Computational Classification Research Articles

Classification of tomato leaf disease using Transductive Long Short-Term Memory with an attention mechanism.

Tomatoes are considered one of the most valuable vegetables around the world due to their usage and minimal harvesting period. However, effective harvesting still remains a major issue because tomatoes are easily susceptible to weather conditions and other types of attacks. Thus, numerous research studies have been introduced based on deep learning models for the efficient classification of tomato leaf disease. However, the usage of a single architecture does not provide the best results due to the limited computational ability and classification complexity. Thus, this research used Transductive Long Short-Term Memory (T-LSTM) with an attention mechanism. The attention mechanism introduced in T-LSTM has the ability to focus on various parts of the image sequence. Transductive learning exploits the specific characteristics of the training instances to make accurate predictions. This can involve leveraging the relationships and patterns observed within the dataset. The T-LSTM is based on the transductive learning approach and the scaled dot product attention evaluates the weights of each step based on the hidden state and image patches which helps in effective classification. The data was gathered from the PlantVillage dataset and the pre-processing was conducted based on image resizing, color enhancement, and data augmentation. These outputs were then processed in the segmentation stage where the U-Net architecture was applied. After segmentation, VGG-16 architecture was used for feature extraction and the classification was done through the proposed T-LSTM with an attention mechanism. The experimental outcome shows that the proposed classifier achieved an accuracy of 99.98% which is comparably better than existing convolutional neural network models with transfer learning and IBSA-NET.

An Enhanced Two-Way Unet Approach for Copy Move Image Forgery Detection

The paper explores a deep learning-based approach to semantic classification, emphasizing its utility in complex real-world situations. The main aim is to engineer a model that can recognize features in images and distinguish them accurately and efficiently. Leveraging advanced architectures, including convolutional neural networks (CNNs) and their variants, the research combines complex training methods with advanced datasets to achieve the state-of-the-art Includes conceptual techniques and data enhancement methods to the model's capability to generalize to diverse has greatly impressive images. The report describes typical improvements in accuracy and loss coefficients at various stages, and highlights the importance of fine-tuning hyperparameters Analytical metrics such as accuracy, accuracy, loss, and validation loss reveal high model performance displayed, balanced in terms of computational efficiency and classification quality Alongside envisioning predictive coverage, the report offers qualitative and quantitative evidence for on how effective the model is This approach holds particular promise for applications such as autonomous driving, surveillance, and medical imaging. The findings also highlight the importance of continuous innovation in model construction and training techniques to push the limits of logical classification.

An efficient deep learning system for automatic detection of Acute Lymphoblastic Leukemia.

Early and highly accurate detection of rapidly damaging deadly disease like Acute Lymphoblastic Leukemia (ALL) is essential for providing appropriate treatment to save valuable lives. Recent development in deep learning, particularly transfer learning, is gaining a preferred trend of research in medical image processing because of their admirable performance, even with small datasets. It inspires us to develop a novel deep learning-based leukemia detection system in which an efficient and lightweight MobileNetV2 is used in conjunction with ShuffleNet to boost discrimination ability and enhance the receptive field via convolution layer succession. More importantly, the suggested weight factor and an optimal threshold value (which is experimentally selected) is responsible for maintaining a healthy balance between computational efficiency and classification performance. Hence, the benefits of inverted residual bottleneck structure, depthwise separable convolution, tunable hyperparameters, pointwise group convolution, and channel shuffling are integrated to improve the feature discrimination ability and make the proposed system faster and more accurate. The experimental results convey that the proposed framework outperforms others with the best detection performances. It achieves superior performance to its competitors with the best accuracy (99.07%), precision(98.00%), sensitivity (100%), specificity (98.31%), and F1 score (0.9899) in ALLIDB1 dataset. Similarly, it outperforms others with 98.46% accuracy, 98.46% precision, 98.46% specificity, 98.46% sensitivity, and 0.9846 F1 Score in ALLIDB2 dataset.

A Comprehensive CNN Model for Age-Related Macular Degeneration Classification Using OCT: Integrating Inception Modules, SE Blocks, and ConvMixer.

Background/Objectives: Age-related macular degeneration (AMD) is a significant cause of vision loss in older adults, often progressing without early noticeable symptoms. Deep learning (DL) models, particularly convolutional neural networks (CNNs), demonstrate potential in accurately diagnosing and classifying AMD using medical imaging technologies like optical coherence to-mography (OCT) scans. This study introduces a novel CNN-based DL method for AMD diagnosis, aiming to enhance computational efficiency and classification accuracy. Methods: The proposed method (PM) combines modified Inception modules, Depthwise Squeeze-and-Excitation Blocks, and ConvMixer architecture. Its effectiveness was evaluated on two datasets: a private dataset with 2316 images and the public Noor dataset. Key performance metrics, including accuracy, precision, recall, and F1 score, were calculated to assess the method's diagnostic performance. Results: On the private dataset, the PM achieved outstanding performance: 97.98% accuracy, 97.95% precision, 97.77% recall, and 97.86% F1 score. When tested on the public Noor dataset, the method reached 100% across all evaluation metrics, outperforming existing DL approaches. Conclusions: These results highlight the promising role of AI-based systems in AMD diagnosis, of-fering advanced feature extraction capabilities that can potentially enable early detection and in-tervention, ultimately improving patient care and outcomes. While the proposed model demon-strates promising performance on the datasets tested, the study is limited by the size and diversity of the datasets. Future work will focus on external clinical validation to address these limita-tions.

Feature selection based on consistent granulation

Classification learning from large-scale data is a pivotal and complex task in machine learning. Enhancing classifier performance can be achieved through effective feature selection and construction. However, current feature selection methods often encounter difficulties with large-scale data. Considering granulation is essential for developing accurate and efficient models, we firstly introduce a novel granulation technique, termed consistent granulation in this study. Furthermore, an innovative feature construction method leveraging consistent granulation is proposed to enhance computational speed, scalability, and classification performance of feature selection. Consequently, both time and space complexities are reduced to linear levels relative to sample sizes. Consistent granulation also allows for an accurate representation of the data's topological structure and reduces the parameter set from the interval [0,1] to a finite set, simplifying the search for optimal parameters. Experimental results indicate that our proposed algorithm, FCG, surpasses four classical feature selection algorithms—FARNeMF, HANDI, GBNRS, and RMDPS—in handling datasets with millions of samples and tens of thousands of dimensions. Specifically, on the first fourteen datasets, FCG decreases the average runtime by times of 1029.30 (FARNeMF), 703.54 (HANDI), 1253.78 (GBNRS), and 3.08 (RMDPS), while enhancing average classification accuracy by 3.24% (FARNeMF), 2.74% (HANDI), 5.77% (GBNRS), and 15.5% (RMDPS).

M2ara: unraveling metabolomic drug responses in whole-cell MALDI mass spectrometry bioassays.

Fast computational evaluation and classification of concentration responses for hundreds of metabolites represented by their mass-to-charge (m/z) ratios is indispensable for unraveling complex metabolomic drug actions in label-free, whole-cell Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI MS) bioassays. In particular, the identification of novel pharmacodynamic biomarkers to determine target engagement, potency, and potential polypharmacology of drug-like compounds in high-throughput applications requires robust data interpretation pipelines. Given the large number of mass features in cell-based MALDI MS bioassays, reliable identification of true biological response patterns and their differentiation from any measurement artefacts that may be present is critical. To facilitate the exploration of metabolomic responses in complex MALDI MS datasets, we present a novel software tool, M2ara. Implemented as a user-friendly R-based shiny application, it enables rapid evaluation of Molecular High Content Screening (MHCS) assay data. Furthermore, we introduce the concept of Curve Response Score (CRS) and CRS fingerprints to enable rapid visual inspection and ranking of mass features. In addition, these CRS fingerprints allow direct comparison of cellular effects among different compounds. Beyond cellular assays, our computational framework can also be applied to MALDI MS-based (cell-free) biochemical assays in general. The software tool, code, and examples are available at https://github.com/CeMOS-Mannheim/M2ara and https://dx.doi.org/10.6084/m9.figshare.25736541.

MVCLST: A spatial transcriptome data analysis pipeline for cell type classification based on multi-view comparative learning

Recent advancements in spatial transcriptomics sequencing technologies can not only provide gene expression within individual cells or cell clusters (spots) in a tissue but also pinpoint the exact location of this expression and generate detailed images of stained tissue sections, which offers invaluable insights into cell type identification and cell function exploration. However, effectively integratingthegene expression data, spatial location information, and tissue images from spatial transcriptomics data presents a significant challenge for computational methodsin cell classification. In this work, we propose MVCLST, a multi-view comparative learningmethod to analyze spatial transcriptomicsdata for accurate cell type classification. MVCLSTconstructs two views based on gene expression profiles, cell coordinates and image features. The multi-view method we proposed can significantly enhance the effectiveness of feature extraction while avoiding the impact of erroneous information in organizing image or gene expression data. The model employs four separate encoders to capture shared and unique features within each view. To ensure consistency and facilitate information exchange between the two views, MVCLST incorporates a contrastive learning loss function. The extracted shared and private features from both views are fused using corresponding decoders. Finally, the model utilizes the Leiden algorithm to clusterthe learned featuresfor cell type identification. Additionally, we establish a framework called MVCLST-CCFS for spatial transcriptomicsdata analysis based on MVCLST and consistent clustering. Our method achieves excellent results in clustering on human dorsolateral prefrontal cortex data and the mouse brain tissue data. Italso outperforms state-of-the-art techniques in the subsequent search for highly variable genes across cell types on the mouse olfactory bulbdata.

Automatic motion artifact detection in electrodermal activity signals using 1D U-net architecture

We developed a method for automated detection of motion and noise artifacts (MNA) in electrodermal activity (EDA) signals, based on a one-dimensional U-Net architecture. EDA has been widely employed in diverse applications to assess sympathetic functions. However, EDA signals can be easily corrupted by MNA, which frequently occur in wearable systems, particularly those used for ambulatory recording. MNA can lead to false decisions, resulting in inaccurate assessment and diagnosis. Several approaches have been proposed for MNA detection; however, questions remain regarding the generalizability and the feasibility of implementation of the algorithms in real-time especially those involving deep learning approaches. In this work, we propose a deep learning approach based on a one-dimensional U-Net architecture using spectrograms of EDA for MNA detection. We developed our method using four distinct datasets, including two independent testing datasets, with a total of 9602 128-s EDA segments from 104 subjects. Our proposed scheme, including data augmentation, spectrogram computation, and 1D U-Net, yielded balanced accuracies of 80.0 ± 13.7 % and 75.0 ± 14.0 % for the two independent test datasets; these results are better than or comparable to those of other five state-of-the-art methods. Additionally, the computation time of our feature computation and machine learning classification was significantly lower than that of other methods (p < .001). The model requires only 0.28 MB of memory, which is far smaller than the two deep learning approaches (4.93 and 54.59 MB) which were used as comparisons to our study. Our model can be implemented in real-time in embedded systems, even with limited memory and an inefficient microprocessor, without compromising the accuracy of MNA detection.

Advanced Computational Methods for News Classification: A Study in Neural Networks and CNN integrated with GPT

In an era inundated with vast amounts of information, the imperative for efficient news classification is paramount. This research explores the sophisticated integration of neural networks and convolutional neural networks (CNN) with Generative Pre-trained Transformers (GPT) to enhance the precision and efficacy of news categorization. The rapid digital dissemination of news necessitates advanced computational methodologies capable of accurate classification and event prediction that include finance and economic events. Leveraging recent advancements in machine learning and natural language processing (NLP), this study utilizes large language models (LLMs) such as GPT and BERT, known for their exceptional comprehension and generation of human-like text. Over 232 days, our methodology classified 33,979 news articles into Education & Learning, Health & Medicine, and Science & Technology, with further subcategorization into 32 distinct subcategories. For evaluation, a sample of 5,000 articles was assessed using metrics such as True Positive (TP), True Negative (TN), False Positive (FP), False Negative (FN), Precision, Recall, and F1-Score. In comparison with the existing studies, the proposed method achieving significantly higher with average scores of 0.986 (Precision), 0.987 (Recall), and 0.987 (F1-Score). This research offers substantial practical contributions, providing detailed insights into news source contributions, effective anomaly detection, and predictive trend analysis using neural networks. The theoretical contributions are profound, demonstrating the mathematical integration of GPT with CNNs and recurrent neural networks. This integration advances computational news classification and exemplifies how sophisticated mathematical frameworks enhance large-scale text data analysis, marking a pivotal advancement in applying advanced computational methods in real-world scenarios.

Migrating birds optimization-based feature selection for text classification.

Text classification tasks, particularly those involving a large number of features, pose significant challenges in effective feature selection. This research introduces a novel methodology, MBO-NB, which integrates Migrating Birds Optimization (MBO) approach with naïve Bayes as an internal classifier to address these challenges. The motivation behind this study stems from the recognized limitations of existing techniques in efficiently handling extensive feature sets. Traditional approaches often fail to adequately streamline the feature selection process, resulting in suboptimal classification accuracy and increased computational overhead. In response to this need, our primary objective is to propose a scalable and effective solution that enhances both computational efficiency and classification accuracy in text classification systems. To achieve this objective, we preprocess raw data using the Information Gain algorithm, strategically reducing the feature count from an average of 62,221 to 2,089. Through extensive experiments, we demonstrate the superior effectiveness of MBO-NB in feature reduction compared to other existing techniques, resulting in significantly improved classification accuracy. Furthermore, the successful integration of naïve Bayes within MBO offers a comprehensive and well-rounded solution to the feature selection problem. In individual comparisons with Particle Swarm Optimization (PSO), MBO-NB consistently outperforms by an average of 6.9% across four setups. This research provides valuable insights into enhancing feature selection methods, thereby contributing to the advancement of text classification techniques. By offering a scalable and effective solution, MBO-NB addresses the pressing need for improved feature selection methods in text classification, thereby facilitating the development of more robust and efficient classification systems.

Computational techniques, classification, datasets review and way forward with modern analysis of epileptic seizure – a study

Computational techniques, classification, datasets review and way forward with modern analysis of epileptic seizure – a study

Spiking mode-based neural networks.

Spiking neural networks play an important role in brainlike neuromorphic computations and in studying working mechanisms of neural circuits. One drawback of training a large-scale spiking neural network is that updating all weights is quite expensive. Furthermore, after training, all information related to the computational task is hidden into the weight matrix, prohibiting us from a transparent understanding of circuit mechanisms. Therefore, in this work, we address these challenges by proposing a spiking mode-based training protocol, where the recurrent weight matrix is explained as a Hopfield-like multiplication of three matrices: input modes, output modes, and a score matrix. The first advantage is that the weight is interpreted by input and output modes and their associated scores characterizing the importance of each decomposition term. The number of modes is thus adjustable, allowing more degrees of freedom for modeling the experimental data. This significantly reduces the training cost because of significantly reduced space complexity for learning. Training spiking networks is thus carried out in the mode-score space. The second advantage is that one can project the high-dimensional neural activity (filtered spike train) in the state space onto the mode space which is typically of a low dimension, e.g., a few modes are sufficient to capture the shape of the underlying neural manifolds. We successfully apply our framework in two computational tasks-digit classification and selective sensory integration tasks. Our method thus accelerates the training of spiking neural networks by a Hopfield-like decomposition, and moreover this training leads to low-dimensional attractor structures of high-dimensional neural dynamics.

A fuzzy-based emotion detection method to classify the attractiveness of urban green spaces

In European studies, the most used definition of Urban Green Spaces (UGS) is based on the European Urban Atlas, which includes public green areas primarily used for recreation and green areas adjacent to urban areas that are managed or utilized for recreational purposes. UGS play a vital role in creating sustainable and resilient cities, as they provide essential social benefits for the well-being and health of urban residents. Both planners and scientists acknowledge the importance of involving, actively or passively, citizens in defining criteria for designing and managing inclusive and functional UGS. According to a post-normal science approach, the integration of hard data from scientific sources with soft data gathered from citizens’ engagement holds the potential to shape an innovative support system for public policies addressing significant, urgent, and uncertain challenges pertaining to UGS. Nowadays, the abundance of data generated through online reviews, opinions, and comments allows for collecting valuable information about people’s opinions and sentiments towards UGS. This study proposes a methodological framework that utilizes emotion detection techniques to identify and analyze citizens’ emotions concerning UGS through social reviews. To balance computational costs and classification accuracy, the framework introduces a fuzzy emotion-based classification method called FREDoC (Fuzzy Relevance Emotions Document Classification). This method incorporates a lightweight natural language pro-cessing (NLP) approach to detect and annotate terms associated with specific emotional categories within the text. The framework adopts the psycho-evolutionary classification approach based on R. Plutchik’s observations of general emotional responses. This model is implemented within a Geographical Information System (GIS) for the purpose of categorizing UGS, specifically green parks, according both to WHO report key indicators and to the detected relevant emotions. The outcome is a novel classification model of UGS that can assist decision-makers in identifying the attractiveness of UGS as catalysts for urban transformation processes.

Skin Lesion Prediction and Classification Using Innovative Modified Long Short-Term Memory-Based Hybrid Optimization Algorithm

Identification of pigmented skin lesions is necessary for the detection of severe diseases associated with the skin organ, notably malignancy. Accurate skin cancer diagnosis can be improved with the use of image detection approaches and computer classification skills. Therefore, this research work plans to perform skin lesion prediction and classification using a novel deep learning methodology. Initially, the data related to the skin lesion are gathered from the ISIC dataset. After collecting the images, the pre-processing is performed using hair removal and filtering hair removed images via median filtering. These pre-processed images undergo segmentation process accomplished using the U-Net method. Next, the features are extracted from these segmented images with the help of color features, and texture features by GLCM and RGB histogram features. These extracted features undergo the prediction phase that is accomplished using the MLSTM model, in which the parameter optimization is done by the nature inspired novel hybrid metaheuristic algorithm referred as SC-STBO algorithm with the consideration of accuracy maximization and RMSE minimization as the major fitness for the objective function. If the predicted output is returned as the presence of skin lesion, the same novel MLSTM model classifies the final skin lesion output into seven types, such as Vascular Lesions, Melanocytic Nevi, Melanoma, Dermatofibroma, Benign Keratosis-like Lesions, BCC, and Actinic Keratoses, respectively. Seven groups of skin diseases can be identified early thanks to the suggested effort, which can then be tested and properly handled by medical professionals. With an accuracy of 0.9931, the recommended methodology clearly outperforms traditional techniques. Similarly, the suggested methodology clearly beats the conventional methods, with a recall of 0.9825.

Characterizing the consistency of motion of spermatozoa through nanoscale motion tracing

To demonstrate nanoscale motion tracing of spermatozoa and present analysis of the motion traces to characterize the consistency of motion of spermatozoa as a complement to progressive motility analysis. Anonymized sperm samples were videographed under a quantitative phase microscope, followed by generating and analyzing superresolution motion traces of individual spermatozoa. Not applicable. Centrifuged human sperm samples. Not applicable. Precision of motion trace of individual sperms, presence of a helical pattern in the motion trace, mean and standard deviations of helical periods and radii of sperm motion traces, speed of progression. Spatially sensitive quantitative phase imaging with a superresolution computational technique MUltiple SIgnal Classification ALgorithm allowed achieving motion precision of 340 nm using ×10, 0.25 numerical aperture lens whereas the diffraction-limited resolution at this setting was 1,320 nm. The motion traces thus derived facilitated new kinematic features of sperm, namely the statistics of helix period and radii per sperm. Through the analysis, 47 sperms with a speed >25 μm/s were randomly selected from the same healthy donor semen sample, it is seen that the kinematic features did not correlate with the speed of the sperms. In addition, it is noted that spermatozoa may experience changes in the periodicity and radius of the helical path over time. Further, some very fast sperms (e.g., >70 μm/s) may demonstrate irregular motion and need further investigation. Presented computational analysis can be used directly for sperm samples from both fertility patients with normal and abnormal sperm cell conditions. We note that MUltiple SIgnal Classification ALgorithm is an image analysis technique that may vaguely fall under the machine learning category, but the conventional metrics for reporting found in Enhancing the QUAlity and Transparency Of health Research network do not apply. Alternative suitable metrics are reported, and bias is avoided through random selection of regions for analysis. Detailed methods are included for reproducibility. Kinematic features derived from nanoscale motion traces of spermatozoa contain information complementary to the speed of the sperms, allowing further distinction among the progressively motile sperms. Some highly progressive spermatozoa may have irregular motion patterns, and whether irregularity of motion indicates poor quality regarding artificial insemination needs further investigation. The presented technique can be generalized for sperm analysis for a variety of fertility conditions.

Computational intelligence-based classification system for the diagnosis of memory impairment in psychoactive substance users

Computational intelligence techniques have emerged as a promising approach for diagnosing various medical conditions, including memory impairment. Increased abuse of psychoactive drugs poses a global public health burden, as repeated exposure to these substances can cause neurodegeneration, premature aging, and negatively affect memory impairment. Many studies in the literature relied on statistical studies, but they remained inaccurate. Some studies relied on physical data because the time factor was not considered, until Artificial Intelligence (AI) techniques came along that proved their worth in this diagnosis. The variable deep neural network method was used to adapt to the intermediate results and re-process the intermediate in case the result is undesirable. Computational intelligence was used in this study to classify a brain image from MRI or CT scans and to show the effectiveness of the dose ratio on health with treatment time, and to diagnose memory impairment in users of psychoactive substances. Understanding the neurotoxic profiles of psychoactive substances and the underlying pathways is hypothesized to be of great importance in improving the risk assessment and treatment of substance use disorders. The results proved the worth of the proposed method in terms of the accuracy of recognition rate as well as the possibility of diagnosis. It can be concluded that the diagnostic efficiency is increased by increasing the number of hidden layers in the neural network and controlling the weights and variables that control the deep learning algorithm. Thus, we conclude that good classification in this field may save human life or early detection of memory impairment.

Hybrid rater to quantify and measure the severity of infection and spread of infection in muskmelon

Disease severity index (DIS) is a way of calculating the percentage of infection spread across the field. The percentage of infection in each leaf has been considered at a time stamp is being calculated and based on that disease, severity of disease spread is analyzed. With the advancement in machine learning and deep learning algorithms in the field of computer vision, identification and classification of diseases is effortless. Percentage of infection in a particular leaf, disease index (DI) is calculated using image processing techniques like Otsu threshold method. With this DI and scales, grading the severity of the infection across the field can be achieved. In this paper various scales used for grading severity of infection namely Horsfall-Barratt (H-B scale) quantitative ordinal scale, Amended 20% ordinal scale, and nearest percent estimates (NPEs) in muskmelon is explored, and based on the empirical results Amended 20% ordinal scale is most efficient method of estimating the DIS is to use the midpoint of the severity scope for each class with twenty percent adjusted to ordinal scale. The results show that the density of leaves is directly proportional to spread of diseases in muskmelon plant.

Parameterization of manipulative media discourse: possibilities and problems of automatic diagnosis

Abstract The issue of quantitative measurement and automatic processing is a significant problem in determining the markers of the manipulative potential of media texts, since linguistic indicators are the basis of machine parameterization. The purpose of the research is to analyse the possibilities of the main language parameters of the manipulativeness of media discourse, which can be identified using machine learning. To achieve the research goals, the following methods were used: system, content analysis, computer modelling, and comparative. The results of the article determined that such language indicators as use of the subjunctive mood of verbs, capital letters, high frequency of use of the ‘not’ particle, punctuation marks, questions, or exclamations of a rhetorical nature, use of quotation marks for the purpose of irony, double negative sentences, use of the word ‘no’, and verbal structures calling to action act as computer classification parameters. In order to cover the above purpose, PYTHON software was implemented that allowed texts to be analysed and visualized in algorithmic and lexical-vocabulary ways. In addition, it was determined that by integrating the PYTHON tool, it became possible to use language transformation markers that formed linguistic patterns in the analysed text. The list of parameters for diagnosing manipulative texts is non-exhaustive, which emphasizes the possibility of machine measurement of the manipulative component of mass media discourse.

Research on Computer Classification Algorithm of Concrete Crack Based on Deep Learning

In combination with transfer learning under the framework of Unet semantic partitioning, VGG16 pre-trained neural network enhanced encoder is used to extract multi-level high-level semantic information. The cross-entropy loss function is used to eliminate the imbalance between samples, and finally the crack shape is accurately semitone. Combining with the theory of computer vision, a quantitative calculation method of the crack region, length, width and other geometric characteristic parameters based on the binary segmentation template is proposed. Finally, the self-developed dam concrete crack image is taken as an example to carry out simulation and comparison test to check the correctness and superiority of the research results of this project. The research results will reveal that the crack recognition based on deep neural network can achieve a high recognition rate. The calculation results of fracture characteristic parameters meet the requirement of detection accuracy. The research results of this paper are expected to provide a new technical means for the quality control of dam concrete structure.

Computational Analysis and Classification of Hernia Repairs

Problems related to ventral hernia repairs (VHR) are very common, and evaluating them using computational methods can assist in selecting the most appropriate treatment. This study is based upon data from 3339 patients from different European countries observed during the last 12 years (2012–2023), which were collected by specialists in hernia surgery. Most patients underwent standard surgical procedures, with a growing trend towards laparoscopic surgery. This paper focuses on statistically evaluating the treatment methods in relation to patient age, body mass index (BMI), and the type of repair. Appropriate mathematical methods are employed to extract and classify the selected features, with emphasis on computational and machine-learning techniques. The paper presents surgical hernia treatment statistics related to patient age, BMI, and repair methods. The main conclusions point to mean groin hernia repair (GHR) complications of 19% for patients in the database. The accuracy of separating GHR mesh surgery with and without postoperative complications reached 74.4% using a two-layer neural network classification. Robotic surgeries represent 22.9% of all the evaluated hernia repairs. The proposed methodology suggests both an interdisciplinary approach and the utilization of computational intelligence in hernia surgery, potentially applicable in a clinical setting.