Selection and Quantification of Objects in Microscopic Images: from Multi-Criteria to Multi-Threshold Analysis

Bogachev, Mikhail I.; Volkov, V.; Kolaev, Gleb; Chernova, Liliya S.; Vishnyakov, Innokentii E.; Kayumov, Airat R.

doi:10.1007/s12668-018-0588-2

Cited by 15 publications

(4 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At the correction stage, the operator is required to manually examine the images, meticulously, side-image by side-image, to extract the outlines of the target structures and make proper editing adjustments to eliminate the local inaccuracies. Further improvements of the automated segmentation could be achieved with advanced noise reduction [24,25] or multilevel CT scan processing and adaptive object selection algorithms with decision-making based on improved a posteriori statistics, although these algorithms require considerably more computational resources to obtain these statistics in advance [26][27][28]. To date, manual segmentation provides a higher accuracy, although requiring additional activity and more qualified CT scan operators.…”

Section: Discussionmentioning

confidence: 99%

Intraoperative Sphenoid Sinus Volume Measurement as an Alternative Technique to Intraoperative Computer Tomography

et al. 2020

Self Cite

View full text Add to dashboard Cite

Isolated sphenoid sinus disease (ISSD) is where there is a group of pathologies characterized by inflammation in one or both sphenoid sinuses. Although computer tomography (CT)-based 3D reconstruction remains the gold standard among noninvasive approaches to ISSD diagnostics, no standardized techniques for direct intraoperative measurements of the sphenoid sinus volume in ISSD patients have been documented. We suggest a novel technique for the intraoperative measurement of the sphenoid sinus volume. Our technique is based on filling the sinus with 0.01% methylene blue solution after an endoscopic endonasal sphenoidotomy. The proposed technique was applied to 40 ISSD patients during surgery. Obtained intraoperative measurements were compared to noninvasive measurements from 3D reconstructions based on preoperative CT scans. Our results demonstrated that the obtained measurements did not exhibit significant differences exceeding 0.4 cm3, with CT-scan-based measurements in 39 out of 40 cases (p < 10−6, Wilcoxon sign-rank nonparametric test), thus confirming the accuracy of the proposed technique. Disagreements between direct intraoperative and CT-based measurements in a single case have been attributed to the presence of remaining pathological masses in the sinus, which was further confirmed during the secondary check of the operated sinus. Accordingly, we suggest that the agreement between the CT-based and intraoperative volume measurements can be used as an indicator of the successful elimination of all pathological masses from the sinus without having to perform an adequate exposure of the entire sphenoid sinus to reduce intraoperative bleeding. The proposed technique is accurate and does not require the involvement of specialized intraoperative CT scanners and avoids additional radiation exposure for the patient during an additional postoperation CT scan to confirm the success of the surgery.

show abstract

Section: Discussionmentioning

confidence: 99%

Intraoperative Sphenoid Sinus Volume Measurement as an Alternative Technique to Intraoperative Computer Tomography

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…The figure shows that cells and cell clumps are being successfully distinguished from the image background using local edge density, that does not require any physical markup like contrasting or staining. In comparison with the recently reported algorithm (Sinitca et al, 2023), using multi-thresholding opens ways towards multi-class segmentation that is essential in a variety of experimental settings, for example, when specified cells sub-populations have to be distinguished and selected (see, e.g., (Bogachev et al, 2018, Bogachev et al, 2019, Volkov et al, 2020 and references therein).…”

Section: Cells Segmentationmentioning

confidence: 99%

“…Simple computer vision methods such as intensity thresholding are typically efficient only in combination with physical markup based either on staining or contrasting of studied samples, a common approach in microscopic imaging in various biomedical research settings, or on the calculation of specialized indices based on multiple spectral bands, a common approach in remote sensing with multiple applications in agricultural, geological and other environmental sciences (Bogachev et al, 2018, Bogachev et al, 2019, Volkov et al, 2020. Therefore, in recent years deep learning based solutions attracted considerable attention due to their potentially higher accuracy, although at the cost of performance and resource intensiveness.…”

Section: Introductionmentioning

confidence: 99%

Multi-Class Segmentation of Heterogeneous Areas in Biomedical and Environmental Images Based on the Assessment of Local Edge Density

Sinitca,

Lyanova,

Kaplun

et al. 2023

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

Abstract. Imaging techniques employed in biomedical and ecological applications typically require complex equipment and experimental approaches, including sophisticated multispectral cameras, as well as physical markup of samples, altogether limiting their broad availability. Accordingly, computerized methods allowing to obtain similar information from images obtained in visible light spectrum with reasonable accuracy are of considerable interest. Edge detection methods are commonly used to find discriminating curves in image segmentation. Here we follow an alternative route and employ edge detection results as a separate metric characterizing local structural properties of the image. In turn, their characteristics such as density or orientation averaged in a gliding window are used as a virtual channel substituting multispectral imaging and/or physical markup of samples, and the following image segmentation procedures are performed by thresholding. In complex segmentation scenarios, a single fixed threshold often appears sufficient, and thus relevant adaptive multi-threshold algorithms are of interest, with slope difference distribution (SDD) thresholding algorithm representing a prominent example.

show abstract

“…In the following, we show how similar strategy can be applied to the lane selection and band detection in the protein mass spectra analysis problem for the SDS-PAGE fingerprinting. and that one that leads to the best results according to the optimization criteria is finally applied [29,30]. Based on the assumption that lanes have more or less the same width, the minimum of the sum of the standard deviation of the lane width was chosen as the optimization criteria.…”

Section: Lane and Band Selection In The Protein Mass Spectra Analysis...mentioning

confidence: 99%

Improved online event detection and differentiation by a simple gradient-based nonlinear transformation: Implications for the biomedical signal and image analysis

Sokolova

Uljanitski

Kayumov

et al. 2021

Biomedical Signal Processing and Control

View full text Add to dashboard Cite

Selection and Quantification of Objects in Microscopic Images: from Multi-Criteria to Multi-Threshold Analysis

Cited by 15 publications

References 21 publications

Intraoperative Sphenoid Sinus Volume Measurement as an Alternative Technique to Intraoperative Computer Tomography

Intraoperative Sphenoid Sinus Volume Measurement as an Alternative Technique to Intraoperative Computer Tomography

Multi-Class Segmentation of Heterogeneous Areas in Biomedical and Environmental Images Based on the Assessment of Local Edge Density

Improved online event detection and differentiation by a simple gradient-based nonlinear transformation: Implications for the biomedical signal and image analysis

Contact Info

Product

Resources

About