Final Layer Research Articles

The Effect of Different Concentration of NaOH on Mechanical Properties of Allium sativum L. Peels Thin Sheet Paper

Allium sativum L. peels were used as raw material to obtain cellulosic pulps for paper making. The peels were treated with sodium hydroxide at different concentrations (5%, 12.5%, and 20%) before the paper production. The sheets were measured for weight and thickness and evaluated for tensile, tear and bursting strength. The chemical compositions of the peels were analysed. The results indicated that the peels are suitable for papermaking due to the appropriate contents of cellulose (37.22%), hemicellulose (35.21%) and lignin (9.96%) values. The colour of the sheets varies from dark to light brown when the amount of sodium hydroxide solution was increased. It was estimated that the weight of peels was in the range of 5.0 - 6.5 % of garlic clove. All the sheet samples have uniform weight and thickness. The results showed that increasing sodium hydroxide concentration from 5% to 20% decreased tensile and bursting strength, except tearing strength. Two kinds of microscopic observations showed that peels treated with the highest sodium hydroxide concentration had a fine and thin layer of binder within the fibres. The sheets appear to be transparent with due to a decrease in the tensile and bursting strength except for the tear strength.

Novel Strategy to Enhance Human Mesenchymal Stromal Cell Immunosuppression: Harnessing Interferon-Gamma Presentation in Metal-Organic Frameworks Embedded on Heparin/Collagen Multilayers.

The immunomodulatory potential of human mesenchymal stromal cells (hMSCs) can be boosted when exposed to interferon-gamma (IFN-γ). While pretreating hMSCs with IFN-γ is a common practice to enhance their immunomodulatory effects, the challenge lies in maintaining a continuous IFN-γ presence within cellular environments. Therefore, in this research, we investigate the sustainable presence of IFN-γ in the cell culture medium by immobilizing it in water-stable metal-organic frameworks (MOFs) [PCN-333(Fe)]. The immobilized IFN-γ in MOFs was coated on top of multilayers composed of combinations of heparin (HEP) and collagen (COL) that were used as a bioactive surface. Multilayers were created by using a layer-by-layer assembly technique, with the final layer alternating between collagen (COL) and heparin (HEP). We evaluated the viability, differentiation, and immunomodulatory activity of hMSCs cultured on (HEP/COL) coated with immobilized IFN-γ in MOFs after 3 and 6 days of culture. Cell viability, compared to tissue culture plastic, was not affected by immobilized IFN-γ in MOFs when they were coated on (HEP/COL) multilayers. We also verified that the osteogenic and adipogenic differentiation of the hMSCs remained unchanged. The immunomodulatory activity of hMSCs was evaluated by examining the expression of indoleamine 2,3-dioxygenase (IDO) and 11 essential immunomodulatory markers. After 6 days of culture, IDO expression and the expression of 11 immunomodulatory markers were higher in (HEP/COL) coated with immobilized IFN-γ in MOFs. Overall, (HEP/COL) multilayers coated with immobilized IFN-γ in MOFs provide a sustained presentation of cytokines to potentiate the hMSC immunomodulatory activity.

A deep transfer learning approach for COVID-19 detection and exploring a sense of belonging with Diabetes.

COVID-19 is an epidemic disease that results in death and significantly affects the older adult and those afflicted with chronic medical conditions. Diabetes medication and high blood glucose levels are significant predictors of COVID-19-related death or disease severity. Diabetic individuals, particularly those with preexisting comorbidities or geriatric patients, are at a higher risk of COVID-19 infection, including hospitalization, ICU admission, and death, than those without Diabetes. Everyone's lives have been significantly changed due to the COVID-19 outbreak. Identifying patients infected with COVID-19 in a timely manner is critical to overcoming this challenge. The Real-Time Polymerase Chain Reaction (RT-PCR) diagnostic assay is currently the gold standard for COVID-19 detection. However, RT-PCR is a time-consuming and costly technique requiring a lab kit that is difficult to get in crises and epidemics. This work suggests the CIDICXR-Net50 model, a ResNet-50-based Transfer Learning (TL) method for COVID-19 detection via Chest X-ray (CXR) image classification. The presented model is developed by substituting the final ResNet-50 classifier layer with a new classification head. The model is trained on 3,923 chest X-ray images comprising a substantial dataset of 1,360 viral pneumonia, 1,363 normal, and 1,200 COVID-19 CXR images. The proposed model's performance is evaluated in contrast to the results of six other innovative pre-trained models. The proposed CIDICXR-Net50 model attained 99.11% accuracy on the provided dataset while maintaining 99.15% precision and recall. This study also explores potential relationships between COVID-19 and Diabetes.

Enhancing Water Quality Management in Ornamental Fish Aquariums with Pot and Bamboo Filter Systems

The filtration of aquarium wastewater composed of a pot and bamboo is constructed for the purification of microbial activities and to reduce turbidity. The aquarium wastewater is treated with sand to reduce BOD and COD; the usage of coconut husk and vetiver act to decrease the turbidity, nitrite, and pH while raising DO level. Initially, sand was sieved to produce 12 size ranges between 180 and 2000 m. The effectiveness of each range in removing turbidity, BOD, and COD through sand beds 4 or 5 cm deep were tested. The critical size (300 & 425 m) at which no solid particles could penetrate through the sand bed was observed at 3 and 4 cm in depth, respectively. Moreover, the flow of solids through sand filter beds had all particles smaller than 180 m removed from them. A 2 cm bed of sand allowed 0.2-0.5% of solids to flow through, and a 4 cm bed allowed 0.07-0.1 %. Both fine and coarse sand layers are supported by a layer of gravel. In the column approach, the thickness and flow rate of each layer were fixed. The thickness of the layers of sand and charcoal were fixed at 4 cm and 6 cm, respectively. The findings of the column analysis, the combined filter showed decreases in turbidity by 48%, COD by 52%, BOD by 30%, and increasing DO by 48%, respectively. The results reveal that absorbing nutrients can serve as bio-filters to create better conditions for fish growth. The final weight, length, and SGR of fishes in the three aquariums were significantly different after seven weeks of trial, the treated aquarium showed the highest value than the control.

Natural bone inspired core–shell triple-layered gel/PCL/gel 3D printed scaffolds for bone tissue engineering

Despite technological advancements in bone tissue engineering, it is still a challenge to fabricate a scaffold with high bioactivity as well as high mechanical strength that can promote osteogenesis as well as bear load. Here we developed a 3D printed gel-polymer multi-layered hybrid scaffold. The innermost layer is porous gel-based framework made of gelatin/carboxymethyl-chitin/nano-hydroxyapatite and is cryogenically 3D printed. Further, the second and middle layer of micro-engineered polycaprolactone (PCL) is infused in the gel with controlled penetration and tuneable coating thickness. The PCL surface is further coated with a third and final thin layer of gel matrix used for the first layer. This triple-layered structure demonstrates compression strength and modulus of 13.07 ± 1.15 MPa and 21.8 ± 0.82 MPa, respectively, post 8 weeks degradation which is >3000% and >700% than gel scaffold. It also shows degradation of 6.84 ± 0.70% (83% reduction than gel scaffold) after 12 weeks and swelling of 69.09 ± 6.83% (81% reduction) as compared to gel scaffolds. Further, nearly 300%, 250%, 50%, and 440% increase in cellular attachment, proliferation, protein generation, and mineralization, respectively are achieved as compared to only PCL scaffolds. Thus, these hybrid scaffolds offer high mechanical strength, slow degradation rate, high bioactivity, and high osteoconductivity. These multifunctional scaffolds have potential for reconstructing non-load-bearing bone defects like sinus lift, jaw cysts, and moderate load-bearing like reconstructing hard palate, orbital palate, and other craniomaxillofacial bone defects.

Lyme rashes disease classification using deep feature fusion technique.

Automatic classification of Lyme disease rashes on the skin helps clinicians and dermatologists' probe and investigate Lyme skin rashes effectively. This paper proposes a new in-depth features fusion system to classify Lyme disease rashes. The proposed method consists of two main steps. First, three different deep learning models, Densenet201, InceptionV3, and Exception, were trained independently to extract the deep features from the erythema migrans (EM) images. Second, a deep feature fusion mechanism (meta classifier) is developed to integrate the deep features before the final classification output layer. The meta classifier is a basic deep convolutional neural network trained on original images and features extracted from base level three deep learning models. In the feature fusion mechanism, the last three layers of base models are dropped out and connected to the meta classifier. The proposed deep feature fusion method significantly improved the classification process, where the classification accuracy was 98.97%, which is particularly impressive than the other state-of-the-art models.

Experimental Investigation on Reinforcement Application of Newly Permeable Polymers in Dam Engineering with Fine Sand Layers

The grinding reinforcement of fine sand layers is a difficult problem in dam engineering construction. As a new type of grouting material, permeable polymer with excellent impermeability and high strength is widely used in dam engineering. In this paper, a series of compressive tests were designed considering different grouting pressures, curing days, moisture content, and porosity of fine sand. The influence of grouting parameters and sand layer conditions on the strength of fine sand layers reinforced by permeable polymers was analyzed. SEM and XDR tests were conducted to analyze the microscopic characteristics of the grouting stone. The functional calculation model of the strength and the influencing factors was established to explore the main factors influencing grouting stones. The compressive strength of grouted stones increases rapidly from the 7th to the 14th day, reaching about 96% of the maximum strength. The degree of influence of different factors is grouting pressure > moisture content > porosity. The compressive strength of the grouted stones increases with the increase of grouting pressure and the number of curing days. The compressive strength decreases with the sand layer’s increasing moisture content and porosity.

The Performance of a Lip-Sync Imagery Model, New Combinations of Signals, a Supplemental Bond Graph Classifier, and Deep Formula Detection as an Extraction and Root Classifier for Electroencephalograms and Brain–Computer Interfaces

Many current brain–computer interface (BCI) applications depend on the quick processing of brain signals. Most researchers strive to create new methods for future implementation and enhance existing models to discover an optimal feature set that can operate independently. This study focuses on four key concepts that will be used to complete future works. The first concept is related to potential future communication models, whereas the others aim to enhance previous models or methodologies. The four concepts are as follows. First, we suggest a new communication imagery model as a substitute for a speech imager that relies on a mental task approach. As speech imagery is intricate, one cannot imagine the sounds of every character in every language. Our study proposes a new mental task model for lip-sync imagery that can be employed in all languages. Any character in any language can be used with this mental task model. In this study, we utilized two lip-sync movements to indicate two sounds, characters, or letters. Second, we considered innovative hybrid signals. Choosing an unsuitable frequency range can lead to ineffective feature extractions. Therefore, the selection of an appropriate frequency range is crucial for processing. The ultimate goal of this method is to accurately discover distinct frequencies of brain imagery activities. The restricted frequency range combination presents an initial proposal for generating fragmented, continuous frequencies. The first model assesses two 4 Hz intervals as filter banks. The primary objective is to discover new combinations of signals at 8 Hz by selecting filter banks with a 4 Hz scale from the frequency range of 4 Hz to 40 Hz. This approach facilitates the acquisition of efficient and clearly defined features by reducing similar patterns and enhancing distinctive patterns of brain activity. Third, we introduce a new linear bond graph classifier as a supplement to a linear support vector machine (SVM) when handling noisy data. The performance of the linear support vector machine (SVM) significantly declines under high-noise conditions. To complement the linear support vector machine (SVM) in noisy-data situations, we introduce a new linear bond graph classifier. Fourth, this paper presents a deep-learning model for formula recognition that converts the first-layer data into a formula extraction model. The primary goal is to decrease the noise in the formula coefficients of the subsequent layers. The output of the final layer comprises coefficients chosen by different functions at various levels. The classifier then extracts the root interval for each formula, and a diagnosis is established based on these intervals. The final goal of the last idea is to explain the main brain imagery activity formula using a combination formula for similar and distinctive brain imagery activities. The results of implementing all of the proposed methods are reported. The results range between 55% and 98%. The lowest result is 55% for the deep detection formula, and the highest result is 98% for new combinations of signals.

The transferability of transfer learning model based on ImageNet for medical image classification tasks

Transfer learning with pretrained weights is commonly based on the ImageNet dataset. However, ImageNet does not contain medical images, leaving the transferability of these pretrained weights for medical image classification an open question. The core purpose of this study is to investigate the impact of transfer learning on the accuracy of medical image classification, utilizing ResNet18, VGG11, AlexNet, and MobileNet, which are four of the most widely used neural network models. Specifically, this study aims to determine whether the incorporation of transfer learning techniques leads to significant improvements in the performance of image classification tasks, as compared to traditional methods that do not utilize transfer learning. The dataset consists of approximately 4,000 chest X-ray images with labels of healthy, COVID, or Viral Pneumonia. The final layer's output neurons of the networks architecture were revised to three to accommodate the ternary classification task. Preprocessing techniques include downsampling and normalization of the pixel values. By maintaining the same dataset and preprocessing methods, this study compares the performance of the models with and without pretrained weights. The results demonstrate that, compared to not using transfer learning, all four network models converge more quickly and achieve higher validation accuracy in the initial epochs when transfer learning is employed. Furthermore, the models exhibit higher prediction accuracy in the final test set. This study suggests that using transfer learning with pretrained weights based on ImageNet can boost the efficiency of medical image classification tasks.

Designing superoleophobic and flame retardant coatings from fly ash based on layer-by-layer approach

Abstract The superoleophobic, superhydrophilic, and flame-retardant coatings were obtained through layer-by-layer (LBL) technology using fly ash (FA) as the core. Chitosan (CS), phytic acid (PA), and fluorosurfactant (FS50) were assembled onto FA by LBL assembly using the principle of electrostatic adsorption. The CS and PA were recycled assembled on FA core for 10 layers, which was further deposited with FS50 molecules as the final layer (marked as FA@CS-PA10@FS50). Multifunctional coatings consisted of the FA@CS-PA10@FS50 powders and polymer binders were formed and could be used on a variety of substrates of various shape and size. The obtained coatings presented superoleophobicity, superhydrophilicity, and flame retardance. The contact angles (CAs) of coated substrates were all greater than 150° towards various oils (rapeseed oil, diesel, hexadecane, and dodecane), while achieved water CAs as low as 0°. The as-fabricated coatings could efficiently separate different oil/water mixtures and lotions while avoiding pollution of the separated materials. The coatings also achieved superior flame retardance, and the fire was self-extinguished within 13 s. Moreover, it also obtained excellent mechanical durability.

Material characterization uncovers sophisticated mould-making techniques of the middle-Shang period (14th-13th BCE)

Recent archaeological excavations revealed a significant amount of bronze casting moulds and cores dated to the Middle-Shang period and demonstrated it as a vital stage in the development of casting technology in Bronze Age China. This study has analyzed fifty-three mould and eight core fragments from the Taijiasi site, dated to the Middle Shang III period, using micro-CT scanning, SEM-EDS, and p-XRD. The results show that four types of materials combined in distinct ways to form three mould types at this site. Moulds for smaller bronze vessels with band-like patterns were constructed with a sand tempered material while only pattern sections were replaced with a c. 5–10 mm layer of fine silt-rich material. In contrast, moulds for larger vessels with extensive decorative patterns have a double-layered structure, comprising a fine silt-rich internal layer and a sand-grog tempered external layer. Moulds for tools and weapons were made of unprocessed loess. Casting cores display a texture similar to sand tempered moulds. This research uncovered a nuanced material selection strategy for mould manufacturing at the site of Taijiasi and highlighted the innovations of mould-making techniques in the Middle Shang period.

Axial stress strain behavior of FRP-confined rectangular rubber concrete columns with different aspect ratio

At present, the mechanical properties of FRP-confined rectangular rubber concrete columns have not been studied. To fill this gap, thirty-six concrete columns were prepared and tested under axial compression in this study. The effects of the aspect ratio (1, 1.25 and 1.5), rubber volume replacement ratios of fine aggregates (0 %, 40 % and 60 %) and FRP layer number (2 and 3) on the failure state and stress–strain relationship were investigated. The test results showed that with increasing section aspect ratio and rubber volume replacement ratio, the damage zone of the specimen increased, and the compressive strength decreased. As the number of FRP layers increasing, the damage zone and compressive strength of the specimens both showed an upward trend. Based on the test results, the stress–strain model, compressive strength model and ultimate strain model of FRP-confined rubber concrete columns were proposed. The proposed models can accurately predict the stress–strain relationship, compressive strength and ultimate strain of FRP-confined rectangular rubber concrete.

Study on Spatiotemporal Transport Characteristics of Soil Moisture in Layered Heterogeneous Vadose Zone Based on HYDRAS-3D

The heterogeneity of layered soils affects the transport processes of water in the vadose zone. However, the mechanism of soil moisture transport in the vadose zone under heterogeneous media conditions, especially the three-dimensional transport mechanism, is a frontier problem to be studied. In order to reveal the law of soil water transport in the vadose zone under heterogeneous media, this paper constructs a three-dimensional model of water transport using HYDRAS-3D (V 2.x) software through in situ tests of water transport in 3 × 3 × 4 m sample and verifies and analyzes the model. The vertical distribution and variability of soil water content and the temporal and spatial rules of soil water transversal transport at the soil–layer interface were analyzed. The results indicate that after adjusting the parameters in the HYDRAS-3D model using measured values, the simulation results are reasonable, and the model has high reliability. To represent the water content variability characteristics of the test tube profile, water movement in the vadose zone was classified as follows: (I) steady period (April–September, less affected by rainfall), (II) slow change period (January–March), and (III) rainfall rapid change period (June, when the impact of rainfall on the test cylinder was greater). The two largest values of the soil water potential variation gradient, 19.9 and 17.8 cm/d, were observed in the silty clay layer of the test cylinder, and the influence of evaporation and infiltration in the silty clay layer was most notable at the interface. The lateral transport of soil water at the interface was influenced to a certain degree by the layered heterogeneity, wherein the fine sand layer exhibited the most pronounced impact. This was followed by silt and silty clay layers, with the silt layer showing the lowest degree of influence. The research results can provide scientific reference for the rational planning of water resources in the Yinchuan Plain.

A grinding force model and surface formation mechanism of cup wheels considering crystallographic orientation

As typical two-phase materials, tungsten heavy alloys (WHAs) are widely used in industry owing to their excellent mechanical properties, which also challenge high-quality machining. Grinding can lead to a better machining quality. Prediction and evaluation of grinding forces are essential for grinding quality control and process optimization. This study proposes a flow stress model for the WHA considering the strain hardening, strain rate hardening, and thermal softening effects. The difference between the two phases of WHA was confirmed by calculating the Taylor factors, where the average Taylor factor of the W phase was 3.011 and that of the matrix phase was 3.414. Therefore, dislocations are more likely to be generated and aggregated in the matrix phase during grinding process. The mechanism of subsurface formation during the grinding of WHA was analyzed by transmission electron microscopy. The results show that the plastic deformation layer consists of the fine grain layer, the high density dislocation zone and the substrate. A grinding force model for cup wheels in the vertical-spindle face grinding of WHA considering the grain orientation was developed, and the error between the model and experimental value was within 10%. This model can provide an in-depth understanding of the effects caused by the difference between the two phases during the grinding process, thus provide a theoretical basis for the realization of efficient and low-damage grinding of WHA and other composite materials.

DMF-Net: a deep multi-level semantic fusion network for high-resolution chest CT and X-ray image de-noising

Medical images such as CT and X-ray have been widely used for the detection of several chest infections and lung diseases. However, these images are susceptible to different types of noise, and it is hard to remove these noises due to their complex distribution. The presence of such noise significantly deteriorates the quality of the images and significantly affects the diagnosis performance. Hence, the design of an effective de-noising technique is highly essential to remove the noise from chest CT and X-ray images prior to further processing. Deep learning methods, mainly, CNN have shown tremendous progress on de-noising tasks. However, existing CNN based models estimate the noise from the final layers, which may not carry adequate details of the image. To tackle this issue, in this paper a deep multi-level semantic fusion network is proposed, called DMF-Net for the removal of noise from chest CT and X-ray images. The DMF-Net mainly comprises of a dilated convolutional feature extraction block, a cascaded feature learning block (CFLB) and a noise fusion block (NFB) followed by a prominent feature extraction block. The CFLB cascades the features from different levels (convolutional layers) which are later fed to NFB to attain correct noise prediction. Finally, the Prominent Feature Extraction Block(PFEB) produces the clean image. To validate the proposed de-noising technique, a separate and a mixed dataset containing high-resolution CT and X-ray images with specific and blind noise are used. Experimental results indicate the effectiveness of the DMF-Net compared to other state-of-the-art methods in the context of peak signal-to-noise ratio (PSNR) and structural similarity measurement (SSIM) while drastically cutting down on the processing power needed.

A novel technique for CAD-CAM assisted digital ocular prosthesis.

There are various techniques available for constructing a custom ocular prosthesis. The present technique report describes a digital workflow for constructing an ocular prosthesis using computed tomography integrated with digital technologies. The outer region of the healthy eyeball was segmented to produce the top surface of the prosthesis, while the contour and bottom surface were segmented from the tissue bed of the anophthalmic socket. The iris position was determined by tracing the optical nerve and confirmed by superimposing the patient's facial scan onto the soft tissue model of the computed tomography. Using these parameters, a standard tessellation language file for the ocular prosthesis was generated and 3D printed. The iris was then printed via UV technology using digital photographs of the patient's contralateral eye; characterization of the sclera and the final layer of clear acrylic resin were done conventionally.

An infiltration model for inclined covers with consideration of capillary barrier effect

Covers with capillary barrier effect (CCBEs) are widely used as surface cover systems in landfills, mine tailings and water-sensitive soil slopes to prevent rainwater from infiltrating into the protected materials below. In the engineering design, the water-prevention performance of the CCBEs must be quantificationally evaluated according to the local meteorological conditions. In this study, according to the moisture distribution during infiltration, the fine-grained soil layer was divided into saturated, transition and initial zones. Capillary barrier effects were analyzed by considering water migration and redistribution at the interface between fine and coarse-grained layers. Then, the total time, including the time to reach the interface (tHf), the time for overcoming the capillary barrier effect (tc) and the time for passing through the coarse-grained layer (tg), was determined to quantificationally evaluate the water-prevention capacity of the inclined CCBEs under different rainfall conditions. For the case studies selected from literature, the calculated results by the proposed model agree well with the measured data. Finally, some parameters and the factors influencing the capillary barrier effect were further analyzed.

Thickness, Adhesion and Microscopic Analysis of the Surface Structure of Single-Layer and Multi-Layer Metakaolin-Based Geopolymer Coatings

This work is focused on creating coating layers made of a metakaolin-based geopolymer suspensions (GP)-formed Al matrix modified using H3PO4 acid with Al(OH)3 in isopropyl alcohol, named GP suspension I, and H3PO4 acid with nano Al2O3 in isopropyl alcohol, named GP suspension J. The selected GP suspensions were applied on aluminum and steel underlying substrates as single-layer coatings and multi-layer coatings, where multi-layer coatings included three and five layers that were polymerized by a curing process. Curing was divided into two types with every layer curing process and final layer curing process. For both GP suspensions I and J, the effect of the number of layers and the type of substrate on adhesion was investigated. The prepared samples on underlying substrates were characterized on the microscopy analysis including SEM for high-resolution images and 3D laser confocal microscopy (CLSM) for the 3D visualization of the coatings structure. Microscopy analysis showed structural defects such as porosity, cracks and peeling, which increase with a greater number of applied layers. However, these defects were only evident on a micro scale and did not seem to be fatal for the performance of the surface stability. The EDS mapping of the prepared layer showed inhomogeneity in the distribution of elements caused by the brush application. A grid test and thickness measurement were performed to complete the microscopy analysis. The grid test confirmed a very high adhesion of GP coatings on the aluminum substrate with a rating of one (only in one case was there a rating of two) and a lower adhesion on the steel substrate with the most frequent rating of three (in one case, there were ratings of two and one). The thickness measurement proved a noticeably thicker thickness of the prepared layer on the Fe substrate compared to the Al substrate by 20%–30% in the case of suspension I and by 70%–10% in the case of suspension J. The thickness of the layer also showed a dependence on the method of application and curing, as a thicker layer was always achieved when curing after the final layer of the GP suspension, compared to curing after each applied layer. The resulting single-layer and multi-layer thicknesses ranged from approx. 7 to 30 µm for suspension I and from approx. 3 to 11 µm for suspension J. A non-linear increase in thickness was also evident from the thickness measurement data.

DenSplitnet: Classifier-invariant neural network method to detect COVID-19 in chest CT data

Objective:COVID-19 has made an unprecedented impact on humanity. The Healthcare sector, in an effort to curb COVID-19, could leverage Artificial Intelligence (AI) to its aid, especially in diagnosing it through the classification of Chest-CT scans. However, data scarcity plagues the medical domain. Therefore, any AI solution must be capable of learning from limited data. Also the resulting AI could relieve radiologists from exhaustion and aid medical practitioners as a valuable diagnostic tool. Methods:Our proposed model DenSplitnet uses Dense blocks to learn image features, Self-Supervised Learning for pre-training to learn the context, and a novel two-way split branch at the final classification layer for classifier-invariant generalization ability. Results:DenSplitnet achieves state-of-the-art performance on four benchmark chest-CT scan datasets for COVID-19. The model performs well on the Sars-cov-2 ct-scan dataset. The Test accuracy is 91.92%, F1-Score is 91.30%, Precision is 96.55%, and Recall is 87.04%. The model achieves test accuracy of 86.17%, Precision of 82.94%, Recall of 83.72%, and F1-Score of 83.23% on the Sars-cov-2 ct-scan multiclass dataset. The model obtains a test accuracy of 86.21%, an F1-Score of 83.91%, and an AUC of 0.95 in the Covid-ct-dataset. The model obtains a test accuracy that is 73.11% and an F1-Score of 70.97% in the TransferLearn-ct dataset. Conclusion:The findings support the practical usability of the DenSplitnet as a technological AI assistance to radiologists and other medical professionals, alongside Grad-CAM plots for explainable AI and the theoretical examination of the classifier-invariant generalization capacity of the network. The healthcare sector can benefit from this technology in a number of ways.

Hafting adhesives from the final Later Stone Age layers at Sehonghong, Lesotho: a case study and commentary on the use of imported resources by nineteenth-century San

ABSTRACT There is longstanding evidence in the archaeological record of southern Africa for the application of hafting adhesives in the production of composite tool technology. Despite the limitations in the preservation of organics in archaeological contexts, the preservation of these materials can inform us about aspects of technological production and function, including hafting arrangements, adhesive production and resource acquisition. In some instances where we see the prevalence of organic materials that are at odds with the ecological landscape, these same materials can also elucidate social interactions between cultural groups. This paper presents the results of the analysis by microscopy and gas chromatography-mass spectrometry (GC-MS) of hafting residues from stone tools from the ceramic final Later Stone Age assemblages at Sehonghong, Lesotho. It argues that an adoption of novel resources including Pinus sp. resin, castor oil and paraffin by nineteenth-century San facilitated the continued existence of hunter-gatherer lifeways in increasingly marginalised circumstances amid tense relationships between the last hunter-gatherers of the Maloti-Drakensberg, newly arriving Basotho communities and European colonists.