Low Accuracy Rate Research Articles

MS-YOLOv5: a lightweight algorithm for strawberry ripeness detection based on deep learning

The existing ripeness detection algorithm for strawberries suffers from low detection accuracy and high detection error rate. Considering these problems, we propose an improvement method based on YOLOv5, named MS-YOLOv5. The first step is to reconfigure the feature extraction network of MS-YOLOv5 by replacing the standard convolution with the depth hybrid deformable convolution (Ms-MDconv). In the second step, a double cooperative attention mechanism (Bc-attention) is constructed and implemented in the CSP2 module to improve the feature representation in complex environments. Finally, the Neck section of MS-YOLOv5 has been enhanced to use the fast-weighted fusion of cross-scale feature pyramid networks (FW-FPN) to replace the CSP2 module. It not only integrates multi-scale target features but also significantly reduces the number of parameters. The method was tested on the strawberry ripeness dataset, the mAP reached 0.956, the FPS reached 76, and the model size was 7.44M. The mAP and FPS are 8.4 and 1.3 percentage higher than the baseline network, respectively. The model size is reduced by 6.28M. This method is superior to mainstream algorithms in detection speed and accuracy. The system can accurately identify the ripeness of strawberries in complex environments, which could provide technical support for automated picking robots.

Using Twitter to collect a multi-dialectal corpus of Albanian using advanced geotagging and dialect modeling.

In this study, we present the acquisition and categorization of a geographically-informed, multi-dialectal Albanian National Corpus, derived from Twitter data. The primary dialects from three distinct regions-Albania, Kosovo, and North Macedonia-are considered. The assembled publicly available dataset encompasses anonymized user information, user-generated tweets, auxiliary tweet-related data, and annotations corresponding to dialect categories. Utilizing a highly automated scraping approach, we initially identified over 1,000 Twitter users with discernible locations who actively employ at least one of the targeted Albanian dialects. Subsequent data extraction phases yielded an augmentation of the preliminary dataset with an additional 1,500 Twitterers. The study also explores the application of advanced geotagging techniques to expedite corpus generation. Alongside experimentation with diverse classification methodologies, comprehensive feature engineering and feature selection investigations were conducted. A subjective assessment is conducted using human annotators, which demonstrates that humans achieve significantly lower accuracy rates in comparison to machine learning (ML) models. Our findings indicate that machine learning algorithms are proficient in accurately differentiating various Albanian dialects, even when analyzing individual tweets. A meticulous evaluation of the most salient attributes of top-performing algorithms provides insights into the decision-making mechanisms utilized by these models. Remarkably, our investigation revealed numerous dialectal patterns that, despite being familiar to human annotators, have not been widely acknowledged within the broader scientific community.

Perbandingan Kinerja K-Nearest Neighbors dan Naive Bayes Untuk Klasifikasi Perilaku Nasabah Pada Pembayaran Kredit Bank

:Credit customers are people who use banking services or other financial services. to use bank money in its business activities, so it expects that the bank's credit can meet business capital needs. To reach information to increase profits and reduce company losses, we need a method that can provide knowledge to support the company's data. Research data can be obtained from processing classification data from credit customer data that are categorized as potential or not potential in the next credit grant. Data processing can be done using machine learning, namely classification techniques. This technique will produce a predictive churn model to determine which customer categories belong to a group. potential smooth or jammed. The Naive Bayes method was chosen because it can produce maximum accuracy with little training data. Meanwhile, the K-Nearest Neighbor method was chosen because it is robust against noise data. The performance of the two methods will be compared, so that it can be seen which method is better in classifying documents. The results obtained show that the Naive Bayes method has better performance with an accuracy rate of 70%, while the K-Nearest Neighbor method has a fairly low accuracy rate of 40%. Thus, it can be seen the accuracy value displayed by applying the classification algorithm. K-Nears Neighbors and Naïve Bayes. Parameter category. which in this study are account numbers, names of debtors, collectibility in the categories: current, DPK (on special mention), substandard, doubtful, loss. Then clarified with a description of the type of loan, collectability of ADK (computer data archive), type of business.

Benign and Malignant Oral Lesion Image Classification Using Fine-Tuned Transfer Learning Techniques.

Oral lesions are a prevalent manifestation of oral disease, and the timely identification of oral lesions is imperative for effective intervention. Fortunately, deep learning algorithms have shown great potential for automated lesion detection. The primary aim of this study was to employ deep learning-based image classification algorithms to identify oral lesions. We used three deep learning models, namely VGG19, DeIT, and MobileNet, to assess the efficacy of various categorization methods. To evaluate the accuracy and reliability of the models, we employed a dataset consisting of oral pictures encompassing two distinct categories: benign and malignant lesions. The experimental findings indicate that VGG19 and MobileNet attained an almost perfect accuracy rate of 100%, while DeIT achieved a slightly lower accuracy rate of 98.73%. The results of this study indicate that deep learning algorithms for picture classification demonstrate a high level of effectiveness in detecting oral lesions by achieving 100% for VGG19 and MobileNet and 98.73% for DeIT. Specifically, the VGG19 and MobileNet models exhibit notable suitability for this particular task.

Evaluating the relationship between pain empathy, knowledge and attitudes among nurses in North China: a cross-sectional study

BackgroundEffective pain management is closely related to nurses’ knowledge, attitudes and empathy regarding pain. Nursing educators and managers should understand the relationship between nurses’ pain management knowledge, attitudes and empathy level, and take targeted measures accordingly. Currently, there is limited study exploring the relationship between pain empathy and pain knowledge and attitudes among nurses in North China.ObjectivesThe purpose of this study was to investigate the level of nurses’ pain management knowledge and attitudes and pain empathy, to analyze the factors influencing pain empathy, and to explore the relationship between these two variables.DesignThis study was a quantitative, descriptive-correlation design.Setting and participantsThe study population was registered nurses in North China, the sample included 177 registered nurses in North China.MethodsData were collected with the “General data questionnaire”, “Knowledge and attitudes survey regarding pain” (KASRP) and the “Empathy for pain scale” (EPS) via Wechat mini program “Questionnaire Star”.ResultsThe 177 registered nurses completed the survey. The averege correct rate for KASRP was (51.94 ± 9.44)%, and none of the respondents achieved a percentage score of >80%. The mean score for pain empathy was (2.78 ± 0.78), the empathy reactions dimension was (2.99 ± 0.77), and the body and mind discomfort dimension was (2.71 ± 0.80). The results of multiple stepwise linear regression showed that whether they had received empathy training, whether they had greater trauma or severe pain and whether they had negative emotions were independent influencing factors for EPS scores. Pearson correlation analysis showed that KASRP scores were positively correlated with EPS scores (r = 0.242, P < 0.05).ConclusionsThe pain knowledge and attitudes of nurses in North China are far from optimal. Nurses have a relatively low accuracy rate in areas such as medication knowledge, assessment of patient pain based on case studies, and handling PRN prescriptions. Nursing educators and administrators need to design some pain management courses in a targeted manner. Nurses’ empathy for pain was at a moderate level. Pain empathy was positively correlated with pain knowledge and attitudes, suggesting that empathy for pain can be developed postnatally.

Development of an Algorithm for Detecting Real-Time Defects in Steel

The integration of artificial intelligence with steel manufacturing operations holds great potential for enhancing factory efficiency. Object detection algorithms, as a category within the field of artificial intelligence, have been widely adopted for steel defect detection purposes. However, mainstream object detection algorithms often exhibit a low detection accuracy and high false-negative rates when it comes to detecting small and subtle defects in steel materials. In order to enhance the production efficiency of steel factories, one approach could be the development of a novel object detection algorithm to improve the accuracy and speed of defect detection in these facilities. This paper proposes an improved algorithm based on the YOLOv5s-7.0 version, called YOLOv5s-7.0-FCC. YOLOv5s-7.0-FCC integrates the basic operator C3-Faster (C3F) into the C3 module. Its special T-shaped structure reduces the redundant calculation of channel features, increases the attention weight on the central content, and improves the algorithm’s computational speed and feature extraction capability. Furthermore, the spatial pyramid pooling-fast (SPPF) structure is replaced by the Content Augmentation Module (CAM), which enriches the image feature content with different convolution rates to simulate the way humans observe things, resulting in enhanced feature information transfer during the process. Lastly, the upsampling operator Content-Aware ReAssembly of Features (CARAFE) replaces the “nearest” method, transforming the receptive field size based on the difference in feature information. The three modules that act on feature information are distributed reasonably in YOLOv5s-7.0, reducing the loss of feature information during the convolution process. The results show that compared to the original YOLOv5 model, YOLOv5s-7.0-FCC increases the mean average precision (mAP) from 73.1% to 79.5%, achieving a 6.4% improvement. The detection speed also increased from 101.1 f/s to 109.4 f/s, an improvement of 8.3 f/s, further meeting the accuracy requirements for steel defect detection.

Past and future time reference processing teased apart in Paiwan, an endangered Formosan language

AbstractHow the time reference of a sentence is processed based on the grammatical marking of the verb has already been explored in several languages with grammatical tense and aspect. It can also be grammatically expressed according to the reality status of the event (whether the event exists in time,realismood, or not,irrealis). This study reports results from an acceptability judgment experiment in Paiwan, a Formosan language which exhibits a realis-perfective/irrealis distinction. By placing realis-perfective and irrealis markers after deictic past or future time adverbs and manipulating the grammaticality of the sentences, we asked which temporal concord violation (i.e., realis-perfective or irrealis) was harder to detect. The temporal concord violation of the realis-perfective marker induced greater processing difficulties (interactions between time reference and mood marking revealed lower accuracy rates and longer reaction time), but not the irrealis marker, in line with previous hypotheses. These processing difficulties may be partly due to the Paiwan realis mood marker which also encodes perfective aspect meaning. The reanalysis of the design of previous studies indicates that the interaction with perfective aspect also led to additional processing cost, suggesting that perfective aspect marking plays a crucial role in the processing of time reference.

Development and application of KASP marker for high throughput detection of the seedless trait in grapevine

Molecular marker-assisted selection (MAS) can significantly accelerate and improve the efficiency of the breeding process in seedless grape cultivars. In this study, we developed the KASP_VviAGL11 and VviAGL11_410 markers based on a single nucleotide polymorphism (SNP) site (Chr18: 26889437 (A/C)) of the VviAGL11 gene, and compared them with previously reported SSR markers p3_VvAGL11 and 5U_VviAGL11 by testing 101 cultivars and 81 F1 hybrid progenies. The results showed that both of the proposed markers obtained 100% accuracy rates in detecting allele A, which was closely associated with the seedless trait in grapes, while p3_VvAGL11 and 5U_VviAGL11 had lower accuracy rates due to their tendency to produce false positives. After careful evaluation of the technical advantages and disadvantages associated with these markers, we concluded that KASP_VviAGL11 was superior in terms of simplicity, cost-effectiveness, efficiency, and accuracy. Thus, we optimized the process of molecular MAS for seedless grapes, focusing on the KASP_VviAGL11 marker as a central component, to provide key technical support for the development of new seedless grape cultivars.

LabVIEW based ECG simulator and automatic detection system for cardiac disorders

In the present age, most cardiac arrhythmias are emerging due to irregular rhythmic conditions of the heart. Diagnosis of cardiac arrhythmias plays an essential role in the survival of the human race. If rhythms of nature are distorted for any reason, it leads to several variations or abnormalities known as heart arrhythmias. Irregular heartbeats lead to anomalous P, QRS, and T values, which the patient’s ECG can follow. The growing well-being concerns, especially for cardiac arrhythmias, reflect the requirement of creating a cheap and convenient ECG detection system. The most common heart conditions are tachycardia, bradycardia, atrial fibrillation, myocardial infarction and many other cardiac disorders. These diseases are one of the most significant causes of death worldwide. There is much research for recognizable proof of cardiovascular diseases utilizing ECG diagnostic toolbox and different software. So, in this project, the idea is to design an Automatic Disease Detection System on LabVIEW to detect various cardiac abnormalities such as Atrial Flutter (AF), Supraventricular Tachycardia (SVT), Left Bundle Branch Block (LBBB), Ventricular Fibrillation (VF), Myocardial Infarction (MI), Hyperkalaemia and Digoxin by comparing it to the normal ECG signal. So for this, NI LABVIEW (2018) software has been used for designing. This Automated Disease Detection System, hopefully, gives the best accurate and fast results which are low cost, effective, efficient, easy handled and high accuracy rate detection system which can be used in hospitals for medical professionals.

Performance Comparison of CNN Based Hybrid Systems Using UC Merced Land-Use Dataset

Remote sensing is the technology of collecting and examining data about the earth with special sensors. The data obtained are used in many application areas. The classification success of remote sensing images is closely related to the accuracy and reliability of the information to be used. For this reason, especially in recent studies, it is seen that Convolutional Neural Network (CNN), which has become popular in many fields, is used and high successes have been achieved. However, it is also an important need to obtain this information quickly. Therefore, in this study, it is aimed to get results as successful as CNN and in a shorter time than CNN. Hybrid systems in which features are extracted with CNN and then classification is performed with machine learning algorithms have been tested. The successes of binary combinations of two different CNN architectures (ResNet18, GoogLeNet) and four different classifiers (Support Vector Machine, K Nearest Neighbor, Decision Tree, Discriminant Analysis) have been compared with various metrics. GoogLeNet &amp; Support Vector Machine (93.33%) is the method with the highest accuracy rate, while ResNet18 &amp; Decision Tree (50.95%) is the method with the lowest accuracy rate.

Chronic disease prediction with deep convolution based modified extreme-random forest classifier

A disease is said to be chronic when a disease occurs in an individual and their health conditions due to the disease last one year/more. Predicting chronic diseases have become crucial to save individual’s life and enhance their well-being. Though traditional clinical procedures are considered to perform this, it seems to be a time consuming process. Concurrently, with the progress of data mining algorithms, researchers have attempted to use different algorithms for such prediction. Nevertheless, they have been ineffective in feature extraction that negatively affected the prediction rate. To combat issues with regard to low accuracy rate, the present research intends to perform prediction of four common chronic diseases (breast cancer, heart disease, diabetes and kidney disease) affecting people worldwide. To accomplish this, the research proposes dual Deep CNN (Deep Convolutional Neural Network) for feature extraction. In this case, optimal, maximum and minimum hidden layers are used for extracting relevant features. Further, ME-RF (Modified Extreme-Random Forest) is used for classification. In this process, the research considers XGBoost algorithm comprising of certain innate advantages like high convergence and modest computations. However, when the predictability of this model is poor, it works in an ideal manner with numerous leaves in DT (Decision Tree). Simultaneously, RF comprise of several trees with equal weighted leaves by which, maximum precision and accuracy could be attained flexibly with the prevailing data. Considering this, the trees are built with RF and the research introduces this process as ME-RF. Classification performance is evaluated individually on four different considered datasets under the implementation of minimum and maximum Deep CNN network and also with the use of combined dual deep CNN networks. The overall analytical outcomes confirms the effectiveness of the proposed system.

YOLO v7-ECA-PConv-NWD Detects Defective Insulators on Transmission Lines

This paper proposes an enhanced YOLO v7-based method for detecting insulator defects in transmission lines, addressing the challenges of low accuracy and high leakage rates caused by complex backgrounds and electric poles alongside varying sizes of insulator targets in the image. Firstly, to address the issue of background interference and improve the importance of insulator features, a lightweight attention mechanism named Efficient Channel Attention (ECA) was introduced. With the incorporation of ECA, this model could effectively suppress background noise and provide more focus to insulator regions, thus enhancing its ability to detect insulator defects accurately. Secondly, a partial convolution (PConv) approach was employed in the backbone network instead of conventional convolution, which learned some important channels. This substitution improved both the network model’s accuracy and the training speed. Finally, the Normalized Wasserstein Distance (NWD) prevented insulator features from being lost during pre-feature extraction, which reduced the leakage rate and improved the detection accuracy of small target insulators and defective insulators. The experimental results demonstrate that the improved YOLO v7 network model achieved an average detection accuracy (mAP) of 98.1%, recall of 93.7%, and precision of 96.8% on the TISLTR dataset. On the FISLTR dataset, the average detection accuracy (mAP) for flashover insulators was 93%, with a recall of 92.3% and precision of 87.1%. The average detection accuracy (mAP) for broken insulators was 92.2%, with a recall of 90.3% and a precision of 95.2%. These metrics demonstrate significant improvements in both datasets, highlighting the proposed algorithms’ strong generalization capability and practicable potential to detect insulator targets.

The application of saccades to assess cognitive impairment among older adults: a systematic review and meta-analysis.

Saccade is a novel and feasible method for cognition assessment and has potential to screen older people with cognitive impairment. To systematically summarize the evidence and determine whether different saccade parameters can effectively identify patients with mild cognitive impairment (MCI) and Alzheimer's disease (AD). English and Chinese databases were searched until 19 April 2022. Studies analyzing saccade parameters in older adults with normal cognition, MCI, or AD were included. Two researchers independently performed the screening, data extraction, and quality appraisal. Meta-analyses were conducted and standard mean differences and 95% confidence intervals were estimated with a random effects model. Thirty-five studies were included, and 26 studies were pooled for the meta-analysis. The results demonstrated that patients with cognitive impairment exhibited longer latency and lower accuracy rates in the prosaccade and antisaccade tasks, along with lower corrected error rates in the antisaccade tasks. However, the pooled results for antisaccades were more stable, providing the ability to distinguish patients with cognitive impairment among older adults. The results of the subgroup analyses revealed that only the accuracy rates of the antisaccades differed significantly between people with MCI and AD. Regarding the differences between older adults with normal cognition and those with MCI, the effect sizes of latency and the accuracy rates of saccades as well as the corrected error rates of antisaccades were significant. Saccades, especially antisaccades, are a potential screening and assessment tool for distinguishing older adults with MCI or AD from those with normal cognition.

The data preprocessing in improving the classification quality of network intrusion detection systems

Stream-based intrusion detection is a growing problem in computer network security environments. Many previous researches have applied machine learning as a method to detect attacks in network intrusion detection systems. However, these methods still have limitations of low accuracy and high false alarm rate. To improve the quality of classification, this paper proposes two solutions in the data preprocessing stage, that is, the solution of feature selection and resampling of the training dataset before they are used for training the classifiers. This is based on the fact that there is a lot of class imbalanced data in the training dataset used for network intrusion detection systems, as well as that there are many features in the dataset that are irrelevant to the classification goal, this reduces the quality of classification and increases the computation time. The data after preprocessing by the proposed algorithms is used to train the classifiers using different machine learning algorithms including: Decision Trees, Naive Bayes, Logistic Regression, Support Vector Machines, k Nearest Neighbor and Artificial Neural Network. The training and testing results on the UNSW-NB15 dataset show that: as with the Reconnaissance attack type, the proposed feature selection solution for F-Measure achieves 96.31%, an increase of 19.64%; the proposed oversampling solution for F-Measure achieves 6.99%, an increase of 3.17% and the proposed undersampling solution for F-Measure achieves 94.65%, an increase of 11.42%.

Transfer learning-based fault detection in wind turbine blades using radar plots and deep learning models

ABSTRACT Faults in wind turbine blades are considered a critical issue that can affect the safety and performance of wind turbines. The proposed research aimed to monitor wind turbine blades and identify fault conditions using a transfer learning approach. The study utilized one good and four faulty blade conditions: bend, hub-blade loose connection, erosion, and pitch angle twist. Vibration signals for each blade condition were collected and converted as radar plots that were fed and analyzed using pre-trained deep learning models including ResNet-50, AlexNet, VGG-16, and GoogleNet. Hyperparameters including optimizer, train-test split ratio, batch size, epochs, and learning rate were examined to determine the optimal configuration for each network. The study’s core findings indicate that ResNet-50 outperformed all other models, achieving an impressive accuracy rate of 99.00%. The other models achieved lower accuracy rates, with AlexNet achieving 96.70%, GoogleNet achieving 97.00%, and VGG-16 achieving 95.00%. These findings highlight the potential of using deep learning models for wind turbine monitoring and fault detection, which could significantly improve the efficiency and reliability of wind turbines.

Adaptability of Pediatric Residents for the International League Against Epilepsy-2017 Seizure Classification with a Modular Education Program.

The aim of this study was to evaluate the adaptability of pediatric residents to the current seizure classification of the International League Against Epilepsy-2017 (ILAE-2017) using a modular education program (MEP). The MEP design consisted of 8 modules, including 5 modules for the current version of the ILAE-2017 seizure classification and 3 modules for the older ILAE-1981 version. The MEP was implemented with a group of pediatric residents, and it comprised 50 illustrative pediatric seizure videos along with an instruction manual kit that included a seizure determinator. Following a 3-month follow-up period, a posttest was conducted using 58 new videos in the MEP. The overall success rates of the participants were similar both ILAE-2017 (41%) and ILAE-1981 (38.5%) seizure classifications in the post-MEP test. Regarding the ILAE-2017 mod- ules, the participants demonstrated a higher proficiency in classifying focal nonmotor seizures (56.3%) compared to focal motor seizures (34.9%). However, when it came to generalized seizures, the participants had significantly lower accuracy rates for generalized nonmotor seizures (26%) compared to generalized motor seizures (46%) with the ILAE-2017 classifica- tion. The seizure types that were most commonly misclassified, with an error rate exceeding 50%, were automatisms and myoclonic seizures within the focal seizure modules and atypical absences in generalized seizure modules of ILAE-2017. The single-day MEP yielded modest results, with a success rate of 41% in terms of the initial adaptability of pediatric residents to the ILAE-2017 seizure classification. However, to ensure successful implementation of the ILAE-2017 classification in clinical practice, additional booster applications of the MEP are required.

Harmonizing Heritage and Artificial Neural Networks: The Role of Sustainable Tourism in UNESCO World Heritage Sites

The classification of the United Nations Educational, Scientific, and Cultural Organization (UNESCO) World Heritage Sites (WHS) is essential for promoting sustainable tourism and ensuring the long-term conservation of cultural and natural heritage sites. Therefore, two commonly used techniques for classification problems, multilayer perceptron (MLP) and radial basis function (RBF) neural networks, were utilized to define the pros and cons of their applications. Then, according to the findings, both correlation attribute evaluator (CAE) and relief attribute evaluator (RAE) identified the region and date of inscription as the most prominent features in the classification of UNESCO WHS. As a result, a trade-off condition arises when classifying a large dataset for sustainable tourism between MLP and RBF regarding evaluation time and accuracy. MLP achieves a slightly higher accuracy rate with higher processing time, while RBF achieves a slightly lower accuracy rate but with much faster evaluation time.

CXRNet: CNN‐attention based CXR image classifier

AbstractChest X‐ray (CXR) images are widely accepted for the diagnosis of lung diseases. The X‐ray machinery is widely available but the number of radiologists who interpret these images is very limited. Therefore, the development of an automated disease classification system is a need for the healthcare industry. The existing testing methods take hours to days to generate the testing result and have low detection accuracy and high false detection rate. Furthermore, the testing kits are costly, and availability is limited. Therefore, Convolutional Neural Network (CNN) based framework is proposed to address these limitations. Four pre‐trained frameworks, ResNet50V2, InceptionV3, NASNetMobile, and Xception are used to generate highly dense features, refine the features using the attention module, and then fused these features to classify the diseases using CXR images. In addition, extensive experiments were carried out on the activation functions (Relu, Leaky Relu, and Tanh), which help to improve the results. Evaluation is done on the three enriched CXR datasets DS‐1, DS‐2, and DS‐3 to examine the performance of the proposed framework in terms of binary classification and multi‐class classification. The proposed model achieved a class‐5 accuracy of 92.89% on dataset DS‐3 and the class‐3 classification accuracy of 95.22%, 92.25% on dataset DS‐1 and DS‐2, respectively. The class‐2 classification accuracy on dataset DS‐1 and DS‐2 is 99.23% and 98.85%, respectively.

A Prospective Observational Study of Diagnostic Reliability of Semiquantitative and Quantitative High b-Value Diffusion-Weighted MRI in Distinguishing between Benign and Malignant Lung Lesions at 3 Tesla.

Aim The aim of this study was to evaluate the usefulness of high b-value diffusion-weighted imaging (DWI) to differentiate benign and malignant lung lesions in 3 Tesla magnetic resonance imaging (MRI). Materials and Methods Thirty-one patients with lung lesions underwent a high b-value (b= 1000 s/mm 2 ) DW MRI in 3 Tesla. Thirty lesions were biopsied, followed by histopathological analysis, and one was serially followed up for 2 years. Statistical analysis was done to calculate the sensitivity, specificity, and accuracy of different DWI parameters in distinguishing benign and malignant lesions. Receiver operating characteristic (ROC) curves were used to determine the cutoff values of different parameters. Results The qualitative assessment of signal intensity on DWI based on a 5-point rank scale had a mean score of 2.71 ± 0.75 for benign and 3. 75 ± 0.60 for malignant lesions. With a cutoff of 3.5, the sensitivity, specificity, and accuracy were 75, 86, and 77.6%, respectively. The mean ADC min (minimum apparent diffusion coefficient) value of benign and malignant lesions was 1. 49 ± 0.38 × 10-3 mm 2 /s and 1.11 ± 0.20 ×10-3 mm 2 /s, respectively. ROC curve analysis showed a cutoff value of 1.03 × 10-3 mm 2 /s; the sensitivity, specificity, and accuracy were 87.5, 71.4, and 83.3%, respectively. For lesion to spinal cord ratio and lesion to spinal cord ADC ratio with a cutoff value of 1.08 and 1.38, the sensitivity, specificity, and accuracy were 83.3 and 87.5%, 71.4 and 71.4%, and 80.6 and 83.8%, respectively. The exponential ADC showed a low accuracy rate. Conclusion The semiquantitative and quantitative parameters of high b-value DW 3 Tesla MRI can differentiate benign from malignant lesions with high accuracy and make it a reliable nonionizing modality for characterizing lung lesions.

Probabilistic neural network-based diagnosis of blockage faults in air conditioning and refrigeration units

Abstract The refrigeration unit is a core component of the air conditioner, and its typical faults are pipe scaling and blockage, which can cause refrigerant leakage in serious cases, and these faults can affect the cooling effect of the air conditioner and even affect the safety of using the air conditioner in serious cases. Therefore, it is necessary to study the fault diagnosis system of the refrigeration unit. The traditional fault diagnosis system for refrigeration units, which generally uses wireless sensor networks to transmit data, is a challenge because of its slow diagnosis speed and the low accuracy rate of fault diagnosis for refrigeration units. This paper uses a probabilistic neural network (PNN) to debug a blockage fault diagnosis algorithm system for simulated fault training. The system successfully achieves four levels of pre-warning for blockage faults: “uncertainty, micro-blockage, blockage, and severe blockage”, with the fastest pre-warning time of “uncertainty, micro-blockage” being less than 2 seconds. The experimental results of this paper show that the PNN system can take timely measures against blockage faults, and the fault diagnosis accuracy of the model reaches 97.97%, which is 13.34% higher than the traditional fault identification algorithm. The PNN model can pinpoint the blockage faults of the refrigeration unit, which enables the working state of the refrigeration unit to be continuously optimized, and the air conditioner to maintain energy-efficient operation. The fault simulation of the air conditioning refrigeration unit in automatic control mode verifies the blockage fault diagnosis’s accuracy and the algorithm’s feasibility, providing reference and learning for the development of fault diagnosis algorithms.