Scopus İndeksli Yayınlar Koleksiyonu / Scopus Indexed Publications Collection

Permanent URI for this collectionhttps://hdl.handle.net/11147/7148

Browse

Filters

2021 - 20242

Settings

Institution Author: search.filters.institutionAuthor.Önal, SevgiPublication Index: search.filters.publicationIndex.TR-Dizin

Search Results

Now showing 1 - 2 of 2
  • Article
    Blank Frame and Intensity Variation Distortion Detection and Restoration Pipeline for Phase-Contrast Microscopy Time-Lapse Images
    (Aves, 2024) Ucar, Mahmut; Iheme, Leonardo O.; Onal, Sevgi; Pesen-Okvur, Devrim; Yalcin-Ozuysal, Ozden; Toreyin, Behcet U.; Unay, Devrim
    In this study, we propose a preprocessing pipeline for the detection and correction of distorted frames in time-lapse images obtained from phase-contrast microscopy. The proposed pipeline employs the average intensities of frames as a foundational element for the analysis. In order to evaluate the degree of correction required for intensity variance, a normalization technique is applied to the difference between the average intensity of a specific frame and the median average intensity of all frames within the study. Our restoration method increases the histogram similarity between the distorted and non-distorted frames, preserves trans-passing pixels in regions of interest, and mitigates the development of additional distortions. The efficacy of the proposed method was evaluated using 15 395 time-lapse image frames from 27 experiments using our own dataset and 830 time-lapse images from four different experiments obtained from the cell tracking challenge. The results of the validation demonstrate a high degree of numerical and visual accuracy of the proposed pipeline.
  • Article
    Citation - WoS: 3
    Citation - Scopus: 6
    Improved Cell Segmentation Using Deep Learning in Label-Free Optical Microscopy Images
    (TÜBİTAK - Türkiye Bilimsel ve Teknolojik Araştırma Kurumu, 2021) Ayanzadeh, Aydın; Yalçın Özuysal, Özden; Pesen Okvur, Devrim; Önal, Sevgi; Töreyin, Behçet Uğur; Ünay, Devrim
    The recently popular deep neural networks (DNNs) have a significant effect on the improvement of segmentation accuracy from various perspectives, including robustness and completeness in comparison to conventional methods. We determined that the naive U-Net has some lacks in specific perspectives and there is high potential for further enhancements on the model. Therefore, we employed some modifications in different folds of the U-Net to overcome this problem. Based on the probable opportunity for improvement, we develop a novel architecture by using an alternative feature extractor in the encoder of U-Net and replacing the plain blocks with residual blocks in the decoder. This alteration makes the model superconvergent yielding improved performance results on two challenging optical microscopy image series: a phase-contrast dataset of our own (MDA-MB-231) and a brightfield dataset from a well-known challenge (DSB2018). We utilized the U-Net with pretrained ResNet-18 as the encoder for the segmentation task. Hence, following the modifications, we redesign a novel skip-connection to reduce the semantic gap between the encoder and the decoder. The proposed skip-connection increases the accuracy of the model on both datasets. The proposed segmentation approach results in Jaccard Index values of 85.0% and 89.2% on the DSB2018 and MDA-MB-231 datasets, respectively. The results reveal that our method achieves competitive results compared to the state-of-the-art approaches and surpasses the performance of baseline approaches.