Colorectal cancer is still an important cause of death from cancer. Early polyp detection can play an important role in avoiding colorectal cancer. Colonoscopy has been the most common approach for detecting polyps; however, the results of the examination rely on the expertise of the operator. Computer-aided detection systems are useful for automatic polyp segmentation because they can detect some polyps which might be missed by clinical practitioners. Polyp segmentation is faced with challenges such as the lack of clear boundaries, low background-foreground contrast, and morphological heterogeneity. Although recent deep learning-based methods aim to address the challenges of colorectal polyp segmentation, they still struggle to achieve high accuracy and efficiency. This research introduces a novel encoder-decoder segmentation architecture, the Boundary-Assisted Multiscale Contextual Network for Polyp Segmentation (BAMCN). The proposed network aims to achieve high polyp segmentation accuracy without increasing computational complexity. The suggested architecture integrates the advantages of a pre-trained encoder, our innovative hybrid multiscale feature fusion block (HMFF), multitask learning, depthwise separable convolutions, channel attention, spatial attention, and deep supervision. The hybrid multiscale feature fusion block improves the feature representation by extracting and fusing the features at various scales. This helps the model learn the finer as well as the coarser aspects of the input data, thus improving the robustness of segmentation. Multitasking learning helps in extracting boundary features in order to localize the polyps accurately. Various polyp segmentation datasets have been employed in the assessment of the proposed architecture. The findings indicate that the proposed framework significantly improves polyp segmentation compared to many leading approaches.
Skeletal muscle exhibits a hierarchical and heterogeneous architecture in which fascicles and their surrounding extracellular matrix (ECM) compose the interplay of biomechanical functions. While experimental studies indicate that muscle morphology, including ECM content and fascicle architecture, influences muscle mechanics, the direct contributions of these mesoscale features remain poorly quantified. Here, we develop a mesoscale finite element framework to investigate how fascicle number and ECM volume fraction affect passive and active muscle responses. Representative volume elements were generated using centroidal Voronoi tessellation with Lloyd's method, combined with a customized algorithm to replicate fascicle–ECM geometries. Parametric variations spanned fascicle counts from 15 to 70 and ECM volume fractions from 2% (healthy) to 20% (fibrotic or aged), covering physiological and pathological ranges. Simulation results revealed that active peak stresses remained stable across configurations (∼0.85 MPa at 10% elongation) with <1% variability, whereas passive responses were highly sensitive to ECM volume fraction. Increasing ECM content reduced passive stiffness and altered fascicle–ECM load sharing, with up to 8.36% variability across the evaluated range. Reaction force analyses demonstrated that excessive ECM, as observed in fibrosis or aging, amplifies the load carried by connective tissue, potentially impairing force transmission. Global stress transmission in both active and passive states showed a notable dependence on ECM volume fraction: average stress at the functional surface was inversely related to ECM content, particularly under active conditions. The framework links mesoscale heterogeneity to macroscopic function, bridging a critical gap in skeletal muscle modeling, with potential extensions to multiscale models of adaptation, degeneration, and therapeutic intervention.
Background Wear and wear debris generated from the bearing components of artificial hip prostheses may contribute to osteolysis, implant loosening, and eventual prosthesis failure. To address these limitations, an elastic composite based on polyurethane dispersion (PUD) reinforced with reduced graphene oxide (rGO) was investigated as a potential elastic cartilage pad (ECP) material for artificial hip prostheses. Method PUD/rGO composite specimens with different formulations were fabricated using ultrasonic dispersion and freeze-drying. The cytocompatibility of PUD-based test articles was evaluated according to ISO 10993-5. Tensile and compression tests were performed to determine the mechanical properties of the composite, and pin-on-disc friction tests were conducted under dry and phosphate-buffered saline (PBS)-lubricated conditions. The experimentally determined material properties were subsequently incorporated into a three-dimensional finite element model to evaluate the stress and displacement responses of the artificial hip prosthesis under a simplified static loading condition. Result The PUD-based test articles satisfied the ISO 10993-5 acceptance criterion, providing preliminary evidence of the cytocompatibility of the PUD substrate; however, cytotoxicity was not independently evaluated for each PUD/rGO formulation. The simplified two-stage nominal tensile representation was derived from representative points of the experimental tensile load–displacement response. The transition point and ultimate tensile point corresponded to nominal tensile stresses of approximately 0.55 and 1.17 MPa, respectively. Based on the nominal initial specimen length of 16 mm, the first-stage nominal secant modulus (Et1) and second-stage nominal slope (Et2) were approximately 4.40 and 0.33 MPa, respectively. The friction tests demonstrated load- and lubrication-dependent frictional behavior. Finite element analysis showed distinct stress and displacement responses among the investigated material configurations. Under the simplified static loading condition, the PUD/rGO-containing model exhibited limited displacement and redistributed stress among the prosthetic components, supporting the feasibility of incorporating experimentally derived PUD/rGO material properties into an elastic cartilage-pad model. Conclusion The developed PUD/rGO composite demonstrated measurable mechanical and frictional properties relevant to its preliminary evaluation as an elastic cartilage-pad material, while the PUD-based cytotoxicity results provided preliminary evidence of substrate cytocompatibility. Integration of the experimentally determined material parameters with finite element analysis demonstrated the feasibility of the proposed material–numerical evaluation framework. However, the present findings should be interpreted as preliminary because the friction tests and finite element analyses were conducted under simplified experimental and loading conditions that do not fully reproduce physiological hip-joint conditions; further validation using independently fabricated specimens and physiologically relevant testing conditions is required.
Additive Manufacturing (AM) has transformed modern manufacturing by enabling the creation of complex, customized structures with specific geometrical configurations. Despite its revolutionary impact, AM faces significant challenges such as optimizing process parameters, material customization, quality assurance, and reproducibility. Each AM process involves over 250 parameters that must be tailored for different materials, making the process labor-intensive and reliant on highly skilled professionals. To address these issues, Artificial Intelligence (AI) and Machine Learning (ML) are being explored as transformative tools. AI, particularly through deep neural networks (DNN) and artificial neural networks (ANN), has shown potential to simplify AM processes, automate decision-making, and enhance manufacturing efficiency. These models can learn from large datasets to predict optimal process parameters, thus reducing human dependency and improving product quality. The synergy between AI and AM is promising in areas such as process optimization, defect detection, material development, quality control, topology optimization, and predictive maintenance. By integrating AI-driven systems, AM can evolve into a more intelligent, data-driven production environment where parameters are optimized in real-time, improving reproducibility and reducing defects. Combining AI with AM paves the way for more efficient, innovative, and scalable manufacturing, particularly in the context of Industry 4.0.
