-
When does a bent concatenation not belong to the completed Maiorana-McFarland class?
Authors:
Sadmir Kudin,
Enes Pasalic,
Alexandr Polujan,
Fengrong Zhang
Abstract:
Every Boolean bent function $f$ can be written either as a concatenation $f=f_1||f_2$ of two complementary semi-bent functions $f_1,f_2$; or as a concatenation $f=f_1||f_2||f_3||f_4$ of four Boolean functions $f_1,f_2,f_3,f_4$, all of which are simultaneously bent, semi-bent, or 5-valued spectra-functions. In this context, it is essential to ask: When does a bent concatenation $f$ (not) belong to…
▽ More
Every Boolean bent function $f$ can be written either as a concatenation $f=f_1||f_2$ of two complementary semi-bent functions $f_1,f_2$; or as a concatenation $f=f_1||f_2||f_3||f_4$ of four Boolean functions $f_1,f_2,f_3,f_4$, all of which are simultaneously bent, semi-bent, or 5-valued spectra-functions. In this context, it is essential to ask: When does a bent concatenation $f$ (not) belong to the completed Maiorana-McFarland class $\mathcal{M}^\#$? In this article, we answer this question completely by providing a full characterization of the structure of $\mathcal{M}$-subspaces for the concatenation of the form $f=f_1||f_2$ and $f=f_1||f_2||f_3||f_4$, which allows us to specify the necessary and sufficient conditions so that $f$ is outside $\mathcal{M}^\#$. Based on these conditions, we propose several explicit design methods of specifying bent functions outside $\mathcal{M}^\#$ in the special case when $f=g||h||g||(h+1)$, where $g$ and $h$ are bent functions.
△ Less
Submitted 24 April, 2024;
originally announced April 2024.
-
CPAISD: Core-penumbra acute ischemic stroke dataset
Authors:
D. Umerenkov,
S. Kudin,
M. Peksheva,
D. Pavlov
Abstract:
We introduce the CPAISD: Core-Penumbra Acute Ischemic Stroke Dataset, aimed at enhancing the early detection and segmentation of ischemic stroke using Non-Contrast Computed Tomography (NCCT) scans. Addressing the challenges in diagnosing acute ischemic stroke during its early stages due to often non-revealing native CT findings, the dataset provides a collection of segmented NCCT images. These inc…
▽ More
We introduce the CPAISD: Core-Penumbra Acute Ischemic Stroke Dataset, aimed at enhancing the early detection and segmentation of ischemic stroke using Non-Contrast Computed Tomography (NCCT) scans. Addressing the challenges in diagnosing acute ischemic stroke during its early stages due to often non-revealing native CT findings, the dataset provides a collection of segmented NCCT images. These include annotations of ischemic core and penumbra regions, critical for developing machine learning models for rapid stroke identification and assessment. By offering a carefully collected and annotated dataset, we aim to facilitate the development of advanced diagnostic tools, contributing to improved patient care and outcomes in stroke management. Our dataset's uniqueness lies in its focus on the acute phase of ischemic stroke, with non-informative native CT scans, and includes a baseline model to demonstrate the dataset's application, encouraging further research and innovation in the field of medical imaging and stroke diagnosis.
△ Less
Submitted 3 April, 2024;
originally announced April 2024.
-
GigaPevt: Multimodal Medical Assistant
Authors:
Pavel Blinov,
Konstantin Egorov,
Ivan Sviridov,
Nikolay Ivanov,
Stepan Botman,
Evgeniy Tagin,
Stepan Kudin,
Galina Zubkova,
Andrey Savchenko
Abstract:
Building an intelligent and efficient medical assistant is still a challenging AI problem. The major limitation comes from the data modality scarceness, which reduces comprehensive patient perception. This demo paper presents the GigaPevt, the first multimodal medical assistant that combines the dialog capabilities of large language models with specialized medical models. Such an approach shows im…
▽ More
Building an intelligent and efficient medical assistant is still a challenging AI problem. The major limitation comes from the data modality scarceness, which reduces comprehensive patient perception. This demo paper presents the GigaPevt, the first multimodal medical assistant that combines the dialog capabilities of large language models with specialized medical models. Such an approach shows immediate advantages in dialog quality and metric performance, with a 1.18\% accuracy improvement in the question-answering task.
△ Less
Submitted 26 February, 2024;
originally announced February 2024.
-
Vectorial Negabent Concepts: Similarities, Differences, and Generalizations
Authors:
Nurdagül Anbar,
Sadmir Kudin,
Wilfried Meidl,
Enes Pasalic,
Alexandr Polujan
Abstract:
In Pasalic et al., IEEE Trans. Inform. Theory 69 (2023), 2702--2712, and in Anbar, Meidl, Cryptogr. Commun. 10 (2018), 235--249, two different vectorial negabent and vectorial bent-negabent concepts are introduced, which leads to seemingly contradictory results. One of the main motivations for this article is to clarify the differences and similarities between these two concepts. Moreover, the neg…
▽ More
In Pasalic et al., IEEE Trans. Inform. Theory 69 (2023), 2702--2712, and in Anbar, Meidl, Cryptogr. Commun. 10 (2018), 235--249, two different vectorial negabent and vectorial bent-negabent concepts are introduced, which leads to seemingly contradictory results. One of the main motivations for this article is to clarify the differences and similarities between these two concepts. Moreover, the negabent concept is extended to generalized Boolean functions from \(\mathbb{F}_2^n\) to the cyclic group \(\mathbb{Z}_{2^k}\). It is shown how to obtain nega-\(\mathbb{Z}_{2^k}\)-bent functions from \(\mathbb{Z}_{2^k}\)-bent functions, or equivalently, corresponding non-splitting relative difference sets from the splitting relative difference sets. This generalizes the shifting results for Boolean bent and negabent functions. We finally point to constructions of \(\mathbb{Z}_8\)-bent functions employing permutations with the \((\mathcal{A}_m)\) property, and more generally we show that the inverse permutation gives rise to \(\mathbb{Z}_{2^k}\)-bent functions.
△ Less
Submitted 8 February, 2024;
originally announced February 2024.
-
Bent functions satisfying the dual bent condition and permutations with the $(\mathcal{A}_m)$ property
Authors:
Alexandr Polujan,
Enes Pasalic,
Sadmir Kudin,
Fengrong Zhang
Abstract:
The concatenation of four Boolean bent functions $f=f_1||f_2||f_3||f_4$ is bent if and only if the dual bent condition $f_1^* + f_2^* + f_3^* + f_4^* =1$ is satisfied. However, to specify four bent functions satisfying this duality condition is in general quite a difficult task. Commonly, to simplify this problem, certain connections between $f_i$ are assumed, as well as functions $f_i$ of a speci…
▽ More
The concatenation of four Boolean bent functions $f=f_1||f_2||f_3||f_4$ is bent if and only if the dual bent condition $f_1^* + f_2^* + f_3^* + f_4^* =1$ is satisfied. However, to specify four bent functions satisfying this duality condition is in general quite a difficult task. Commonly, to simplify this problem, certain connections between $f_i$ are assumed, as well as functions $f_i$ of a special shape are considered, e.g., $f_i(x,y)=x\cdotπ_i(y)+h_i(y)$ are Maiorana-McFarland bent functions. In the case when permutations $π_i$ of $\mathbb{F}_2^m$ have the $(\mathcal{A}_m)$ property and Maiorana-McFarland bent functions $f_i$ satisfy the additional condition $f_1+f_2+f_3+f_4=0$, the dual bent condition is known to have a relatively simple shape allowing to specify the functions $f_i$ explicitly. In this paper, we generalize this result for the case when Maiorana-McFarland bent functions $f_i$ satisfy the condition $f_1(x,y)+f_2(x,y)+f_3(x,y)+f_4(x,y)=s(y)$ and provide a construction of new permutations with the $(\mathcal{A}_m)$ property from the old ones. Combining these two results, we obtain a recursive construction method of bent functions satisfying the dual bent condition. Moreover, we provide a generic condition on the Maiorana-McFarland bent functions stemming from the permutations of $\mathbb{F}_2^m$ with the $(\mathcal{A}_m)$ property, such that their concatenation does not belong, up to equivalence, to the Maiorana-McFarland class. Using monomial permutations $π_i$ of $\mathbb{F}_{2^m}$ with the $(\mathcal{A}_m)$ property and monomial functions $h_i$ on $\mathbb{F}_{2^m}$, we provide explicit constructions of such bent functions. Finally, with our construction method, we explain how one can construct homogeneous cubic bent functions, noticing that only very few design methods of these objects are known.
△ Less
Submitted 16 October, 2023;
originally announced October 2023.
-
Design and analysis of bent functions using $\mathcal{M}$-subspaces
Authors:
Enes Pasalic,
Alexandr Polujan,
Sadmir Kudin,
Fengrong Zhang
Abstract:
In this article, we provide the first systematic analysis of bent functions $f$ on $\mathbb{F}_2^{n}$ in the Maiorana-McFarland class $\mathcal{MM}$ regarding the origin and cardinality of their $\mathcal{M}$-subspaces, i.e., vector subspaces on which the second-order derivatives of $f$ vanish. By imposing restrictions on permutations $π$ of $\mathbb{F}_2^{n/2}$, we specify the conditions, such th…
▽ More
In this article, we provide the first systematic analysis of bent functions $f$ on $\mathbb{F}_2^{n}$ in the Maiorana-McFarland class $\mathcal{MM}$ regarding the origin and cardinality of their $\mathcal{M}$-subspaces, i.e., vector subspaces on which the second-order derivatives of $f$ vanish. By imposing restrictions on permutations $π$ of $\mathbb{F}_2^{n/2}$, we specify the conditions, such that Maiorana-McFarland bent functions $f(x,y)=x\cdot π(y) + h(y)$ admit a unique $\mathcal{M}$-subspace of dimension $n/2$. On the other hand, we show that permutations $π$ with linear structures give rise to Maiorana-McFarland bent functions that do not have this property. In this way, we contribute to the classification of Maiorana-McFarland bent functions, since the number of $\mathcal{M}$-subspaces is invariant under equivalence. Additionally, we give several generic methods of specifying permutations $π$ so that $f\in\mathcal{MM}$ admits a unique $\mathcal{M}$-subspace. Most notably, using the knowledge about $\mathcal{M}$-subspaces, we show that using the bent 4-concatenation of four suitably chosen Maiorana-McFarland bent functions, one can in a generic manner generate bent functions on $\mathbb{F}_2^{n}$ outside the completed Maiorana-McFarland class $\mathcal{MM}^\#$ for any even $n\geq 8$. Remarkably, with our construction methods it is possible to obtain inequivalent bent functions on $\mathbb{F}_2^8$ not stemming from two primary classes, the partial spread class $\mathcal{PS}$ and $\mathcal{MM}$. In this way, we contribute to a better understanding of the origin of bent functions in eight variables, since only a small fraction, of which size is about $2^{76}$, stems from $\mathcal{PS}$ and $\mathcal{MM}$, whereas the total number of bent functions on $\mathbb{F}_2^8$ is approximately $2^{106}$.
△ Less
Submitted 26 April, 2023;
originally announced April 2023.
-
Whole-body PET image denoising for reduced acquisition time
Authors:
Ivan Kruzhilov,
Stepan Kudin,
Luka Vetoshkin,
Elena Sokolova,
Vladimir Kokh
Abstract:
This paper evaluates the performance of supervised and unsupervised deep learning models for denoising positron emission tomography (PET) images in the presence of reduced acquisition times. Our experiments consider 212 studies (56908 images), and evaluate the models using 2D (RMSE, SSIM) and 3D (SUVpeak and SUVmax error for the regions of interest) metrics. It was shown that, in contrast to previ…
▽ More
This paper evaluates the performance of supervised and unsupervised deep learning models for denoising positron emission tomography (PET) images in the presence of reduced acquisition times. Our experiments consider 212 studies (56908 images), and evaluate the models using 2D (RMSE, SSIM) and 3D (SUVpeak and SUVmax error for the regions of interest) metrics. It was shown that, in contrast to previous studies, supervised models (ResNet, Unet, SwinIR) outperform unsupervised models (pix2pix GAN and CycleGAN with ResNet backbone and various auxiliary losses) in the reconstruction of 2D PET images. Moreover, a hybrid approach of supervised CycleGAN shows the best results in SUVmax estimation for denoised images, and the SUVmax estimation error for denoised images is comparable with the PET reproducibility error.
△ Less
Submitted 28 March, 2023;
originally announced March 2023.
