Classification Accuracy Values Research Articles

Identification of Leaf Diseases from Figs Using Deep Learning Methods

Early detection of plant diseases is of great importance for agricultural production and plant health. Early detection is important to prevent the spread of diseases and reduce agricultural losses. The aim of this study is to use artificial intelligence technologies for the early detection of diseased fig plants and reduce agricultural losses. The fig leaf dataset used in the study has two classes: healthy and diseased leaves. There are a total of 2321 images in the dataset. Among these images, there are 1350 images representing diseased leaves and 971 images representing healthy leaves. The dataset is divided into 80% training data and 20% test data. DarkNet-19, ResNet50, VGG-19, VGG-16, ShuffleNet, GoogLeNet, MobileNet-v2, EfficientNet-b0, and DarkNet-53 algorithms were used to analyze the fig leaves dataset using a MATLAB graphical user interface (GUI). The classification accuracy values of each algorithm are as follows: DarkNet-19 90.3%, ResNet50 90.95%, VGG-19 93.32%, VGG-16 92.89%, ShuffleNet 89.44%, GoogLeNet 87.5%, MobileNet-v2 87.5%, EfficientNet-b0 85.56%, and DarkNet53 91.59%. These results evaluate the usability and performance of different algorithms for the early detection of plant diseases. The research emphasizes the importance of the effective use of artificial intelligence technologies in the agricultural industry.

Monitoring the Urban Heat Island Phenomenon in Pontianak City, West Kalimantan, using Google Earth Engine

The urban heat island (UHI) phenomenon can cause the land surface temperature (LST) in the city center to be higher than in the surrounding area. The population of Pontianak City increases yearly, resulting in land cover changes and LST distribution. This research aims to monitor the UHI phenomenon by identifying the temporal distribution of LST in Pontianak City. This research analyses the Landsat 8 thermal band, which is converted into LST value based on the normalised difference vegetation index (NDVI). The land cover classification was conducted using a random forest algorithm and acquisition of confusion matrix classification accuracy value. This research utilizes the Google Earth Engine (GEE) platform because of its ability to process satellite data continuously and has data with a sizeable spatiotemporal dimension. The results show a strong relationship between urban development and LST distribution. Based on the land cover classification in 2014 and 2019, there is a significant change in the composition of the surface, namely, built-up land has increased, while open space and vegetation land have decreased. Time series analysis also shows that the average LST of built-up land, open space, and vegetation are 28.10℃, 22.80℃, and 21.56℃, respectively. The results of monitoring the UHI phenomenon can help in better urban spatial planning, especially in increasing green space, because it is proven effective in reducing land surface temperature.

An adaptation of hybrid binary optimization algorithms for medical image feature selection in neural network for classification of breast cancer

The performance of neural network is largely dependent on their capability to extract very discriminant features supporting the characterization of abnormalities in the medical image. Several benchmark architectures have been proposed and the use of transfer learning has further made these architectures return good performances. Study has shown that the use of optimization algorithms for selection of relevant features has improved classifiers. However continuous optimization algorithms have mostly been used though it allows variables to take value within a range of values. The advantage of binary optimization algorithms is that it allows variables to be assigned only two states, and this have been sparsely applied to medical image feature optimization. This study therefore proposes hybrid binary optimization algorithms to efficiently identify optimal features subset in medical image feature sets. The binary dwarf mongoose optimizer (BDMO) and the particle swarm optimizer (PSO) were hybridized with the binary Ebola optimization search algorithm (BEOSA) on new nested transfer functions. Medical images passed through convolutional neural networks (CNN) returns extracted features into a continuous space which are piped through these new hybrid binary optimizers. Features in continuous space a mapped into binary space for optimization, and then mapped back into the continuous space for classification. Experimentation was conducted on medical image samples using the Curated Breast Imaging Subset of Digital Database for Screening Mammography (DDSM+CBIS). Results obtained from the evaluation of the hybrid binary optimization methods showed that they yielded outstanding classification accuracy, fitness, and cost function values of 0.965, 0.021 and 0.943. To investigate the statistical significance of the hybrid binary methods, the analysis of variance (ANOVA) test was conducted based on the two-factor analysis on the classification accuracy, fitness, and cost metrics. Furthermore, results returned from application of the binary hybrid methods medical image analysis showed classification accuracy of 0.8286, precision of 0.97, recall of 0.83, and F1-score of 0.99, AUC of 0.8291. Findings from the study showed that contrary to the popular approach of using continuous metaheuristic algorithms for feature selection problem, the binary metaheuristic algorithms are well suitable for handling the challenge. Complete source code can be accessed from: https://github.com/NathanielOy/hybridBinaryAlgorithm4FeatureSelection

A two-stage deep-learning model for determination of the contact of mandibular third molars with the mandibular canal on panoramic radiographs

ObjectivesThis study aimed to assess the accuracy of a two-stage deep learning (DL) model for (1) detecting mandibular third molars (MTMs) and the mandibular canal (MC), and (2) classifying the anatomical relationship between these structures (contact/no contact) on panoramic radiographs.MethodMTMs and MCs were labeled on panoramic radiographs by a calibrated examiner using bounding boxes. Each bounding box contained MTM and MC on one side. The relationship of MTMs with the MC was assessed on CBCT scans by two independent examiners without the knowledge of the condition of MTM and MC on the corresponding panoramic image, and dichotomized as contact/no contact. Data were split into training, validation, and testing sets with a ratio of 80:10:10. Faster R-CNN was used for detecting MTMs and MCs and ResNeXt for classifying their relationship. AP50 and AP75 were used as outcomes for detecting MTMs and MCs, and accuracy, precision, recall, F1-score, and the area-under-the-receiver-operating-characteristics curve (AUROC) were used to assess classification performance. The training and validation of the models were conducted using the Python programming language with the PyTorch framework.ResultsThree hundred eighty-seven panoramic radiographs were evaluated. MTMs were present bilaterally on 232 and unilaterally on 155 radiographs. In total, 619 images were collected which included MTMs and MCs. AP50 and AP75 indicating accuracy for detecting MTMs and MCs were 0.99 and 0.90 respectively. Classification accuracy, recall, specificity, F1-score, precision, and AUROC values were 0.85, 0.85, 0.93, 0.84, 0.86, and 0.91, respectively.ConclusionDL can detect MTMs and MCs and accurately assess their anatomical relationship on panoramic radiographs.

Analysing crash severity on expressways in India using statistical and ML approaches

The high injury severity of traffic crashes on Indian expressways is a significant concern for road safety experts, although studies dedicated to this critical issue are limited. The present study investigates the factors influencing crash severity using multinomial logit (MNL), decision tree (DT) and random forest (RF) models on a dataset of 2,747 crashes on three selected expressways. The dependent variable, crash severity, had four injury severity categories: fatal, severe, minor, and property damage only (PDO). Various explanatory variables, included traffic and speed characteristics, temporal and geometric characteristics, primary contributing factors, crash type, and the vehicles involved. Synthetic minority oversampling (SMOTE) and randomise class balancing (RCB) techniques were also employed to tackle the class imbalance issue in the dataset. The predictive performance of models was evaluated using classification accuracy and Kappa value. The RF model showed the highest predictive accuracy on the RCB dataset. The key findings highlight the critical need for enforcement of speed limits and entry restrictions, lane discipline, and improvement of design deficiencies such as provisions of truck laybys and bus bays. These measures can help in policy-making and engineering improvements to enhance road safety on expressways in India.

Rapid qualitative and quantitative detection for adulteration of Atractylodis Rhizoma using hyperspectral imaging combined with chemometric methods

In the field of traditional Chinese medicine, Atractylodis Rhizoma (AR) is commonly used for various diseases due to its excellent ability to dry dampness and strengthen the spleen, especially popular in East Asia. The aim of this study is to proposed Hyperspectral Imaging (HSI) in combination with chemometric methods for the rapid qualitative and quantitative detection of AR adulteration with other types of powder. Partial Least Squares Discriminant Analysis (PLS-DA) was used to construct the classification models the best, with the First-order Derivative (F-D) preprocessing method. The accuracy values of the test sets for classification models were above 99%. Furthermore, Partial Least Squares Regression (PLSR), Random Forest Regression (RFR), and BP Neural Network (BPNN) were used to quantitatively analyze the adulteration level. On the whole, the BPNN model has a relatively stable effect. The R-square (R2) values of different models were all greater than 0.97, the Root Mean Square Error (RMSE) values were all less than 0.0300, and the Relative Percentage Difference (RPD) values were over 6.00. After applying three characteristic wavelength selection algorithms, namely Iterative Retained Information Variable (IRIV), Successive Projections Algorithm (SPA), and Variable Iterative Space Shrinkage Approach (VISSA) algorithms, the classification accuracy values remained over 99.00% while the quantification models’ RPD values were over 4.00. These results demonstrate the reliability of using hyperspectral imaging combined with chemometrics methods for the adulteration problems in AR.

Spline Estimator in Nonparametric Ordinal Logistic Regression Model for Predicting Heart Attack Risk

In Indonesia, one of the main causes of death for both young and elderly people is heart attacks, and the main cause of heart attacks is non-communicable diseases such as hypertension. Deaths due to heart attacks caused by non-communicable diseases, namely hypertension, rank first in Indonesia. Therefore, predictions of the risk of having a heart attack caused by hypertension need serious attention. Further, for determining whether a patient is experiencing a heart attack, an effective method of prediction is required. One efficient approach is to use statistical models. This study discusses predicting risk of heart attack via modeling and classifying hypertension risk based on factors that influence it, namely, age, cholesterol levels, and triglyceride levels by using the spline estimator of the Nonparametric Ordinal Logistic Regression (NOLR) model. In this study, we assume an ordinal scale response variable with q categories to have an asymmetric distribution, namely, a multinomial distribution. The data used in this study are secondary data from medical records of cardiac poly patients at the Haji General Hospital in Surabaya, Indonesia. The results show that the proposed model approach has the greatest classification accuracy and sensitivity values compared to NOLR model approach using GAM, and the classical model approach, namely the Parametric Ordinal Logistic Regression (POLR) model. This means that the NOLR model approach is suitable for predicting hypertension and heart attack risks. Also, the NOLR model estimated using the LS-Spline estimator obtained is valid for predicting the risk of heart attack with accuracy value of 85% and sensitivity value of 100%.

Research on Amplification Algorithm of Small Sample Soil Composite Data Based on Probability Distribution

Abstract Analyzing and studying the data of soil compounds from 10 cities, two important features were found in this project. First, the data for all compounds in soil follow a normal distribution with different parameters. Therefore, this paper argues that small-sample soil compound amplification data should also have these two characteristics. Based on this, this paper proposes a new algorithm called PDWCC (probability distribution with correlation coefficient) for amplification of small-sample soil compound data. The algorithm combines a probability distribution with a Pearson correlation coefficient. In addition, an optimized genetic algorithm incorporating a mutation operator was introduced to improve the efficiency of data amplification. Finally, experimental comparisons were made with conventional Bootstrap and SMOTE amplification algorithms using SVM classifiers. The results show that the PDWCC amplification algorithm outperforms the other algorithms in terms of classification accuracy, recall and F1 value with mean values of 81.77%, 79.34% and 78.65%, respectively.

Gradient Boosting Prediction of Overlapping Genes From Weighted Co-expression and Differential Gene Expression Analysis of Wnt Pathway: An Artificial Intelligence-Based Bioinformatics Study.

Introduction The Wnt (wingless-related integration site) signalling pathway is crucial for bone formation and remodelling, regulating the commitment of mesenchymal stem cells (MSCs) to the osteoblastic lineage. It triggers the transcriptional activation of Wnt target genes and promotes osteoblast proliferation and survival. Weighted co-expression network analysis (WGCNA) and differential gene expression analysis help researchers understand gene roles. Gradient boosting, a machine learning technique, enhances understanding of genetic and molecular mechanisms contributing to overlap genes, improving gene regulation and functional genomics. The aim is to predict overlapping genes in the Wnt signalling pathway. Methods Differential gene expression analysis was performed using the National Center for Biotechnology Information (NCBI) geo dataset-GSE251951, focusing on the effect of Wnt signaling on treatment. The WGCNA module was analyzed using the iDEP tool to identify interconnected gene clusters. Hub genes were identified by calculating module eigengenes, correlated with external traits, and ranked based on module membership values. The study utilized gradient boosting, an ensemble learning method, to predict models, evaluate their performance using metrics like accuracy, precision, recall, and F1 score, and adjust predictions based on gradient and learning rate. Results The dendrogram uses the "Dynamic TreeCut" algorithm to analyze gene clusters, aiding researchers in understanding gene modules and biological processes, identifying co-expressed genes, and discovering new pathways. The confusion matrix displays 88 actual and predicted cases. The gradient boosting model achieves 78.9% accuracy in predicting Wnt pathway overlapping genes, with a respectable area under the curve (AUC) and classification accuracy values. It accurately predicts 73.9% of samples, with a high precision ratio and low recall. Conclusion Future research should enhance differential expression analysis and WGCNA to identify key Wnt pathway genes, improve sensitivity, specificity, hyperparameter tuning, and validation experiments, and use larger datasets.

Construction and Validation of a General Medical Image Dataset for Pretraining.

In the field of deep learning for medical image analysis, training models from scratch are often used and sometimes, transfer learning from pretrained parameters on ImageNet models is also adopted. However, there is no universally accepted medical image dataset specifically designed for pretraining models currently. The purpose of this study is to construct such a general dataset and validate its effectiveness on downstream medical imaging tasks, including classification and segmentation. In this work, we first build a medical image dataset by collecting several public medical image datasets (CPMID). And then, some pretrained models used for transfer learning are obtained based on CPMID. Various-complexity Resnet and the Vision Transformer network are used as the backbone architectures. In the tasks of classification and segmentation on three other datasets, we compared the experimental results of training from scratch, from the pretrained parameters on ImageNet, and from the pretrained parameters on CPMID. Accuracy, the area under the receiver operating characteristic curve, and class activation map are used as metrics for classification performance. Intersection over Union as the metric is for segmentation evaluation. Utilizing the pretrained parameters on the constructed dataset CPMID, we achieved the best classification accuracy, weighted accuracy, and ROC-AUC values on three validation datasets. Notably, the average classification accuracy outperformed ImageNet-based results by 4.30%, 8.86%, and 3.85% respectively. Furthermore, we achieved the optimal balanced outcome of performance and efficiency in both classification and segmentation tasks. The pretrained parameters on the proposed dataset CPMID are very effective for common tasks in medical image analysis such as classification and segmentation.

Breaking Barriers in Cancer Diagnosis: Super‐Light Compact Convolution Transformer for Colon and Lung Cancer Detection

ABSTRACTAccording to the World Health Organization, lung and colon cancers are known for their high mortality rates which necessitate the diagnosis of these cancers at an early stage. However, the limited availability of data such as histopathology images used for diagnosis of these cancers, poses a significant challenge while developing computer‐aided detection system. This makes it necessary to keep a check on the number of parameters in the artificial intelligence (AI) model used for the detection of these cancers considering the limited availability of the data. In this work, a customised compact and efficient convolution transformer architecture, termed, C3‐Transformer has been proposed for the diagnosis of colon and lung cancers using histopathological images. The proposed C3‐Transformer relies on convolutional tokenisation and sequence pooling approach to keep a check on the number of parameters and to combine the advantage of convolution neural network with the advantages of transformer model. The novelty of the proposed method lies in efficient classification of colon and lung cancers using the proposed C3‐Transformer architecture. The performance of the proposed method has been evaluated on the ‘LC25000’ dataset. Experimental results shows that the proposed method has been able to achieve average classification accuracy, precision and recall value of 99.30%, 0.9941 and 0.9950, in classifying the five different classes of colon and lung cancer with only 0.0316 million parameters. Thus, the present computer‐aided detection system developed using proposed C3‐Transformer can efficiently detect the colon and lung cancers using histopathology images with high detection accuracy.

Klasifikasi Perilaku Pemain Game Online Menggunakan Naïve Bayes Berbasis Particle Swarm Optimization

Much research has been conducted to understand player behavior as a result of the rapid growth of online gaming. In this research, we use the Naive Bayes method optimized using Particle Swarm Optimization (PSO) to analyze the behavior classification of online game players. The classification accuracy value of the baseline method is 75.09% and the Area Under the Curve (AUC) value is 0.798. We use PSO-based optimization on Naïve Bayes to improve model performance. The results showed that the combination of Naïve Bayes and PSO increased classification accuracy to 95.28% with an AUC value of 0.990. This is a major advance that shows that combining the PSO algorithm with Naive Bayes can enable better classification of online game player behavior. These findings will make a significant contribution to the process of making plans that can improve the gaming experience.

Enhancing UWB Indoor Positioning Accuracy through Improved Snake Search Algorithm for NLOS/LOS Signal Classification.

Non-line-of-sight (NLOS) errors significantly impact the accuracy of ultra-wideband (UWB) indoor positioning, posing a major barrier to its advancement. This study addresses the challenge of effectively distinguishing line-of-sight (LOS) from NLOS signals to enhance UWB positioning accuracy. Unlike existing research that focuses on optimizing deep learning network structures, our approach emphasizes the optimization of model parameters. We introduce a chaotic map for the initialization of the population and integrate a subtraction-average-based optimizer with a dynamic exploration probability to enhance the Snake Search Algorithm (SSA). This improved SSA optimizes the initial weights and thresholds of backpropagation (BP) neural networks for signal classification. Comparative evaluations with BP, Particle Swarm Optimizer-BP (PSO-BP), and Snake Optimizer-PB (SO-BP) models-performed using three performance metrics-demonstrate that our LTSSO-BP model achieves superior stability and accuracy, with classification accuracy, recall, and F1 score values of 90%, 91.41%, and 90.25%, respectively.

Implementasi Decision Tree untuk Prediksi Kelahiran Bayi Prematur

The early birth of baby in Indonesia is a case that has a very high incidence rate. According to data from the Ministry of Health in 2021, the presentation of premature babies in Indonesia is 84%. The number of infant deaths in Indonesia is still relatively high compared to other ASEAN countries. The purpose of this study was to predict the birth of premature babies with the implementation of decision tree, with this type of predictive analysis research. The population in this study is pregnant women patients with a sample of 350 pregnant women patient data covering the variables studied Age, BMI, Vaginal Discharge, History of Miscarriage, History of Prematurity and Pregnancy Spacing. The prediction was made by halving the training data by 245 and the testing data by 105. The results obtained are the variable Body Mass Index (BMI) is the riskiest factor for premature birth The decision tree model yields an AUC of 91.7%, it can be concluded that the decision tree has a good classification accuracy value.

Multi-strategy enhanced Marine Predators Algorithm with applications in engineering optimization and feature selection problems

This paper introduces a novel meta-heuristic optimization algorithm named Clustering Wavelet Opposition-based Marine Predators Algorithm (CWOMPA) to address some limitations present in the well-established Marine Predators Algorithm (MPA). CWOMPA incorporates three key strategies: a fuzzy clustering approach for escaping local optima, using wavelet basis function-based impact coefficient adjustment for elites to prevent premature convergence, and finally an adaptive opposition-based learning strategy for maintaining population diversity. Compared with some recent meta-heuristic algorithms, extensive evaluations conducted affirm that CWOMPA achieves the best Friedman rank, 4.30 and 1.95 respectively, on 23 benchmark functions and the CEC 2017 benchmark set. Not only does CWOMPA demonstrate significant effectiveness on six constrained problems from the CEC 2020 real-world benchmarks, but also when applied to 14 medical datasets, it gains superior Friedman ranks in terms of selected features, classification accuracy, F-Score, and objective function value, 2.5, 2.25, 2.96, 2.93 respectively, outperforming other common methods. Finally, CWOMPA exhibits the highest classification accuracy across all datasets and the best F-Score performance in nine datasets compared to traditional feature selection algorithms. The obtained results reveal that CWOMPA is a powerful and versatile optimization algorithm with significant potential in various real-world applications, including feature selection for medical datasets.

LAND USE LAND COVER CHANGE MAPPING FROM SENTINEL 1B < 2A IMAGERY USING RANDOM FOREST ALGORITHM IN CÔTE D’IVOIRE

Monitoring crop condition, soil properties, and mapping tillage activities can be used to assess land use, forecast crops, monitor seasonal changes, and contribute to the implementation of sustainable development policy. Agricultural maps can provide independent and objective estimates of the extent of crops in a given area or growing season, which can be used to support efforts to ensure food security in vulnerable areas. Satellite data can help detect and classify different types of soil. The evolution of satellite remote sensing technologies has transformed techniques for monitoring the Earth’s surface over the last several decades. The European Space Agency (ESA) and the European Union (EU) created the Copernicus program, which resulted in the European satellites Sentinel-1B (S1B) and Sentinel-2A (S2A), which allow the collection of multi-temporal, spatial, and highly repeatable data, providing an excellent opportunity for the study of land use, land cover, and change. The goal of this study is to map the land cover of Côte d’Ivoire’s West Central Soubre area (5°47′1′′ North, 6°35′38′′ West) between 2014 and 2020. The method is based on a combination of S1B and S2A imagery data, as well as three types of predictors: the biophysical indices Normalized Difference Vegetation Index “(NDVI)”, Modified Normalized Difference Water Index “(MNDWI)”, Normalized Difference Urbanization Index “(NDBI)”, and Normalized Difference Water Index “(NDWI)”, as well as spectral bands (B1, B11, B2, B3, B4, B6, B7, B8) and polarization coefficients VV. For the period 2014–2020, six land classifications have been established: Thick_Forest, Clear_Drill, Urban, Water, Palm_Oil, Bareland, and Cacao_Land. The Random Forest (RF) algorithm with 60 numberOfTrees was the primary categorization approach used in the Google Earth Engine (GEE) platform. The results show that the RF classification performed well, with outOfBagErrorEstimates of 0.0314 and 0.0498 for 2014 and 2020, respectively. The classification accuracy values for the kappa coefficients were above 95%: 96.42% in 2014 and 95.28% in 2020, with an overall accuracy of 96.97% in 2014 and 96 % in 2020. Furthermore, the User Accuracy (UA) and Producer Accuracy (PA) values for the classes were frequently above 80%, with the exception of the Bareland class in 2020, which achieved 79.20%. The backscatter coefficients of the S1B polarization variables had higher GINI significance in 2014: VH (70.80) compared to VH (50.37) in 2020; and VV (57.11) in 2014 compared to VV (46.17) in 2020. Polarization coefficients had higher values than the other spectral and biophysical variables of the three predictor variables. During the study period, the Thick_Forest (35.90% ± 1.17), Palm_Oil (57.59% ± 1.48), and Water (5.90% ± 0.47) classes experienced a regression in area, while the Clear_Drill (16.96% ± 0.80), Urban (2.32% ± 0.29), Bareland (83.54% ± 1.79), and Cacao_Land (35.14% ± 1.16) classes experienced an increase. The approach used is regarded as excellent based on the results obtained.

Analysis of the Effectiveness of Using Two-Stage Neural Network Models for Early Detection of Forest Fires

The purpose of the research – analysis of the effectiveness of two-stage neural network models for solving the problem of detecting forest fires in images obtained from unmanned aerial vehicles.Methods. А training dataset was synthesized for training neural network models for the purpose of detection and semantic segmentation of forest fires in images. Тwo-stage neural network models (“Faster R-CNN”, “Mask RCNN” and “Retina-Net”) were used for training. Тhe neural network models were trained according to the same parameters set for all models in order to ensure consistency and a common basis for experiments. Optimization of model parameters during the training process was carried out to minimize the classification loss function. Тo synthesize the test sample, we used a video sequence covering the events of forest fires in the /rkutsk region, which was filmed by an unmanned aerial vehicle. Using a specially developed script in the Рython programming language, the process of dividing this video sequence into separate frames was carried out, which were used as a test data set when assessing the quality of classification of trained neural network models.Results. Based on the analysis of the obtained values of the quality criterion, as well as visual analysis on the test data set produced as part of testing neural network models, the effectiveness of the studied models for detecting forest fires in images was assessed. Тo assess the quality of binary classification of neural network models, the quality criterion “Accuracy” (classification accuracy) was used.Conclusion. Еxperimental studies on a test data set showed that the Retina-Net model demonstrates the lowest, but acceptable, performance compared to other studied neural network models. Тhe two-stage neural network models “Faster R-CNN” and “Mask R-CNN” demonstrate similar classification accuracy values (0.9492 and 0.9521, respectively), which allows us to recommend them for use in early detection systems for forest fires.

YOLOX-SwinT algorithm improves the accuracy of AO/OTA classification of intertrochanteric fractures by orthopedic trauma surgeons

PurposeIntertrochanteric fracture (ITF) classification is crucial for surgical decision-making. However, orthopedic trauma surgeons have shown lower accuracy in ITF classification than expected. The objective of this study was to utilize an artificial intelligence (AI) method to improve the accuracy of ITF classification. MethodsWe trained a network called YOLOX-SwinT, which is based on the You Only Look Once X (YOLOX) object detection network with Swin Transformer (SwinT) as the backbone architecture, using 762 radiographic ITF examinations as the training set. Subsequently, we recruited 5 senior orthopedic trauma surgeons (SOTS) and 5 junior orthopedic trauma surgeons (JOTS) to classify the 85 original images in the test set, as well as the images with the prediction results of the network model in sequence. Statistical analysis was performed using the Statistical Package for the Social Sciences (SPSS) 20.0 (IBM Corp., Armonk, NY, USA) to compare the differences among the SOTS, JOTS, SOTS + AI, JOTS + AI, SOTS + JOTS, and SOTS + JOTS + AI groups. All images were classified according to the AO/OTA 2018 classification system by 2 experienced trauma surgeons and verified by another expert in this field. Based on the actual clinical needs, after discussion, we integrated 8 subgroups into 5 new subgroups, and the dataset was divided into training, validation, and test sets by the ratio of 8:1:1. ResultsThe mean average precision at the intersection over union (IoU) of 0.5 (mAP50) for subgroup detection reached 90.29%. The classification accuracy values of SOTS, JOTS, SOTS + AI, and JOTS + AI groups were 56.24% ± 4.02%, 35.29% ± 18.07%, 79.53% ± 7.14%, and 71.53% ± 5.22%, respectively. The paired t-test results showed that the difference between the SOTS and SOTS + AI groups was statistically significant, as well as the difference between the JOTS and JOTS + AI groups, and the SOTS + JOTS and SOTS + JOTS + AI groups. Moreover, the difference between the SOTS + JOTS and SOTS + JOTS + AI groups in each subgroup was statistically significant, with all p < 0.05. The independent samples t-test results showed that the difference between the SOTS and JOTS groups was statistically significant, while the difference between the SOTS + AI and JOTS + AI groups was not statistically significant. With the assistance of AI, the subgroup classification accuracy of both SOTS and JOTS was significantly improved, and JOTS achieved the same level as SOTS. ConclusionIn conclusion, the YOLOX-SwinT network algorithm enhances the accuracy of AO/OTA subgroups classification of ITF by orthopedic trauma surgeons.

Discriminant Decision Making of Cardiovascular Diseases Based on Cloud-Based Convolutional Attention Network

Cardiovascular diseases (CVDs) have become the number one killer affecting human health. In order to reduce the burden of medical workers, facilitate government screening of the population and enable patients to conduct their own health status checks, there is an urgent need for a complementary diagnostic system to predict the occurrence of CVD. In this study, a new cloud-based convolutional attention network (C-CAN) model is proposed for the discriminant decision making of CVD. In this model, the indicator data for discriminant decision making of CVD are trained using an improved one-dimensional convolutional neural network (1D CNN) model structure based on the correlation of factors influencing CVD given by decision-making trial and evaluation laboratory (DEMATEL) and cloud models. This 1D CNN model consists of a convolutional pooling module, an attention module and a fully connected module. The cloud model is used to process the original data based on the discriminating opinion of experts, so as to select the important factors that affect CVD. The attention mechanism is effective in augmenting attention to the essential elements of the data and reducing attention to the less important features. Both have similarities in that they are effective in augmenting the important features in the data and combine with each other to achieve better results. Moreover, the C-CAN is compared with decision tree (DT), [Formula: see text]-nearest neighbors (KNN), random forests (RF) and normal CNN according to the CVD dataset from the Kaggle platform. The results show that the classification accuracy, precision, recall and F1 value of C-CAN are all higher than that of all compared models. Further, the proposed model is further externally validated using other imbalanced datasets, and the results indicate that C-CAN has good resilience for imbalanced data. Our findings suggest that C-CAN represents a promising new approach that may somehow address the challenges associated with deep learning (DL) in the medical field.

Classifying Raman Spectra of Colon Cells Based on Machine Learning Algorithms

Colorectal cancer is very widespread in developed countries. Its diagnosis partly depends on pathologists’ experience and their laboratories’ instrumentation, producing uncertainty in diagnosis. The use of spectroscopic techniques sensitive to the cellular biochemical environment could aid in achieving a reliable diagnosis. So, we used Raman micro-spectroscopy, combined with a spectral analysis by means of machine learning methods, to build classification models, which allow colon cancer to be diagnosed in cell samples, in order to support such methods as complementary tools for achieving a reliable identification of colon cancer. The Raman spectra were analyzed in the 980–1800 cm−1 range by focusing the laser beam onto the nuclei and the cytoplasm regions of single FHC and CaCo-2 cells (modelling healthy and cancerous samples, respectively) grown onto glass coverslips. The comparison of the Raman intensity of several spectral peaks and the Principal Component Analysis highlighted small biochemical differences between healthy and cancerous cells mainly due to the larger relative lipid content in the former cells with respect to the latter ones and to the larger relative amount of nucleic acid components in cancerous cells compared with healthy ones. We considered four classification algorithms (logistic regression, support vector machine, k nearest neighbors, and a neural network) to associate unknown Raman spectra with the cell type to which they belong. The built machine learning methods achieved median values of classification accuracy ranging from 95.5% to 97.1%, sensitivity values ranging from 95.5% to 100%, and specificity values ranging from 93.9% to 97.1%. The same median values of the classification parameters, which were estimated for a testing set including unknown spectra, ranged between 93.1% and 100% for accuracy and between 92.9% and 100% for sensitivity and specificity. A comparison of the four methods pointed out that k nearest neighbors and neural networks better perform the classification of nucleus and cytoplasm spectra, respectively. These findings are a further step towards the perspective of clinical translation of the Raman technique assisted by multivariate analysis as a support method to the standard cytological and immunohistochemical methods for diagnostic purposes.