Medical image segmentation is a key technology in computer-aided diagnosis systems. However, the accuracy of the existing segmentation models often falls short in practical applications. To improve the accuracy of the model, this paper proposes a novel U-shaped network called MMP-Net, meticulously crafted for medical image segmentation tasks. The MMP-Net incorporates three core modules: the multi-kernel convolution module, which enhances the multi-receptive field representation of the model; the multi-layer-guided channel attention module, which combines the features from different encoder layers, and the combined features are used to guide the channel attention; the phase-guided Laplacian convolution module, which leverages the boundary sensitivity of Laplacian convolution kernels to effectively capture edge gradient changes and local detail textures in images. The proposed MMP-Net has been validated with three metrics (Dice, HD95, and IOU) across five public medical image datasets (ISIC2017, ISIC2018, BUSI, COVID-19, PH2). Experimental results demonstrate that the MMP-Net outperforms other popular models in all these three metrics with moderate model parameters (2.03M) and computation (5.36G FLOPs). This achievement offers an efficient and accurate solution for medical image segmentation tasks, making it particularly suitable for mobile healthcare and edge computing scenarios. The source code will be available at https://github.com/liyiwei-png/MMP-Net.git
The paper presents the patterns of change of the electromagnetic radiation absorption coefficient values of in the frequency range of 2.0–17.0 GHz for an aluminum-foiled polymer film onto which three-dimensional spiral elements formed from aluminum foil with a thickness of 20.0 μm are attached. These patterns are presented depending on the diameter of these elements. Taking these patterns into account, the authors developed a technology for the manufacture of multilayer microwave absorbers based on three-dimensional spiral elements made of aluminum foil. This technology involves thermopressing the construction in the form of identical fragments of a synthetic non-woven fibrous material, between which 3D spiral elements made of aluminum foil are distributed, and adhesively bonding the resulting structure to a fragment of aluminum-foiled polymer film. Absorbers manufactured in accordance with the developed technology are characterized by a wider effective absorption band (12.0 GHz) compared to similar ones. They are also flexible and lightweight. This makes them suitable for use in creating partitions for zoning anechoic chambers or for covering the walls of these chambers.
This study investigates the structural, electrical properties, and impedance characteristics of thin graphite-like carbon films deposited by electron-beam evaporation and annealed in vacuum. Raman spectroscopy reveals that the films consist mainly of sp2-hybridized carbon with graphite-like clusters of approximately 9.5 nm. Current-voltage measurements indicate space-charge-limited conduction and trap states. Impedance spectroscopy shows two overlapping relaxation processes, modeled with an equivalent circuit of two parallel branches with resistive and constant phase elements. The calculated effective capacitances were Ceff 1 = 4.09*10-9 F and Ceff 2 = 7.87*10-8 F, with corresponding relaxation times of approximately tau eff 1 = 1.12*10-6 sand tau eff 2 = 3.6* 10-5 s respectively. The observed low-frequency nonlinear dielectric response exhibits strong dependence on the applied DC bias, leading to a pronounced decrease in the real part and increase in the imaginary part of the dielectric permittivity. This bias-controlled behavior suggests great potential for implementation in tunable capacitors and frequency-dependent energy storage components. The obtained results demonstrate the interplay between conduction and polarization processes in nanostructured carbon films and validate the relevance of the cluster-based structural model.
The current–voltage characteristics (IVCs) of anodic TiO2 films in a thin-film structure (Carbon paste/TiO2/Ti/Al) were investigated in the temperature range of T = 80–300 K with bias voltages from −0.5 V to +0.5 V. Anodic oxide film, with a thickness of 14 nm, was obtained by electrochemical oxidation of Ti at a voltage of 10 V. The obtained data for various temperatures showed that the IVCs in the forward (negative on the Ti electrode) and reverse (positive on the Ti electrode) bias of the thin film structure are not symmetrical. Based on the analysis, three temperature ranges (sections) were identified in which the IVCs differ in their behavior. Examination of the IVCs revealed that the conductivity mechanism in Section I (temperature range from 298 to 263 K) is determined by the Space Charge Limited Current (SCLC). Section II, in the temperature range from 243 to 203 K, is characterized by the onset of conductivity involving donor centers, in the case where the concentration of electrons on traps is significantly higher than the concentration of electrons in the conduction band. In Section III, within the temperature range from 183 to 90 K, the conduction mechanism is the Poole–Frenkel process involving donor centers. These donor centers are located below the level of traps in the forbidden band. The results obtained indicate that anodic TiO2 is an n-type semiconductor, in the bandgap of which there are both electron traps and donor centers formed by anionic (oxygen) vacancies. The different behavior of the characteristic energy with different sample biasing in the case of the Poole–Frenkel mechanism indicates a two-layer structure of anodic TiO2.
