Weakly supervised semantic segmentation of tomographic images in the diagnosis of stroke

Результат исследования: Научные публикации в периодических изданияхстатья по материалам конференциирецензирование

Аннотация

This paper presents an automatic algorithm for the segmentation of areas affected by an acute stroke in the non-contrast computed tomography brain images. The proposed algorithm is designed for learning in a weakly supervised scenario when some images are labeled accurately, and some images are labeled inaccurately. Wrong labels appear as a result of inaccuracy made by a radiologist in the process of manual annotation of computed tomography images. We propose methods for solving the segmentation problem in the case of inaccurately labeled training data. We use the U-Net neural network architecture with several modifications. Experiments on real computed tomography scans show that the proposed methods increase the segmentation accuracy.

Язык оригиналаанглийский
Номер статьи012021
ЖурналJournal of Physics: Conference Series
Том2099
Номер выпуска1
DOI
СостояниеОпубликовано - 13 дек. 2021
СобытиеInternational Conference on Marchuk Scientific Readings 2021, MSR 2021 - Novosibirsk, Virtual, Российская Федерация
Продолжительность: 4 окт. 20218 окт. 2021

Предметные области OECD FOS+WOS

  • 1.03 ФИЗИЧЕСКИЕ НАУКИ И АСТРОНОМИЯ

Fingerprint

Подробные сведения о темах исследования «Weakly supervised semantic segmentation of tomographic images in the diagnosis of stroke». Вместе они формируют уникальный семантический отпечаток (fingerprint).

Цитировать