Szczegóły publikacji
Opis bibliograficzny
A preliminary approach to plaque detection in MRI brain images / Karolina Milewska, Rafał Obuchowicz, Adam PIÓRKOWSKI // W: Innovations and developments of technologies in medicine, biology and healthcare : proceedings of the IEEE EMBS International Student Conference (ISC) : [Zabrze, December 11–12, 2020] / ed. Natalia Piaseczna, Magdalena Gorczowska, Agnieszka Łach. — Cham : Springer Nature Switzerland, cop. 2022. — (Advances in Intelligent Systems and Computing ; ISSN 2194-5357 ; vol. 1360). — ISBN: 978-3-030-88975-3; e-ISBN: 978-3-030-88976-0. — S. 94–105. — Bibliogr., Abstr. — Publikacja dostępna online od: 2021-10-28
Autorzy (3)
- AGHMilewska Karolina
- Obuchowicz Rafał
- AGHPiórkowski Adam
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 137390 |
|---|---|
| Data dodania do BaDAP | 2021-11-05 |
| DOI | 10.1007/978-3-030-88976-0_13 |
| Rok publikacji | 2022 |
| Typ publikacji | materiały konferencyjne (aut.) |
| Otwarty dostęp |  |
| Wydawca | Springer |
| Czasopismo/seria | Advances in Intelligent Systems and Computing |
Abstract
Demyelinating plaques are areas of perineural demyelination, i.e., areas where myelin has disappeared due to regional ischemia or autoimmune diseases; they are visible as white matter hyperintensities in MR imaging. Regional blood flow reduction provokes a state in which the metabolic needs of the neuron are not met. The aforementioned metabolic imbalance of the neuron is the most frequent causative factor of demyelination. Local thrombosis, small vessel disease, or vessel compression are possible causative factors that provoke regional ischemia. The autoimmune background is also a possible cause as it usually results in numerous foci in brain white matter and the spinal cord. This article presents the experimental results of various algorithms applied for the segmentation of plaques. This issue is briefly discussed from a medical point of view, after which data obtained from magnetic resonance imaging, in particular from the Turbo Inversion Recovery Magnitude (TIRM) protocol, is shown. Then, experimental test results for convolutional filtering and a group of local binarization algorithms are presented. Algorithm sequences are also proposed that achieve an important increase in accuracy. The results are evaluated and the parameter spaces are defined for the best algorithms, including Sauvola, Local Normalization and SDA.
