Introduction/Objectives Cadmium (Cd) is a ubiquitous environmental toxicant that causes adverse effects on female reproductive organs. Curcumin (Cm) has potent antioxidant and anti-inflammatory properties, but it has low solubility and bioavailability. The present study aimed to evaluate and compare the protective effects of native Cm and polycaprolactone/polyvinyl alcohol-based curcumin nanoparticles (PCL/PVA/Cm NPs) against Cd-induced ovarian and uterine toxicity in female rats through histopathological, semi-quantitative, and morphometric analyses. Methods Forty-eight adult female rats were randomly assigned to six groups (n = 8/group): control, Cd, Cm, Cd + Cm, PCL/PVA/Cm NPs, and Cd + PCL/PVA/Cm NPs. Treatments were given orally for four weeks. Ovarian and uterine tissues were collected at the end of the experimental period for histopathological analysis, semi-quantitative histological scoring, and morphometric evaluation. Results Cd exposure induced marked ovarian and uterine histopathological changes, including follicular atresia, degeneration of corpora lutea, vascular congestion, thinning of the endometrial epithelium, glandular atrophy and inflammatory cell infiltration, and reduced endometrial and myometrial thickness. Interestingly, treatment with free Cm maintained normal granulosa cell proliferation and uterine morphology. Administration of Cm along with Cd revealed partial histological recovery compared with the group treated with Cd alone. On the other hand, PCL/PVA-encapsulated Cm showed the most significant histological improvements with signs of active folliculogenesis and normal uterine structure. Interestingly, the group treated with Cd + PCL/PVA/Cm NPs showed almost complete recovery of ovarian and uterine structure with histological scores comparable to those of the control groups. Discussion The significant restoration of ovarian and uterine tissues and folliculogenesis in PCL/PVA/Cm NPs-treated animals could be attributed to the enhanced bioavailability and sustained release of Cm. Conclusion PCL/PVA/Cm NPs exhibited greater protective and restorative effects against Cd-induced reproductive toxicity than native Cm, indicating that nanoparticle-mediated delivery of curcumin may represent an effective strategy for protecting the female reproductive system against heavy metal-induced damage.
IntroductionThe prevalence of mental health disorders such as anxiety, depression, and stress is on the rise. Persistent anxiety adversely affects individuals' quality of life and overall productivity. Early detection employing novel methodologies can enhance the effectiveness of intervention strategies. This study utilized cardio-spatiotemporal features to accurately identify anxiety states. MethodsBased on the dataset, 22 participants were continuously monitored over a 24-hour period to collect cardiac and locomotor data. Variables such as heart rate, R-R intervals, and step-related metrics were recorded. Subsequently, machine learning techniques—including K-Nearest Neighbors (KNN), Linear Discriminant Analysis (LDA), and Support Vector Machines (SVM) were employed to classify anxiety levels, and the performance of these models was evaluated using cross-validation methods. ResultsIn the Fine Tree and Boosted Tree methods, the area under the curve (AUC) outputs were 76% and 80%, respectively, while the other algorithms demonstrated significantly lower accuracy. DiscussionThe findings of this study demonstrated an association between cardio-spatiotemporal features and anxiety states. Furthermore, the application of machine learning techniques provided a robust, balanced approach to classifying anxiety. ConclusionThis study used machine learning to classify and diagnose anxiety by analyzing both muscle and heart characteristics together. Results showed that both traits indicate anxiety behaviors, with certain models achieving up to 76% accuracy. Future research should check anxiety levels beforehand and improve data collection to distinguish normal heart rate changes from those related to anxiety.
IntroductionWith the proliferation of communication applications, security has become a critical concern for the transmission and storage of files. Encryption is required to protect data communication in computer networks. Asymmetric encryption has the disadvantage of high computational complexity, making it unsuitable for resource-constrained communication applications. The main advantage of symmetric encryption is its lower computational cost compared with asymmetric encryption. However, existing symmetric algorithms vary significantly in efficiency and security, necessitating comparative analysis to determine their practical applicability. MethodsTherefore, it is necessary to evaluate the performance of common symmetric encryption algorithms to establish their suitability for various efficiency and security requirements. This study assessed the performance of three symmetric encryption techniques: Advanced Encryption Standard (AES), Data Encryption Standard (DES), and Blowfish for the security of plaintext files, and the simulation program was implemented using the Python programming language. ResultsThe avalanche effect, execution time, and throughput performance of AES, DES, and Blowfish were evaluated based on 0.5MB, 1MB, 2MB, 5MB, 10MB, and 20MB of text files. A key novelty of this study lies in the systematic comparison of these algorithms under uniform conditions across file sizes to evaluate their real-world suitability. DiscussionThe novelty of this research lies in its systematic, file-size-based comparative evaluation of symmetric encryption algorithms using uniform testing conditions. By employing consistent implementation methods and benchmarking metrics, this study offers practical insights into how these algorithms perform under real-world data loads, providing developers and system designers with a valuable tool for selecting the most appropriate encryption algorithm for their specific use cases. ConclusionAES performed better than DES and Blowfish algorithms for security performance, while Blowfish performed better for both time and throughput.
BackgroundFluorescent imaging of nanoparticles (NPs) in organs and tumors is an important part of diagnosing and treating cancers. Our study investigated the differences in imaging depending on the accumulation of NaYF4 upconversion nanoparticles (UCNPs) in rat organs (heart, lung, liver, spleen, kidneys) and tumor (liver cancer model), based on their shell type, such as human serum albumin (HSA), HSA with folic acid (HSA+FA), or HSA, FA, and the cyanine dye Cy3 (HSA+FA+Cy3). MethodsWe performed simultaneous rapid imaging of NPs using a standard microscope with field-excited luminescence excitation. Histological sections were then prepared according to standard methods, followed by hematoxylin and eosin (H&E) staining. ResultsIt was found that the NPs accumulated more in tumors. When using UCNPs-HSA and UCNPs-HSA+FA, similar changes were detected in rat organs. There were differences in the kidneys depending on the type of particle that was used. When UCNPs-HSA+FA+Cy3 particles were injected into the internal organs, signs of circulatory disorders and minor morphological signs of kidney damage were observed. DiscussionVisualizing the accumulation boundaries of NPs was achieved through image processing. This type of particle may be promising for future clinical trials and photodynamic therapy (PDT), as it demonstrates a correlation between particle accumulation and tumor necrosis. ConclusionThe data obtained will enable us to enhance the method of PDT using NPs and a photosensitizer (PS) with additional visualization capabilities.
Atherosclerosis is a risk factor for ischemic heart disease, and its progression has been associated with wall shear stress (WSS). While pulse waveform analysis has garnered increasing attention as a diagnostic indicator for atherosclerosis, the influence of the mechanical properties of arterial walls and the frequency of pulse waves on the pulse waveform and wall shear stress distribution has not been completely elucidated. In this study, three-dimensional fluid-structure interaction analysis was used to simulate the pulse wave propagation phenomena in the aorta. The effects of the mechanical properties of the arterial wall and the frequency of the pulse waves on the wall shear stress distribution and pulse wave dynamics were investigated. In the investigation of wall shear stress distribution, under conditions of low Young’s modulus or high frequency, the time-averaged wall shear stress (TAWSS) was low and the oscillatory shear index (OSI) became high. In contrast, under conditions of high Young’s modulus or low frequency, TAWSS was high and OSI became low. In the pulse waveform analysis, the influence of viscous friction incorporated in the three-dimensional fluid-structure interaction analysis confirmed that the pulse waveform attenuated and diffused during propagation. A causal relationship between the TAWSS values and the attenuation or diffusion of the pulse waveform was not observed. These results suggest that changes in arterial wall properties and differences in pulse wave frequency significantly influence WSS distribution. Furthermore, viscous friction in the three-dimensional FSI simulation led to attenuation and diffusion of the waveforms. However, the accuracy of waveform separation was insufficient, highlighting the need for improved methods that consider three-dimensional effects.
Immune Checkpoint Inhibitors (ICIs) have transformed the field of oncology by improving the functional ability of the immune system to combat malignancies. This review investigates the mechanisms of ICIs, their adverse effects, resistance mechanisms, and the role of Artificial Intelligence (AI) And Machine Learning (ML) in predicting treatment outcomes. A literature search was conducted using PubMed, Google Scholar, and Web of Science to identify pertinent studies, clinical trials, and review articles. The study concentrated on seven ICIs that have been approved and are designed to target the PD-1, PD-L1, and CTLA-4 pathways. The data were derived from clinical guidelines and expert opinions. ICIs have illustrated efficacy in a variety of malignancies, such as renal cell carcinoma, non-small cell lung cancer, and melanoma. Their utilization, whether as monotherapy or in conjunction with chemotherapy, radiotherapy, or targeted therapies, has substantially enhanced survival. Nevertheless, the management of Immune-Related Adverse Events (irAEs) that affect multiple organ systems is imperative. In certain patients, the efficacy of ICI is also restricted by resistance mechanisms. AI/ML-driven models demonstrate potential for anticipating patient responses, optimizing treatment strategies, and reducing toxicity risks. ICIs have revolutionized cancer therapy; however, there are still obstacles in predicting responses and managing adverse effects. This review emphasizes the innovative use of AI/ML to improve the precision and safety of ICI. Nevertheless, additional research is required due to the absence of reliable predictive biomarkers and the variability of patient responses. In order to enhance treatment outcomes and reduce toxicity, future research should enhance AI-driven models and incorporate multi-omics approaches.
Introduction Human daily activities and businesses generate a significant volume of data, which is expected to be transformed for the benefit of both businesses and humanity. Organisations utilise machine learning platforms to make informed decisions based on well-gleaned insights from their real-time data. The process of learning machine learning is challenging, making it difficult for employees to learn quickly and efficiently. Meanwhile, the introduction of automated machine learning (AutoML) has simplified this process. However, it is essential to understand how users adopt and implement the AutoML platform to address their real-world problems. Methods To achieve this, we conducted a quantitative study with 38 users focusing on four key areas: (1) the learning curve in ML and AutoML environments, (2) the design and usability strengths and weaknesses of AutoML platforms, (3) disparities in user experience between novices and professionals, and (4) design factors to enhance usability. Result Our findings revealed that users, particularly those with limited programming experience, have high expectations for the usability of AutoML; however, they also exhibit low awareness and adoption rates in the African context. Discussion The study illuminates gender disparities in technology adoption and identifies critical usability concerns, including the need for improved interpretability, feature engineering modules, and code integration for learning purposes. Additionally, we provide empirical evidence demonstrating AutoML’s advantages regarding training time and reproducibility compared to traditional machine learning tools. Conclusion This work offers novel insights into human-centered AutoML design, emphasizing inclusivity, explainability, and user-friendly interfaces. By addressing regional and gender-specific challenges, we propose actionable recommendations to democratize ML and enhance AutoML platforms. Future research should expand upon these findings by engaging frequent AutoML users to further refine usability and satisfaction metrics.
Adipose tissue and platelet-rich plasma (PRP) have gained significant attention in regenerative medicine and plastic surgery due to their potential in tissue repair and wound healing. This review aimed to analyze the biological properties of adipose tissue, lipofilling techniques, and the role of PRP in enhancing fat graft survival and tissue regeneration. A comprehensive review of the literature was carried out to evaluate the cellular composition, regenerative mechanisms, and clinical applications of adipose-derived therapies and PRP.Adipose tissue contains multipotent stem cells that contribute to angiogenesis, immunomodulation, and tissue remodeling. PRP enhances fat graft retention by promoting vascularization and reducing inflammatory responses. The combined use of PRP and adipose tissue has shown promising outcomes in wound healing, plastic surgery, and reconstructive procedures. The integration of adipose tissue derivatives and PRP holds significant potential for improving surgical outcomes. However, further research is needed to standardize protocols, optimize therapeutic strategies, and ensure reproducible clinical benefits.
Background Post-traumatic stress disorder (PTSD) is caused by depression and stress affecting the brain's emotional, memory, and sensory processes. Materials and Methods This study investigates a stacked deep learning model for trauma-based PTSD disorder diagnosis using rs-fMRI scans. Twenty-eight individual subjects, fourteen PTSD, and fourteen healthy controls were used, and each subject had 140 Resting-State Functional MRI (rs-fMRI) scans. The selected subjects were assessed to obtain brain activation from twelve brain regions of interest. Results The boxplot was used to check the performance of twelve ROI brain regions. Different deep learning algorithms were used for classification through a 10-fold cross-validation approach. This study examines the efficacy of employing a stacked deep approach with two models in the realm of predictive modeling. Discussion The objective of the proposed tacking model is to enhance the overall prediction accuracy and durability by using the complementary attributes of each model. The stacked model achieved a 98.30% accuracy rate on the training dataset and 96.60% on the test dataset. Conclusion Using the proposed approach, we could detect PTSD at an early stage. The selected ROI regions could also discriminate healthy PTSD from infected regions due to trauma events such as violence, accidents, and terrorism.
Background Paralyzed individuals, depending on their severity, are usually incapable of operating an electric wheelchair because it requires a common method of maneuvering, such as a joystick with buttons to control the chair. In such a case, an eye-controlled wheelchair can be utilized as it functions to facilitate mobility assistance for paralyzed or elderly individuals with limited movement within their natural environment. Objective This study aimed to explore the impact of using an eye-controlled wheelchair in a home-care environment on the quality of life of patients with neurological disorders. Method This case study was conducted by two neurological condition patients from a local home-care setting. To achieve the research objective, online questionnaires via Google Form were administered verbally after the eye-controlled wheelchair usage and subjects’ feedback was filled by the researchers. The efficiency of using the eye-controlled wheelchair was measured by the subject’s exhaustion level and workload. The total workload needed for wheelchair usage was measured using the National Aeronautics and Space Administration Task Load Index (NASA-TLX) tool, and a self-designed questionnaire was developed and validated (face and content validation) before the commencement of the study to measure subjective quality of life. Results The relationship between the quality of life and the total subjective workload was calculated using the Pearson correlation coefficient. Subject A displays a strong positive correlation (r =0.8476, n =8, P <.05), while, for Subject B (r =0.6196, n =8, P >.05), a moderate positive correlation was found between total subjective workload and quality of life. Overall, a positive correlation was observed between subjective workload and quality of life: as the workload decreased through the use of the eye-controlled wheelchair, the quality of life for both patients and caregivers significantly improved. Conclusion This study concluded that an eye-controlled wheelchair has a positive impact on the quality of life of patients with neurological disorder. This wheelchair will be beneficial for individuals with limited hand strength who are unable to operate a manual wheelchair or an electric one that uses a joystick or buttons.
Background Brain tumor identification at an early stage is a challenging task that increases the lifetime of patients. Specialists' conclusions on recognizing brain tumors are difficult, as they are based on their theoretical knowledge. It takes a huge amount of time to diagnose the patient. Recently, research has suggested an automated technique that is dependent on convolutional neural networks. Medical pictures are a set of accumulations of data that are hard to store and process, expending broad registering time. The decreased infiltrated systems are normally utilized as an information pre-preparing venture to make the picture information less mind-boggling with the goal that high-dimensional information may be recognized by a fitting and apt low-dimensional portrayal. Objective This study proposes an optimization-based dimensionality reduction and brain tumor segmentation using ensemble convolutional neural networks in MRI images to enhance disease diagnosis and extend healthcare accessibility. Methods Cuckoo-based dimensionality reduction and Ensemble CNN are proposed to segment the tumor region . The cuckoo-based optimization search technique is used to reduce the dimensionality of MRI Brain Images to perform better segmentation. The proposed technique is evaluated on the BRATS database, which contains two datasets: the Leaderboard and Challenge datasets. The outcomes are estimated utilizing the Dice Similarity Coefficient (DSC), Positive Predictive Value (PPV), and Sensitivity. Results The Experimental analysis shows promising results on the leaderboard dataset and the BRATS Challenge dataset. The proposed method outperformed the leaderboard dataset with a greater 91% Dice Similarity Coefficient (DCE), 95% Positive Predictive Value, and 87% Sensitivity of High-Grade Glioma (HGG). Seventy-two percent Dice Similarity Coefficient (DCE), 70% Positive Predictive Value, and 93% Sensitivity of Low-Grade Glioma (LGG). 88% Dice Similarity Coefficient (DCE), 90% Positive Predictive Value, and 91% Sensitivity of combined High-Grade glioma and Low-Grade glioma. For the BRATS Challenge dataset, the proposed method provides a 92% Dice Similarity Coefficient (DCE), 93% Positive Predictive Value, and 95% Sensitivity of High-Grade Glioma (HGG). 86% Dice Similarity Coefficient (DCE), 88% Positive Predictive Value and 93% Sensitivity of Low-Grade glioma (LGG). 85% Dice Similarity Coefficient (DCE), 89% Positive Predictive Value, and 92% Sensitivity of combined High-Grade glioma and Low-Grade glioma. Conclusion In this study, MRI Brain tumor segmentation using Cuckoo-based dimensionality reduction and Ensemble Convolutional Neural Network is proposed. The cuckoo search algorithm used for dimensionality reduction is performed in MRI images to reduce the dimensions. We also compared two of the existing methods with our proposed method. The leaderboard dataset and challenge dataset have been discussed. The challenge dataset for HGG provided good results in terms of dice similarity coefficient and positive predictive value. The sensitivity alone gets reduced when compared with the CNN and random forest methods. Experimental analysis shows promising results on the leaderboard dataset and the BRATS Challenge dataset.
Introduction Biological Named Entity Recognition (BioNER) is a crucial preprocessing step for Bio-AI analysis. Methods Our paper explores the field of Biomedical Named Entity Recognition (BioNER) by closely analysing two advanced models, SciSpaCy and BioBERT. We have made two distinct contributions: Initially, we thoroughly train these models using a wide range of biological datasets, allowing for a methodical assessment of their performance in many areas. We offer detailed evaluations using important parameters like F1 scores and processing speed to provide precise insights into the effectiveness of BioNER activities. Results Furthermore, our study provides significant recommendations for choosing tools that are customised to meet unique BioNER needs, thereby enhancing the efficiency of Named Entity Recognition in the field of biomedical research. Our work focuses on tackling the complex challenges involved in BioNER and enhancing our understanding of model performance. Conclusion The goal of this research is to drive progress in this important field and enable more effective use of advanced data analysis tools for extracting valuable insights from biomedical literature.
Introduction In this work, calibration-free blood pressure estimation using wavelet scalograms of PPG signals using Convolutional Neural Network (CNN) has been proposed. The PPG signal, easily obtained from a subject, serves as a reliable indicator for predicting blood pressure (BP). Methods The proposed methodology involves employing Continuous Wavelet Transform (CWT) scalograms of the PPG signal as inputs for the CNN. Two distinct architectures for BP estimation are explored: one employing regression with a fully connected neural network and another utilizing CNN with Support Vector Regression (SVR). Results The results demonstrate superior BP estimation with the CNN-SVR architecture. With the CNN-SVR model, the Systolic Blood Pressure (SBP) and Diastolic Blood Pressure (DBP) are estimated with a Root Mean Square Error (RMSE) of 6.7 mmHg and 8.9 mmHg, respectively. Conclusion The proposed CNN-SVR model gives 52% better estimation error performance in SBP estimation compared to a machine learning model reported in a previous work.
Introduction Traditional feed-forward neural networks (FFNN) have been widely used in image processing, but their effectiveness can be limited. To address this, we develop two deep learning models based on FFNN: the deep backpropagation neural network classifier (DBPNN) and the deep radial basis function neural network classifier (DRBFNN), integrating convolutional layers for feature extraction. Methods We apply a training algorithm to the deep, dense layers of both classifiers, optimizing their layer structures for improved classification accuracy across various hyperspectral datasets. Testing is conducted on datasets including Indian Pine, University of Pavia, Kennedy Space Centre, and Salinas, validating the effectiveness of our approach in feature extraction and noise reduction. Results Our experiments demonstrate the superior performance of the DBPNN and DRBFNN classifiers compared to previous methods. We report enhanced classification accuracy, reduced mean square error, shorter training times, and fewer epochs required for convergence across all tested hyperspectral datasets. Conclusion The results underscore the efficacy of deep learning feed-forward classifiers in hyperspectral image processing. By leveraging convolutional layers, the DBPNN and DRBFNN models exhibit promising capabilities in feature extraction and noise reduction, surpassing the performance of conventional classifiers. These findings highlight the potential of our approach to advance hyperspectral image classification tasks.
Introduction Obesity is a prevalent and multifaceted health hazard globally, necessitating effective predictive models to mitigate its impact on chronic diseases. Methods This paper introduces the Protein Food Item Prediction Regression (PIPR) model, employing machine learning techniques to analyze the influence of protein-rich foods on obesity. The model undergoes rigorous preprocessing and iterative refinement to identify correlated variables and predict obesity trends. Results The PIPR model demonstrates superior performance in predicting obesity trends, showcasing lower error rates and high adjusted R 2 values. For instance, for the most correlated variables like Meat and Milk (including butter), the model exhibits impressive performance with an MSE of 49.59, RMSE of 7.04, MAE of 5.08, and MAPE of 29%. Similarly, for the least correlated variables like oil crops and vegetable products, the PIPR model maintains excellence with an MSE of 52.51, RMSE of 7.24, MAE of 5.39, and MAPE of 31%. Conclusion The PIPR model emerges as a promising tool for understanding and addressing obesity's complexities, offering valuable insights into dietary patterns and potential interventions. Further research and validation could enhance its applicability and effectiveness in combating obesity on a global scale.
Background Spinal cord injuries (SCI) are debilitating conditions affecting individuals worldwide annually, leading to physical, emotional, and cognitive challenges. Effective rehabilitation for SCI patients is crucial for restoring motor function and enhancing their overall quality of life. Advances in technology, including machine learning (ML) and computer vision, offer promising avenues for personalized SCI treatment. Aims This paper aimed to propose an automated and cost-effective system for spinal cord injury (SCI) rehabilitation using machine learning techniques, leveraging data from the Toronto Rehab Pose dataset and Mediapipe for real-time tracking. Objective The objective is to develop a system that predicts rehabilitation outcomes for upper body movements, highlighting the transformative role of ML in personalized SCI treatment and offering tailored strategies for improved outcomes. Methods The proposed system utilized data from the Toronto Rehab Pose dataset and Mediapipe for real-time tracking. Machine learning models, including Support Vector Machines (SVM), Logistic Regression, Naive Bayes, and XGBoost, were employed for outcome prediction. Features such as joint positions, angles, velocities, and accelerations were extracted from movement data to train the models. Results Statistical analysis revealed the ability of the system to accurately classify rehabilitation outcomes, with an average accuracy of 98.5%. XGBoost emerged as the top-performing algorithm, demonstrating superior accuracy and precision scores across all exercises. Conclusion This paper emphasizes the importance of continuous monitoring and adjustment of rehabilitation plans based on real-time progress data, highlighting the dynamic nature of SCI rehabilitation and the need for adaptive treatment strategies. By predicting rehabilitation outcomes with high accuracy, the system enables clinicians to devise targeted interventions, optimizing the efficacy of the rehabilitation process.
Introduction A malignant abnormal growth that starts in the tissues of the lungs is called Lung Cancer. It ranks among the most common and lethal cancers globally. Lung Cancer is particularly dangerous because of its aggressive nature and how quickly it can extend to other areas of the body. We propose a two-step verification architecture to check the presence of Lung Cancer. The model proposed by this paper first assesses the patient based on a few questions about the patient's symptoms and medical background. Then, the algorithm determines whether the patient has a low, medium, or high risk of developing lung cancer by diagnosing the response using the “Decision Tree” classification at an accuracy of 99.67%. If the patient has a medium or high risk, we further validate the finding by examining the patient's CT scan image using the “VGG16” CNN model at an accuracy of 92.53%. Background One of the key areas of research on Lung Cancer prediction is to identify patients based on symptoms and medical history. Its subjective nature makes it challenging to apply in real-world scenarios. Another research area in this field involves forecasting the presence of cancer cells using CT scan imagery, providing high accuracy. However, it requires physician intervention and is not appropriate for early-stage prediction. Objective This research aims to forecast the severity of Lung Cancer by analyzing the patient with a few questions regarding the symptoms and past medical conditions. If the patient has a medium or a high risk, we further examine their CT scan, validate the result and also predict the type of Lung Cancer. Methodology This paper uses the “Decision Tree” algorithm and the Customised “VGG16” model of CNN for the implementation. The “Decision Tree” algorithm is used to analyze the answers given by the patient to distinguish the severity of Lung Cancer. We further use Convolution Neural Networks with a Customised “VGG16” model to examine the patient's CT scan image, validate the result and categorize the type of Lung Cancer. Results The “Decision Tree” approach for forecasting the severity of lung cancer yields an accuracy of 99.67%. The accuracy of the customized “VGG16” CNN model to indicate the type of Lung Cancer suffered by the patient is 92.53% Conclusion This research indicates that our technique provides greater accuracy than the prior approaches for this problem and has extensive use in the prognosis of Lung Cancer.
Background For the purpose of diagnosing diseases and developing treatment plans, blood cell pictures must be accurately classified. This procedure can be greatly enhanced by automated systems that make use of deep learning and the Internet of Medical Things (IoMT). Objective In order to improve illness detection and increase healthcare accessibility, this work suggests an IoMT-based system for remote blood cell picture transmission and classification utilizing deep learning algorithms. Methods High-resolution pictures of blood cells are captured by an IoMT tiny camera and wirelessly sent to a cloud-based infrastructure. The blood cells are divided into groups according to a, deeplearning classification algorithm, including neutrophils, lymphocytes, monocytes, and eosinophils. Results The IoMT-enabled system excels in transmitting and analyzing blood cell images, achieving precise classification. Utilizing deep learning models with multi-scale feature extraction and attention mechanisms, the system demonstrates robust performance. Numerical results showcase a high accuracy of approximately 97.21%, along with noteworthy precision, recall, and F1 scores for individual blood cell classes. Eosinophil, Lymphocyte, Monocyte, and Neutrophil classes exhibit strong performance metrics, emphasizing the system's effectiveness in accurate blood cell classification. Conclusion By combining IoMT and deep learning with blood cell image analysis, diagnostic accessibility and efficiency are improved. The suggested approach has the potential to completely transform healthcare by facilitating prompt interventions, individualized treatment regimens, and better patient outcomes. It is essential to continuously enhance and validate the system in order to maximize its efficacy and dependability in a variety of healthcare settings.
