DMDmitry M. Murashov

Dmitry M. Murashov

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Dmitry M. Murashov has a total of 39 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • On Assessing the Boundaries of the Information-Theoretical Measures of Digital Image Segmentation Quality

    by D. M. Murashov

    This paper extends the previous study on estimating the bounds of segmentation quality metrics, specifically variation of information and normalized variation of information between ground truth partition of an original image and its segmentation maps. Estimates for the lower and upper bounds of these quality metrics are derived based on an information-theoretic model of segmentation, where image transformations are treated as information channels, and assuming a deterministic segmentation algorithm. The estimates are computed as functions of the number of partition segments and the segmentation error probability, given the entropies of the original image and its ground truth partition. Examples of constructing the bounds for test images are provided. The established bounds are independent of the specific segmentation algorithm employed, allowing for an evaluation of the range of possible quality measure values for a given original image and its ground truth partition without performing the actual segmentation. These bounds facilitate the assessment of available ground truth partitions and the selection of the most suitable partition for the problem to be solved. The proposed bounds can be utilized both for segmentation algorithm selection and for hyperparameter tuning of the chosen segmentation method.

  • A new algorithm for detecting motion artifacts in video-EEG monitoring data

    by D.M. Murashov, Yu.V. Obukhov, 125009 Kotelnikov Ire Ras, I.A. Kershner, M.V. Sinkin, I.V Okuneva

    The article proposes a new algorithm for detecting artifacts in video recordings of long-term video-EEG monitoring data in the context of the problem of diagnosing delayed cerebral ischemia after subarachnoid hemorrhage. The main idea of the new algorithm is as follows. First, to ensure the invariance of the algorithm with respect to illumination variations, we calculate the motion measures from the edge map of the region of interest. Secondly, to expand the set of motion measures by including variation of information between the blurred edge maps of two consecutive frames. Third, to combine detectors that use different measures of scene motion. An experiment on clinical video-EEG monitoring data showed that the proposed algorithm provides sensitivity of 0.94, specificity of 0.94, accuracy of 0.94, and F1 score of 0.91. The obtained characteristics correspond to the level of known motion detectors in video-EEG data.

  • Open Access

    A New Algorithm for Automated Diagnosis of Delayed Cerebral Ischemia in Video-EEG Monitoring Data

    by Dmitry Murashov, Yury Obukhov, Ivan Kershner, Mikhail Sinkin, Irina Okuneva

    The paper proposes a new algorithm for automated diagnosis of delayed cerebral ischemia in video-EEG monitoring data. The algorithm uses the indicator of the delayed cerebral ischemia which is the number of epileptiform discharges per hour over a sufficiently long observation interval. It combines a detector of epileptiform discharges and a new motion artifact detector. The epileptiform discharge detection algorithm analyzes the cross-correlation function of EEG signals with a reference fragment pre-selected by medical experts. In the motion artifact detection algorithm, we propose to calculate the mobility index from blurred edge maps of the region of interest, which are less sensitive to fluctuations in illumination level. As an indicator of mobility, we use the variation of information, which characterizes the informational difference of video frames. We obtained preliminary results that confirm the effectiveness of the proposed algorithm. The developed algorithm for early automated diagnostics of delayed cerebral ischemia provides sensitivity equal to 0.9 and specificity equal to 0.79. The obtained quality measures correspond to the quality level of known methods for diagnosing delayed cerebral ischemia based on EEG signals. At the same time, the accuracy of detecting motion artifacts is 0.9, and the F1 score is 0.93. The detection algorithm for epileptiform discharges provides sensitivity of 0.95 and specificity of 0.85.

  • Assessing the Quality of Digital Image Segmentation

    by D. M. Murashov

    The article addresses the problem of assessing the boundaries of segmentation quality indicators: variation of information and normalized variation of information between the original image and its segmentation maps, as well as the segmentation error probability‒mutual information function. We propose an information segmentation model, in which image transformations are considered as information channels. We estimate the lower and upper bounds of the quality indicators, which are calculated as functions of the segmentation error probability for the specified information characteristics of the input image and its ground truth partition. Examples of constructing boundaries for test images are given. The constructed boundaries make it possible to estimate the range of possible values of segmentation quality measures for a given original image and its reference partition without performing segmentation. These bounds can be used to estimate the quality of existing ground truth partitions and select the most appropriate one to the problem being solved. The proposed bounds can be applied both to select a segmentation algorithm and to adjust the parameters of the selected algorithm.


  • An Algorithm for Measuring Texture Characteristics by Combining Fourier Spectra of Images Obtained in Raking Light

    by A. V. Berezin, E. Yu. Ivanova, D. M. Murashov

    This article proposes a new algorithm for measuring the texture characteristics of objects in images. The novelty of the algorithm lies in using pairs of images obtained in raking light and combining their Fourier spectra. The application of the developed algorithm is demonstrated in solving the problem of counting the threads of the woven bases of paintings from images. A computational experiment shows that the error in measuring the canvas density does not exceed 0.4 threads/cm. The accuracy of measuring the canvas density using different methods is compared. The proposed algorithm measures the characteristics of painting canvases with the texture distorted by pollution, ground penetration, or treatment with restoration materials, and is more accurate than the previously used algorithms. Using a new algorithm, the density of the canvases of seven paintings from the collection of the State Historical Museum is measured.

  • Accelerated Algorithms for Growing Segments from Image Regions

    by D. M. Murashov

    New algorithms for merging superpixel regions into segments are proposed. The main idea of merging superpixels is as follows. Firstly, a strategy is used in which a segment is grown from neighboring regions while the conditions for merging are met, and, secondly, when merging regions, the applied information quality measure should not increase. Three algorithms based on the specified strategy are proposed, which differ in the conditions for making a decision on merging superpixels. A computational experiment is carried out on test images. The experiment showed that the proposed algorithms accelerate the segmentation process compared to the procedure used earlier with acceptable losses of information quality measures of the resulting partitions.


  • Algorithms for Generating Image Segmentation Maps from Superpixels Based on a Combined Information Quality Measure

    by D. M. Murashov

    New algorithms for merging superpixels into segments using a greedy strategy and a combined quality measure including two components such as the informatio

  • Methods and Algorithms for Extracting and Classifying Diagnostic Information from Electroencephalograms and Videos

    by Yu. V. Obukhov, I. A. Kershner, D. M. Murashov, R. A. Tolmacheva

    This article describes new approaches and methods for analyzing long-term EEG data and synchronous video-EEG monitoring of patients with epilepsy and restoration of cognitive functions after moderate traumatic brain injury. EEG analysis is performed using the ridges of its wavelet spectrograms, the power spectral density, the frequency and phase of which, under certain conditions, corresponds to the square of the amplitude, frequency, and phase of the EEG signal. The results of studies of the frequency characteristics of a video stream when analyzing data from long-term synchronous video-EEG monitoring of patients with epilepsy are presented. Signs were obtained for recognizing epileptic seizures and differentiating them from events of a nonepileptic nature. Periodograms of smoothed optical flow calculated from fragments of patient video recordings were analyzed. Welch’s method was used to obtain periodograms. The values of the power spectral density of the optical flow at selected frequencies were used as features. A joint analysis of interchannel frequency synchronization, power spectral density of wavelet spectrogram ridges, and synchronous video made it possible to identify fragments with epileptic seizures on a long-term EEG, excluding various artifacts from consideration. Interchannel phase connectivity of the ridges makes it possible to observe the dynamics of EEG synchronization in patients with moderate traumatic brain injury during cognitive tests. Analysis of a network of phase-related pairs of EEG channels allows determining the positive dynamics of patient rehabilitation.


  • Open Access

    AN ALGORITHM FOR AUTOMATED DETECTION OF DELAYED BRAIN ISCHEMIA INDICATOR FROM VIDEO-EEG MONITORING DATA

    by D. Murashov, Y. Obukhov, I. Kershner, M. Sinkin

    In this work, we propose a new algorithm for detecting the indicator of delayed cerebral ischemia from video-EEG monitoring data. The proposed algorithm combines an algorithm for detecting the effect of interchannel time-frequency synchronization of wavelet spectrogram ridges of EEG signals and an algorithm for detecting motion artifacts in video recording frames. Using the EEG analysis algorithm, we identify an indicator that is used to predict the occurrence of ischemia. By analyzing the optical flow, we exclude time intervals in which erroneous fixation of ischemia indicators is possible. The developed algorithm was tested on the clinical data of video-EEG monitoring. The preliminary results obtained during testing confirm the fundamental possibility of detecting the ischemia indicators and the acceptable accuracy of detecting motion artifacts leading to an erroneous diagnosis.