Detail publikace

Weakly Supervised Deep Learning-based Intracranial Hemorrhage Localization

NEMČEK, J. VIČAR, T. JAKUBÍČEK, R.

Originální název

Weakly Supervised Deep Learning-based Intracranial Hemorrhage Localization

Typ

článek ve sborníku ve WoS nebo Scopus

Jazyk

angličtina

Originální abstrakt

Intracranial hemorrhage is a life-threatening disease, which requires fast medical intervention. Owing to the duration of data annotation, head CT images are usually available only with slice-level labeling. However, information about the exact position could be beneficial for a radiologist. This paper presents a fully automated weakly supervised method of precise hemorrhage localization in axial CT slices using only position-free labels. An algorithm based on multiple instance learning is introduced that generates hemorrhage likelihood maps for a given CT slice and even finds the coordinates of bleeding. Two different publicly available datasets are used to train and test the proposed method. The Dice coefficient, sensitivity and positive predictive value of 58.08 %, 54.72 % and 61.88 %. respectively, are achieved on data from the test dataset.

Klíčová slova

Intracranial Hemorrhage; Computed Tomography; Deep Learning; Convolutional Neural Network; Weakly Supervised Learning; Localization; Attention; Multiple Instance Learning

Autoři

NEMČEK, J.; VIČAR, T.; JAKUBÍČEK, R.

Vydáno

1. 3. 2022

Nakladatel

SciTePress

Místo

SETUBAL

ISBN

978-989-758-552-4

Kniha

Proceedings of the 15th International Joint Conference on Biomedical Engineering Systems and Technologies - (Volume 2)

Strany od

111

Strany do

116

Strany počet

6

URL

Plný text v Digitální knihovně

BibTex

@inproceedings{BUT178071,
  author="Jakub {Nemček} and Tomáš {Vičar} and Roman {Jakubíček}",
  title="Weakly Supervised Deep Learning-based Intracranial Hemorrhage Localization",
  booktitle="Proceedings of the 15th International Joint Conference on Biomedical Engineering Systems and Technologies - (Volume 2)
",
  year="2022",
  pages="111--116",
  publisher="SciTePress",
  address="SETUBAL",
  doi="10.5220/0010825000003123",
  isbn="978-989-758-552-4",
  url="https://www.scitepress.org/Link.aspx?doi=10.5220/0010825000003123"
}