging protocols. The field of automated image interpretation is very promising, offering a wide range of functions. A lot of work is also being done in the field of image quality enhancement. It is essential that radiographers and specialists of nuclear medicine who work with artificial intelligence receive the necessary education in this field and thus ensure safety.

• Keywords: diploma theses, radiologic technology, artificial intelligence, nuclear medicine, PET/CT, improvement of nuclear medicine examination planning, automatic image interpretation