CC BY
11"time markers" while enzymatic degradation of leaf tissue in order to optimize and test the protoplast isola-
tion protocols.
This work was supported by Genomic Research Center �Kurchatov Genomic Center � NBG-NSC�.
New possibilities of the PCA-Seq method in the analysis of time series
V. M. Efimov1,2,3,4, K. V. Efimov5, D. A. Polunin2, and V. Y. Kovaleva3
1Institute of Cytology and Genetics SB RAS
2Novosibirsk State University
3Institute of Systematics and Ecology SB RAS
4Tomsk State University
5Higher School of Economics
E-mail: efimov@bionet.nsc.ru
DOI 10.24412/cl-35065-2021-1-02-97
When decomposing a one-dimensional time series using SSA, two PCs correspond to one frequency. But
they can be modulated by a third PC without breaking orthogonality [1, 2].
In the PCA-Seq method [3], PCs are calculated through a matrix of distances between fragments of a time
series. The original and modulating signals can be separated by fragments pre-normalizing. A reliable zero cor-
relation with the time axis may testify in favor of the hypothesis that the found regularity really exists and can
be continued in both directions from the interval under study. We use traditional solar activity data as an ex-
ample.
This work was supported by the RFBR grant #19-07-00658-a and the Budget Project of the ICG SB RAS #0259-2021-
0009.
References
1. Efimov V. M., Galaktionov Y. K., Shushpanova N. F. Analiz i prognoz vremennykh ryadov metodom glavnykh
komponent. Novosibirsk: Nauka. Sib. otd-nie, 1988.
2. Courtillot V., Lopes F., Le Mouel J. L. //Solar Physics. 2021. V. 296, N. 1. P. 1-23.
3. Efimov V. M., Efimov K. V., and Kovaleva V. // Vavilov J. Gen. and Breeding. 2019. V. 23. P. 1032-1036.
Automatic abdominal aneurysm segmentation using deep learning
R. UI. Epifanov1, R. I. Mullyadzhanov1,2, A. A. Karpenko1,3
1Novosibirsk State University
2Institute of Thermophysics SB RAS
3Meshalkin National Medical Research Center
Email: rostepifanov@gmail.com
DOI 10.24412/cl-35065-2021-1-02-39
Abdominal aortic aneurysm is a serious disease, delayed diagnosis and surgical treatment of which is often
fatal. Early detection and subsequent observation of the dynamics of the development of aneurysmal trans-
formations of the aorta allows for advance planning of surgical intervention, which significantly reduces the
risk of postoperative complications and mortality in patients. The main method for diagnosing this disease at
the moment is laborious visual analysis of three-dimensional computed tomography images of the patient.
Creation of a tool for automatic segmentation of an abdominal aortic aneurysm will simplify the diagnosis of
this disease. The paper considers the creation of an abdominal aortic aneurysm segmentation tool based on
deep learning methods. The algorithm is built by adapting the U-net architecture, which has proven itself for
solving biomedical problems [1], to the analysis of three-dimensional images of computed tomography of pa-
tients.
This work is supported by the Russian Science Foundation grant 21-15-00091.
References
1. Ronneberger O., Fischer P., Brox T. U-net: Convolutional networks for biomedical image segmentation // Intern.
Conf. on Medical image computing and computer-assisted intervention. Springer, Cham, 2015. P. 234-241.
Three-dimensional shape reconstruction of abdominal aortic aneurysm for hemodynamic modeling
Y. V. Fedotova, R. I. Mullyadzhanov
Novosibirsk State University
Email: i.antonevich@g.nsu.ru
DOI 10.24412/cl-35065-2021-1-03-02
Abdominal aortic aneurysm (AAA) is a degenerative disease that significantly increases the rupture risk of
arterial wall. It�s a localized enlargement of the abdominal aorta such that the diameter is greater than 3 cm or
more than 50 % larger than normal. So an accurate prediction of AAA rupture is critical.
In this study, algorithm for automatic 3D reconstruction of AAA, obtained from tomographic images, was
developed and implemented. Accurate reconstruction of the geometry is necessary for automatic mesh pro-
cessing with a high quality in hemodynamic simulations. Moreover, the developed program can be applied to
measure important volume geometric characteristics of AAA [1]. The results of the work will be included in the
software package for predicting the rupture risk.
This work was supported by the Russian Science Foundation, project 21-15-00091.
References
1. Martufi, G., Di Martino, E. S., Amon, C. H., Muluk, S. C., & Finol, E. A. (2009). Three-dimensional geometrical char-
acterization of abdominal aortic aneurysms: image-based wall thickness distribution.
Formalization of target invariants of the system of anaerobic biological wastewater treatment
A. A. Fomenkova, A. A. Klyucharev, S. I. Kolesnikova
Saint-Petersburg State University of Aerospace Instrumentation
Email: skolesnikova@yandex.ru
DOI 10.24412/cl-35065-2021-1-02-40
A model of a two-stage process of anaerobic fermentation in a bioreactor-mixer is considered, the math-
ematical description of which is a system of nonlinear differential equations [1].
The aim of the study is to apply a new approach to organizing energy-saving control based on the princi-
ples of nonlinear adaptation on target manifolds with an attractive property [2], called invariants or laws of
behavior of the target system of the object under study.
In this regard, a necessary preliminary study is the formalization of these target invariants as given (de-
sired) laws of the control object's behavior.
Further, on their basis, an algorithm is proposed for the analytical synthesis of a vector controller with
compensation for systematic and random disturbances along the control channel ofthe system of anaerobic
biological wastewater treatment [3].
This work was supported by the Russian Foundation for Basic Research (grant 20-08-00747).
