Browsing by Subject "computer science"

Now showing 1 - 4 of 4

Access status: Metadata only ,
Classical and isogeometric finite element method
(Akademia Górniczo-Hutnicza im. St. Staszica w Krakowie, 2021) Paszyński, Maciej; Paszyński, Maciej
Wydział Informatyki, Elektroniki i Telekomunikacji; Centrum e-Learningu i Innowacyjnej Dydaktyki
The textbook presents a comprehensive introduction to the finite element method, including methods of isogeometric analysis. The chapters are supplemented with sample MATLAB codes (runnable in the free version of Octave). The textbook is aimed at students of engineering and computer science, and therefore focuses on the practical and implementation aspects of particular topics related to the finite element method.
Access status: Restricted ,
Informatyka w ogólnym zarysie
(Wydawnictwa Naukowo-Techniczne, 2003) Brookshear, Glenn J.
Access status: Open Access ,
Segmentacja aorty z obrazów tomografii komputerowej przy użyciu głębokich sieci neuronowych
(Wydawnictwa AGH, 2025) Paniec, Piotr; Stanuch, Maciej; Skalski, Andrzej
W artykule zaproponowano wykorzystanie architektury nnU-Net do segmentacji aorty z obrazów tomografii komputerowej. Przetwarzanie wstępne obejmowało transformację różnych zakresów wartości do zakresu odpowiadającemu skali jednostek Hounsfielda oraz procedurę okienkowania. Na etapie treningu sieci przeprowadzono pięciokrotną walidację krzyżową oraz wytrenowano pięć modeli, których predykcje na zbiorze testowym zostały połączone. Dla modeli uzyskano średni współczynnik Dice'a na poziomie 0,921 oraz IoU równe 0,858, oznaczające wysoką jakość predykcji. Wartość odległości Hausdorffa ujawniła istnienie elementów odstających, a analiza trójwymiarowych rekonstrukcji wskazała obszary o niższej dokładności segmentacji.
Access status: Open Access ,
The interdisciplinarity of the publications of the Medical University of Silesia in Katowice based on the analysis of the co-occurrence of issues specific to medicine and computer science
(Wydawnictwa AGH, 2024) Paluch, Ewelina; Mikulska, Marcela
Over the years, the study of the interdisciplinarity of publications has taken various forms, from its identification based on the disciplines represented by the authors, through the examination of citations used when writing the article, to the analysis of the publication text itself. The last of these approaches seems to be the most reliable in the context of verifying the real integration between disciplines in a specific text. The approach utilized in the conducted research facilitates a deeper analysis of integration not only between disciplines in general but also between specific issues within their domains, aiding the examination of the intensity of such connections. The research was aimed at analyzing publications affiliated with the Medical University of Silesia in Katowice in terms of their connection with issues included in the area of Computer Science. OpenAlex, a bibliographic database supported by data mainly from Scopus, WoS and Google Scholar, which uses concepts that make up the Wikidata knowledge base to describe the content of publications was used. A list of 14,136 publications from the Medical University of Silesia in Katowice was downloaded from the OpenAlex bibliographic database including such data as: publication id, title, author, abstract, journal, date of publication, ISSN number or concepts. Overall, the most prevalent concepts in the publications were concepts regarding the field of the medicine (medicine, internal medicine, cardiology). The most prevalent concepts concerning computer science in the publications were: computer science, logistic regression and artificial intelligence. The strenght of the connections between concepts regarding medicine and computer science was calculated by calculating the arithmetic mean of the score value for each pair of IT and medical concepts contained in a single publication. The study showed the importance of computer Science issues in the medical publications and highligted the growing importance of AI in the field of medicine.