Breast cancer is a worldwide concern, emphasizing the importance of early diagnosis for effective treatment. Convolutional networks (CNN) have long dominated medical vision tasks. Recently, vision transformers (VIT) have sparked interest in computer-aided diagnosis (CAD). Nonetheless, when ViTs are trained from scratch, they exhibit poorer performance than (CNNs), particularly in low-data scenarios. This discrepancy is even more pronounced in tasks involving medical image datasets, where data scarcity is a significant challenge. To this end, we proposes a Vision Transformer (ViT) model that has been pre-trained on extensive datasets, including (ImageNet) and is applied directly to sequences of mammography breast image patches for the classification of benign and cancerous tissue. The model was thoroughly assessed using the DDSM dataset containing 5970 benign and 7158 malignant Masses, yielding remarkable results, with an accuracy of 99.96
Melanoma is a pathology that poses a risk to the health of the global population, leading to various complications from skin lesions that can result in patient death if not detected early. Medical imaging diagnosis is an increasingly used procedure in current literature, making the clinical investigation process more effective for high-precision imaging diagnosis. Computational methods have been employed to assist medical diagnosis through Computer-Aided Diagnosis (CAD) systems. This study introduces a new approach called SDA-Detection Melanoma for fully automatic detection and segmentation of melanomas in dermoscopy examination images. In this work, different deep-learning networks were used for melanoma detection, combined with the use of fine-tuning and computational methods based on Parzen windowing, clustering, and region growth for melanoma region segmentation. The results were quite satisfactory, achieving a high accuracy rate of 96.39
Tourism has been an essence for the economic growth. The tourism has been growing steadily and application of technology in tourism has been ever-increasing. The tourism organizations have been putting the technology to forecast and meet the demand. Machine learning has been very helpful for such organizations in understanding the tourism statistics. In this paper, a bibliometric analysis was carried out using Dimensions as the database and VOSviewer as the analytical tool. A total of 271 documents were analyzed. Nearly 18
Cholangiocarcinoma (CC), commonly known as choledochal cancer, is the second most frequent type of liver cancer. It primarily affects the elderly population; however younger people are also being diagnosed with it. Hyperspectral imaging (HS) is a spectroscopy-based technique that consists of many images in a band of neighboring spectra, and the reflection of spectra of all pixels is reconstructed to obtain three-dimensional hypercube data. The goal of this study is to provide a solution for semantically segmenting microscopic HS images of choledochal tissues acquired from 174 patients. Principal Component Analysis was used to preprocess the HS images, followed by min max normalization. The proposed system is based on the UNet architecture. The proposed system achieved an accuracy of 68.09
Inpainting is the process of filling in missing or corrupted parts of an image using the information that is present in the rest of the picture. One application of inpainting is the recovery of incomplete depth maps. Depth maps represent the distance from a sensor to each point in the scene. They are becoming increasingly important in applications such as 3D object reconstruction, depth map completion for control of autonomous vehicles, and others. This paper presents three models for inpainting incomplete depth maps and an anisotropic metric. Our inpainting model is the solution of a second-order degenerate partial differential equation. The three used models are variations of the infinity Laplacian: the biased infinity Laplacian, the balanced biased infinity Laplacian, and the double-balanced infinity Laplacian. The contribution of this paper is threefold. First, we propose two new models that are variations of the infinity Laplacian. Second, we evaluate these models on the public data set KITTI Depth Completion Suite. Results show that the considered biased infinity Laplacian and the anisotropic metric perform better than the other proposed models and contemporary models.
The proliferation of data generated and stored through social media has experienced a significant surge over the past decade. Consequently, the analysis and interpretation of such data have emerged as valuable sources of insights across diverse contexts, serving as aids for researchers and businesses in making informed decisions. However, the data is widespread, stemming from diverse sources with distinct formats, and is generated at a rapid pace. These characteristics collectively contribute to the intricacy of extracting knowledge from this data, transforming the process into one that is both complex and resource-intensive. The central scientific contribution of this paper lies in the formulation of a social media data integration model, built upon the foundation of a data warehouse. This model is designed to alleviate the computational costs associated with data analysis while concurrently facilitating the application of techniques aimed at discovering meaningful insights. Notably, this study differentiates itself from existing literature by concentrating on both the Facebook and Twitter social media platforms. Additionally, we introduce a model that covers data acquisition, transformation, and loading processes, enabling the extraction of valuable insights even when the data’s complexity surpasses human understanding.
The proposed study presents the design and implementation of an innovative home automation system leveraging Augmented Reality (AR) technology to enhance user interaction and control over electrical appliances. Augmented Reality offers a dynamic platform for creating personalized and interactive content, aligning with the goal of making home automation more intuitive and accessible. The primary objective is to develop a user-friendly application that generates AR buttons for controlling various electrical appliances. These AR buttons are displayed as 3D models on mobile devices, accessible via web browsers. The application’s data is securely transmitted to a server, which in turn sends control commands to an ESP-32 microcontroller, effectively replacing traditional switchboards. A significant innovation is the use of image-based targets to enable seamless interaction with virtual buttons. Users can effortlessly control lighting, door locks, and air conditioning through intuitive virtual interfaces, improving the overall efficiency and convenience of home automation. The primary focus lies on the integration of AR technology into a smart home ecospace to create a more intuitive and interactive user experience and contributes to the ongoing efforts to simplify and enhance daily life through the convergence of digital technology and the physical environment.
Radiological chest exams, such as chest X-rays, are critical in the fight against the COVID-19 pneumonia outbreak, which is caused by the coronavirus strain SARS-Cov-2. This study looks into classification models to distinguish chest X-ray images based on Radiomics features in order to understand the unique radiographic characteristics of COVID-19. This study used datasets consisting of 136 segmented chest X-rays to train and test the categorization algorithms. Using the Pyradiomics collection, first and second-order statistical texture characteristics were retrieved from the right (R), left (L), superior, middle, and bottom lung zones for each lung side. For feature selection, data was separated into training (80
Mammogram image analysis is an important domain in the image-based diagnosis process. It is a trusted modality of non-invasive imaging for detecting tumour regions in the breast mass. In this context, the poor contrast in the mammogram images is a key challenging issue. To address the issue, we suggest an Improved Gradient based Joint Histogram Equalization (IGJHE) method for enhancing the poor contrast in the mammogram images while restoring the structural information. The basic idea is to preserve the multi-scale structural details using an improved Gradient filtering approach. Further, to enrich the performance of the histogram equalization, we incorporated the spatial information of each data point in the enhancement process. The suggested method is evaluated using a number of mammogram images from standard datasets. The performance of the suggested approach is validated in comparison to the state-of-the-art schemes. The quantitative assessment is performed using some extensive validation metrics. The results indicate the outperformance of the suggested approach.
The Philippines is experiencing a booming industrial sector, encompassing construction, automotive, manufacturing, petrochemicals, energy production, waste treatment facilities, and more. However, this growth brings forth a critical challenge: the need to prioritize worker safety, environmental preservation, and uninterrupted operations amid the presence and utilization of hazardous gases in production, manufacturing processes, and their by-products. This paper introduces a systems engineering approach tailored to the complexities of Philippine industries. It presents a comprehensive framework for the design, implementation, and performance evaluation of gas detection systems. These systems play a pivotal role in ensuring that the country’s manufacturing and production processes operate at peak efficiency, all the while optimizing worker safety, enhancing the efficiency of equipment and processes, and steadfastly upholding environmental integrity. By embracing systems engineering principles, this approach forms the bedrock for the meticulous design of gas detection systems, contributing to a safer, more efficient, and environmentally responsible industrial landscape in the Philippines.
Feature selection stands out to be an important preprocessing step that is used to handle the uncertainty and vagueness in the data. In recent times, the minimum Redundancy and Maximum Relevance (mRMR) approach has been proven to be effective in obtaining the irredundant feature subset. Owing to the generation of voluminous datasets, it is essential to design scalable solutions using distributed/parallel paradigms. MapReduce solutions are proven to be one of the best approaches to designing fault-tolerant and scalable solutions. This work analyses the existing vertical partitioning MapReduce approaches for mRMR feature selection and identifies the limitations thereof. In the current study, we proposed VMR_mRMR, an efficient vertical partitioning-based approach using a memorization approach thereby overcoming the extant approaches limitations. The experiment analysis says that VMR_mRMR significantly outperformed extant approaches and achieved a better computational gain (C.G). We also conducted a comparative analysis with the horizontal partitioning approach HMR_mRMR to assess the strengths and limitations of the proposed approach.
In this research paper, the set union knapsack problem, a more complex variant of the binary knapsack problem, is studied. A hybrid algorithm is proposed, combining machine learning with an iterative search-descent method. This hybrid algorithm is composed of three phases: learning, exploitation, and exploration strategies, which work together to provide high-quality solutions. The algorithm’s effectiveness is demonstrated through a computational analysis of benchmark instances from the literature, highlighting its competitiveness against existing methods.
This study presents a novel hybrid approach for the classification of COVID-19 cases employing a combination between radiomics features and Res-Net-Darknet19 architecture based transfer-learning model to enhance the accuracy of COVID-19 diagnosis. The primary phase of the model demonstrated excellent performance, with 97.33
Retinopathy of Prematurity (ROP) is a vasoproliferative disorder influencing the retinal health of premature neonates with low birth weight. Delay in diagnosis of ROP can lead to permanent blindness of the preterm. This study aims to evaluate the efficiency of the features extracted from three pre-trained networks in the prediction of ROP. Examining the characteristics extracted from different layers of a CNN has the potential to provide deeper insights into the prediction of ROP. Three pre-trained networks such as ConvNeXt, VGG-16, and Inception V3 Convolutional Neural Networks (CNNs), are used to obtain the high-level features from the preterm fundus images and are classified using two methods. The first CNN based classifier, and the second is the Support Vector Machine (SVM). It is observed that the ConvNeXt pre-trained network with SVM classifier outperforms other networks with an 91.6
A Superstreet (or RCUT) represents a contemporary approach to intersection design with an unconventional geometric layout. This intersection model has gained significant traction owing to its capacity to mitigate control delays and enhance traffic safety by facilitating traffic flows on the major road without encountering traffic flows from the opposite direction. This study delves into the environmental ramifications of traffic associated with Superstreets, while also proposing an optimization strategy based on an ecology-driven performance metric. Leveraging Genetic Algorithms (GA), we formulate a mathematical framework for controlling Superstreets under fixed-time control conditions. The frequency of stops on the RCUT approach is assessed with varying penalty factors (K), contingent upon traffic and geometric parameters specific to each demand. To simulate the field-like Superstreets, we encode it into the VISSIM simulation model, utilizing this approach to evaluate both Genetic Algorithms and Webster signal plans. Webster's method, a widely employed technique for devising standard signal plans at signalized intersections, serves as a benchmark, having been extensively utilized in practical applications. Employing a range of traffic demand scenarios, we conduct comprehensive calculations. Our findings demonstrate notable enhancements in evaluation metrics, underscoring the viability of the proposed methodology.
This study aims to assess the economic implications of super-slow steaming on the operational performance of merchant vessels. Super-slow steaming is an area with limited prior research, and our approach leverages Artificial Intelligence (AI) to delve into key aspects of overall ship efficiency, specifically focusing on speed reduction and optimization. Our research collected data from the logbooks of 20 selected merchant vessels, harnessing AI to gather a total of 2226 data points concerning various study variables. The study’s findings offer valuable insights for shippers seeking to optimize fuel consumption, enhance overall vessel performance and efficiency. Furthermore, the research sheds light on the distinctions between slow steaming and super-slow steaming, along with their implications for environmental and economic factors. It’s important to note that this study has a geographical scope limited to a specific international route and a specific type of merchant vessel. The study ultimately concludes that strategic speed reduction has a positive impact on the total fuel consumption, cost-effectiveness, and operating efficiency of merchant vessels.
In-service inspection is performed while the equipment in operation, planned maintenance or breakdown. It entails evaluating the state, functionality, and efficacy of the operational equipment. The pressure vessel and piping are critical components in the process industry. This research study aims to identify and analyze barriers for pressure vessel and piping in-service inspection. This study covers process industry with conventional inspection method/equipment and to overcome this barrier advanced inspection technique or equipment to be established which increases the cost of inspection or equipment downtime. To evaluate and rank various inspection barriers, a Decision-Making Trail and Evaluation Laboratory (DEMATEL) based approach is developed and used. Twenty essential inspection barriers for pressure vessel and piping inspection are chosen based on the literature review and the opinions of experts. This study will be supportive for the inspection organization and user/owner of the equipment to implement the proactive measures. The analysis's findings indicate that the lack of manufacturer design parameters (B3), material specifications (B4), and unavailability of the nameplate (B17) are the most important variables. Cause and effect values and the effect of individual barriers over the other barriers are analyzed in this research.
The paper presents an approach to the optimal design of actively lubricated hybrid bearings with non-circular bore geometry. The novelty of the proposed approach is the automated and joint assessment of the performance of the active lubrication system and the nonlinear dynamic behavior of the rotor on the designed bearings, also considering their non-circular bore shape. The optimization problem is solved by several evolutionary algorithms, including MOGA, NSGA-II and MOPSO to compare their efficiency. Also, a new variation of MOPSO algorithm with the modified method for selecting leading solutions was proposed to improve the quality of solution. Since the calculation results are represented by a three-dimensional Pareto front, several solutions with different specified properties are analyzed, as well as an approach to selecting a single solution using the linear convolution method. The results of comparative analysis of the applying the considered evolutionary algorithms are presented. In particular, the time spent on performing calculations, the degree of minimization of objective functions, as well as the variety of Pareto-optimal solutions obtained were compared. The results obtained allow giving recommendations regarding the application of the considered approach and algorithms to the optimal design of active fluid film bearings.
Plant diseases represent a critical challenge to global food security and agricultural productivity. Timely and accurate disease prediction is essential for effective management. Recent advancements in deep learning techniques, particularly in computer vision, have shown remarkable potential in the context of plant disease prediction. This paper presents a comprehensive exploration of plant disease prediction using cutting-edge deep learning algorithms. Our proposed approach encompasses key stages, beginning with image acquisition, pre-processing, augmentation, and disease prediction. The plant leaf photos are used to train cutting-edge deep learning models, such as EfficientNet with Attention layer and DenseNet. These models are adept at learning the nuanced features that distinguish healthy and diseased plant leaves. To ensure the quality and consistency of our dataset, the employ pre-processing techniques, including image enhancement. Large-scale datasets are utilized to train convolutional neural networks (CNNs), allowing our models to glean insights from a diverse array of healthy and affected plant specimens. The result is a robust predictive framework that empowers farmers to identify diseases affecting their crops. Furthermore, our research introduces the application of image processing, with a particular focus on the EfficientNet and DenseNet Model. By isolating individual leaves within field images, the challenge is detecting diseases within complex visual contexts. This innovative approach holds great promise in enhancing the accuracy and reliability of disease prediction. In the relentless pursuit of agricultural sustainability and food security, the fusion of deep learning and plant pathology emerges as a beacon of hope. This work underscores the pivotal role of technology in safeguarding our agricultural heritage.
Chronic Kidney Disease (CKD) results from kidney injuries and leads to a decline in renal function. This study used unsupervised machine learning techniques, focusing on the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) clustering method. After applying DBSCAN to clinical and laboratory data from patients with or at risk of CKD, we employed Principal Component Analysis (PCA) and t-distributed Stochastic Neighbor Embedding (t-SNE) to visualize the findings. The findings underscore the complexity of analyzing health data. However, the DBSCAN method effectively categorized patients into distinct groups based on the severity of their chronic kidney disease (CKD), allowing for a more precise differentiation of patients according to their risk levels associated with different stages of the disease.
