Heba Kadry, Ahmed H. Samak, Sara Ghorashi, Sarah M. Alhammad, Abdulwahab Abukwaik, Ahmed I. Taloba, Elnomery A. Zanaty
Coronavirus is a new pathogen that causes both the upper and lower respiratory systems. The global COVID-19 pandemic’s size, rate of transmission, and the number of deaths is all steadily rising. COVID-19 instances could be detected and analyzed using Computed Tomography scanning. For the identification of lung infection, chest CT imaging has the advantages of speedy detection, relatively inexpensive, and high sensitivity. Due to the obvious minimal information available and the complicated image features, COVID-19 identification is a difficult process. To address this problem, modified-Deformed Entropy (QDE) algorithm for CT image scanning is suggested. To enhance the number of training samples for effective testing and training, the suggested method utilizes QDE to generate CT images. The retrieved features are used to classify the results. Rapid innovations in quantum mechanics had prompted researchers to use Quantum Machine Learning (QML) to test strategies for improvement. Furthermore, the categorization of corona diagnosed, and non-diagnosed pictures is accomplished through Quanvolutional Neural Network (QNN). To determine the suggested techniques, the results are related with other methods. For processing the COVID-19 imagery, the study relates QNN with other existing methods. On comparing with other models, the suggested technique produced improved outcomes. Also, with created COVID-19 CT images, the suggested technique outperforms previous state-of-the-art image synthesis techniques, indicating possibilities for different machine learning techniques such as cognitive segmentation and classification. As a result of the improved model training/testing, the image classification results are more accurate.
{"title":"Implementation of a quantum machine learning model for the categorization and analysis of COVID-19 cases","authors":"Heba Kadry, Ahmed H. Samak, Sara Ghorashi, Sarah M. Alhammad, Abdulwahab Abukwaik, Ahmed I. Taloba, Elnomery A. Zanaty","doi":"10.3233/jifs-233633","DOIUrl":"https://doi.org/10.3233/jifs-233633","url":null,"abstract":"Coronavirus is a new pathogen that causes both the upper and lower respiratory systems. The global COVID-19 pandemic’s size, rate of transmission, and the number of deaths is all steadily rising. COVID-19 instances could be detected and analyzed using Computed Tomography scanning. For the identification of lung infection, chest CT imaging has the advantages of speedy detection, relatively inexpensive, and high sensitivity. Due to the obvious minimal information available and the complicated image features, COVID-19 identification is a difficult process. To address this problem, modified-Deformed Entropy (QDE) algorithm for CT image scanning is suggested. To enhance the number of training samples for effective testing and training, the suggested method utilizes QDE to generate CT images. The retrieved features are used to classify the results. Rapid innovations in quantum mechanics had prompted researchers to use Quantum Machine Learning (QML) to test strategies for improvement. Furthermore, the categorization of corona diagnosed, and non-diagnosed pictures is accomplished through Quanvolutional Neural Network (QNN). To determine the suggested techniques, the results are related with other methods. For processing the COVID-19 imagery, the study relates QNN with other existing methods. On comparing with other models, the suggested technique produced improved outcomes. Also, with created COVID-19 CT images, the suggested technique outperforms previous state-of-the-art image synthesis techniques, indicating possibilities for different machine learning techniques such as cognitive segmentation and classification. As a result of the improved model training/testing, the image classification results are more accurate.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136231939","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The worldwide deforestation rate worsens year after year, ultimately resulting in a variety of severe implications for both mankind and the environment. In order to track the success of forest preservation activities, it is crucial to establish a reliable forest monitoring system. Changes in forest status are extremely difficult to manually annotate due to the tiny size and subtlety of the borders involved, particularly in regions abutting residential areas. Previous forest monitoring systems failed because they relied on low-resolution satellite images and drone-based data, both of which have inherent limitations. Most government organizations still use manual annotation, which is a slow, laborious, and costly way to keep tabs on data. The purpose of this research is to find a solution to these problems by building a poly-highway forest convolution network using deep learning to automatically detect forest borders so that changes over time may be monitored. Here initially the data was curated using the dynamic decomposed kalman filter. Then the data can be augmented. Afterward the augmented image features can be fused using the multimodal discriminant centroid feature clustering. Then the selected area can be segmented using the iterative initial seeded algorithm (IISA). Finally, the level and the driver of deforestation can be classified using the poly-highway forest convolution network (PHFCN). The whole experimentation was carried out in a dataset of 6048 Landsat-8 satellite sub-images under MATLAB environment. From the result obtained the suggested methodology express satisfied performance than other existing mechanisms.
{"title":"Classifying drivers of deforestation by using the deep learning based poly-highway forest convolution network","authors":"D. Abdus Subhahan, C.N.S. Vinoth Kumar","doi":"10.3233/jifs-233534","DOIUrl":"https://doi.org/10.3233/jifs-233534","url":null,"abstract":"The worldwide deforestation rate worsens year after year, ultimately resulting in a variety of severe implications for both mankind and the environment. In order to track the success of forest preservation activities, it is crucial to establish a reliable forest monitoring system. Changes in forest status are extremely difficult to manually annotate due to the tiny size and subtlety of the borders involved, particularly in regions abutting residential areas. Previous forest monitoring systems failed because they relied on low-resolution satellite images and drone-based data, both of which have inherent limitations. Most government organizations still use manual annotation, which is a slow, laborious, and costly way to keep tabs on data. The purpose of this research is to find a solution to these problems by building a poly-highway forest convolution network using deep learning to automatically detect forest borders so that changes over time may be monitored. Here initially the data was curated using the dynamic decomposed kalman filter. Then the data can be augmented. Afterward the augmented image features can be fused using the multimodal discriminant centroid feature clustering. Then the selected area can be segmented using the iterative initial seeded algorithm (IISA). Finally, the level and the driver of deforestation can be classified using the poly-highway forest convolution network (PHFCN). The whole experimentation was carried out in a dataset of 6048 Landsat-8 satellite sub-images under MATLAB environment. From the result obtained the suggested methodology express satisfied performance than other existing mechanisms.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136232081","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Vo Thi Hong Tuyet, Nguyen Thanh Binh, Dang Thanh Tin
With the medical internet of things, many automated diagnostic models related to eye diseases are easier. The doctors could quickly contrast and compare retina fundus images. The retina image contains a lot of information in the image. The task of detecting diabetic macular edema from retinal images in the healthcare system is difficult because the details in these images are very small. This paper proposed the new model based on the medical internet of things for predicting diabetic macular edema in retina fundus images. The method called DMER (Diabetic Macular Edema in Retina fundus images) to detect diabetic macular edema in retina fundus images based on improving deep residual network being combined with feature pyramid network in the context of the medical internet of things. The DMER method includes the following stages: (i) ResNet101 improved combining with feature pyramid network is used to extract features of the image and obtain the map of these features; (ii) a region proposal network to look for potential anomalies; and (iii) the predicted bounding boxes against the true bounding box by the regression method to certify the capability of macular edema. The MESSIDOR and DIARETDB1 datasets are used for testing with evaluation criteria such as sensitivity, specificity, and accuracy. The accuracy of the DMER method is about 98.08% with MESSIDOR dataset and 98.92% with DIARETDB1 dataset. The results of the method DMER are better than those of the other methods up to the present time with the above datasets.
随着医疗物联网的发展，许多与眼病相关的自动诊断模型变得更加容易。医生可以快速对比和比较视网膜眼底图像。视网膜图像中包含了大量的图像信息。从视网膜图像中检测糖尿病黄斑水肿的任务在医疗系统中是困难的，因为这些图像中的细节非常小。本文提出了一种基于医疗物联网的糖尿病黄斑水肿视网膜眼底图像预测模型。在医疗物联网背景下，基于改进的深度残差网络与特征金字塔网络相结合，检测视网膜眼底图像中糖尿病黄斑水肿的方法称为DMER (Diabetic Macular Edema in Retina fundus images)。DMER方法包括以下几个阶段:(i)利用改进的ResNet101结合特征金字塔网络提取图像的特征，得到这些特征的映射;(ii)寻找潜在异常的区域建议网络;(3)用回归法将预测边界框与真实边界框对比，证明黄斑水肿的能力。MESSIDOR和DIARETDB1数据集用于测试评估标准，如敏感性、特异性和准确性。DMER方法在MESSIDOR数据集和DIARETDB1数据集上的准确率分别为98.08%和98.92%。对于上述数据集，DMER方法的结果优于目前其他方法的结果。
{"title":"Predicting diabetic macular edema in retina fundus images based on optimized deep residual network techniques on medical internet of things","authors":"Vo Thi Hong Tuyet, Nguyen Thanh Binh, Dang Thanh Tin","doi":"10.3233/jifs-234649","DOIUrl":"https://doi.org/10.3233/jifs-234649","url":null,"abstract":"With the medical internet of things, many automated diagnostic models related to eye diseases are easier. The doctors could quickly contrast and compare retina fundus images. The retina image contains a lot of information in the image. The task of detecting diabetic macular edema from retinal images in the healthcare system is difficult because the details in these images are very small. This paper proposed the new model based on the medical internet of things for predicting diabetic macular edema in retina fundus images. The method called DMER (Diabetic Macular Edema in Retina fundus images) to detect diabetic macular edema in retina fundus images based on improving deep residual network being combined with feature pyramid network in the context of the medical internet of things. The DMER method includes the following stages: (i) ResNet101 improved combining with feature pyramid network is used to extract features of the image and obtain the map of these features; (ii) a region proposal network to look for potential anomalies; and (iii) the predicted bounding boxes against the true bounding box by the regression method to certify the capability of macular edema. The MESSIDOR and DIARETDB1 datasets are used for testing with evaluation criteria such as sensitivity, specificity, and accuracy. The accuracy of the DMER method is about 98.08% with MESSIDOR dataset and 98.92% with DIARETDB1 dataset. The results of the method DMER are better than those of the other methods up to the present time with the above datasets.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136232085","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hongyue Guo, Qiqi Deng, Wenjuan Jia, Lidong Wang, Cong Sui
The implied volatility plays a pivotal role in the options market, and a collection of implied volatilities across strike and maturity is known as the implied volatility surface (IVS). To capture the dynamics of IVS, this study examines the latent states of IVS and their relationship based on the regime-switching framework of the hidden Markov model (HMM). The cross-sectional models are first built for daily implied volatilities, and the obtained regression factors are regarded as the proxies of the IVS. Then, having these latent factors, the HMM is employed to model the dynamics of IVS. Take the advantages of HMM, the hidden state for each daily data is identified to achieve the corresponding time distribution, the characteristics, and the transition between the hidden states. The empirical study is conducted on the Shanghai 50ETF options, and the analysis results indicate that the HMM can capture the latent factors of IVS. The achieved states reflect different financial characteristics, and some of their typical features and transfer are associated with certain events. In addition, the HMM exploited to predict the regression factors of the cross-sectional models enables the further forecasting of implied volatilities. The autoregressive integrated moving average model, the vector auto-regression model, and the support vector regression model are regarded as benchmarks for comparison. The results show that the HMM performs better in the implied volatility prediction compared with other models.
{"title":"Hidden Markov model-based modeling and prediction for implied volatility surface1","authors":"Hongyue Guo, Qiqi Deng, Wenjuan Jia, Lidong Wang, Cong Sui","doi":"10.3233/jifs-232139","DOIUrl":"https://doi.org/10.3233/jifs-232139","url":null,"abstract":"The implied volatility plays a pivotal role in the options market, and a collection of implied volatilities across strike and maturity is known as the implied volatility surface (IVS). To capture the dynamics of IVS, this study examines the latent states of IVS and their relationship based on the regime-switching framework of the hidden Markov model (HMM). The cross-sectional models are first built for daily implied volatilities, and the obtained regression factors are regarded as the proxies of the IVS. Then, having these latent factors, the HMM is employed to model the dynamics of IVS. Take the advantages of HMM, the hidden state for each daily data is identified to achieve the corresponding time distribution, the characteristics, and the transition between the hidden states. The empirical study is conducted on the Shanghai 50ETF options, and the analysis results indicate that the HMM can capture the latent factors of IVS. The achieved states reflect different financial characteristics, and some of their typical features and transfer are associated with certain events. In addition, the HMM exploited to predict the regression factors of the cross-sectional models enables the further forecasting of implied volatilities. The autoregressive integrated moving average model, the vector auto-regression model, and the support vector regression model are regarded as benchmarks for comparison. The results show that the HMM performs better in the implied volatility prediction compared with other models.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136232612","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Writing is a crucial component of the language requirement and is an effective method for correctly reflecting language proficiency. Manually evaluating Tamil language exams becomes time-consuming and costly for standardized language administrators as they grow in popularity. Numerous studies on computerized English assessment systems have been conducted in recent years. Due to Tamil text’s complicated grammatical structures, less research has been done on computerized evaluation methods. In this research, we present a Tamil review comment analysis system using a novel multivariate naïve Bayes classifier (mv - NB) where the comments are acquired from an online social network and performed training using the database for further analysis. Experiments show that the graded Kappa of 0.4239, error rate of 2.55 and precision of 85% was achieved on the online dataset by our contents grading system, which is superior in grading compared to the other widely used machine learning algorithms training on big datasets. Our findings are promising. Additionally, our contents analysis may provide beneficial criticism on Tamil writing on YouTube posts including comments, spelling errors and morphological issues that help to analyze thelanguage correlation.
{"title":"Modelling a machine learning based multivariate content grading system for YouTube Tamil-post analysis","authors":"G. Srivatsun, S. Thivaharan","doi":"10.3233/jifs-222504","DOIUrl":"https://doi.org/10.3233/jifs-222504","url":null,"abstract":"Writing is a crucial component of the language requirement and is an effective method for correctly reflecting language proficiency. Manually evaluating Tamil language exams becomes time-consuming and costly for standardized language administrators as they grow in popularity. Numerous studies on computerized English assessment systems have been conducted in recent years. Due to Tamil text’s complicated grammatical structures, less research has been done on computerized evaluation methods. In this research, we present a Tamil review comment analysis system using a novel multivariate naïve Bayes classifier (mv - NB) where the comments are acquired from an online social network and performed training using the database for further analysis. Experiments show that the graded Kappa of 0.4239, error rate of 2.55 and precision of 85% was achieved on the online dataset by our contents grading system, which is superior in grading compared to the other widely used machine learning algorithms training on big datasets. Our findings are promising. Additionally, our contents analysis may provide beneficial criticism on Tamil writing on YouTube posts including comments, spelling errors and morphological issues that help to analyze thelanguage correlation.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136232613","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Point-of-Interest (POI) recommendation is one of the most important tasks in the field of social network analysis. Many efforts have been proposed to enhance the model performance for the POI recommendation task in recent years. Existing studies have revealed that the temporal factor and geographical factor are two crucial contextual factors which influence user decisions. However, they only learn representations of POIs and users from the single contextual factor and fuse the learned representations in the final stage, which ignores the interactions of different contextual factors, leading to learning suboptimal representations of POIs and users. To overcome this gap, we propose a novel Temporal-Geographical Attention-based Transformer (TGAT) for the POI recommendation task. Specifically, TGAT develops a hybrid sequence sampling strategy that samples the sequence of POIs from the different contextual factor POI graphs generated by the users’ check-in records. In this way, the interactions of different contextual factors can be care-fully pre-served. Then TGAT conducts a Transformer-based neural network backbone to learn representations of POIs from the sampling sequences. In addition, a weighted aggregation strategy is proposed to fuse the representations learned from different context factors. The extensive experimental results on real-world datasets have demonstrated the effectiveness of TGAT.
{"title":"Temporal-geographical attention-based transformer for point-of-interest recommendation","authors":"Shaojie Jiang, Jiang Wu","doi":"10.3233/jifs-234824","DOIUrl":"https://doi.org/10.3233/jifs-234824","url":null,"abstract":"Point-of-Interest (POI) recommendation is one of the most important tasks in the field of social network analysis. Many efforts have been proposed to enhance the model performance for the POI recommendation task in recent years. Existing studies have revealed that the temporal factor and geographical factor are two crucial contextual factors which influence user decisions. However, they only learn representations of POIs and users from the single contextual factor and fuse the learned representations in the final stage, which ignores the interactions of different contextual factors, leading to learning suboptimal representations of POIs and users. To overcome this gap, we propose a novel Temporal-Geographical Attention-based Transformer (TGAT) for the POI recommendation task. Specifically, TGAT develops a hybrid sequence sampling strategy that samples the sequence of POIs from the different contextual factor POI graphs generated by the users’ check-in records. In this way, the interactions of different contextual factors can be care-fully pre-served. Then TGAT conducts a Transformer-based neural network backbone to learn representations of POIs from the sampling sequences. In addition, a weighted aggregation strategy is proposed to fuse the representations learned from different context factors. The extensive experimental results on real-world datasets have demonstrated the effectiveness of TGAT.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136232084","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Multiple unmanned aerial vehicles (multi-UAVs) formation shape refers to the geometric shape when multi-UAVs fly in formation and describes their relative positions. It plays a necessary role in multi-UAVs collaboration to improve performance, avoid collision, and provide reference for control. This study aims to determine the most appropriate multi-UAVs formation shape in a specific mission to meet different and even conflicting requirements. The proposed approach introduces requirement satisfaction and spherical fuzzy analytic network process (SFANP) to improve the technique for order preference by similarity to ideal solution (TOPSIS). First, multi-UAVs capability criteria and their evaluation models are constructed. Next, performance data are transformed into requirement satisfaction of capability and unified into a same scale. Qualitative judgments are made and quantified based on spherical fuzzy sets and nonlinear transformation functions are developed for benefit, cost, and interval metrics. Then, SFANP is used to handle interrelationships among criteria and determine their global weights, which takes decision vagueness and hesitancy into account and extends decision-makers’ preference domain onto a spherical surface. Finally, alternative formation shapes are ranked by their distances to the positive and negative ideal solution according to the TOPSIS. Furthermore, a case study of 9 UAVs performing a search-attack mission is set up to illustrate the proposed approach, and a comparative analysis is conducted to verify the applicability and credibility.
{"title":"Determination of multi-UAVs formation shape: Using a requirement satisfaction and spherical fuzzy ANP based TOPSIS approach","authors":"An Zhang, Minghao Li, Wenhao Bi","doi":"10.3233/jifs-231494","DOIUrl":"https://doi.org/10.3233/jifs-231494","url":null,"abstract":"Multiple unmanned aerial vehicles (multi-UAVs) formation shape refers to the geometric shape when multi-UAVs fly in formation and describes their relative positions. It plays a necessary role in multi-UAVs collaboration to improve performance, avoid collision, and provide reference for control. This study aims to determine the most appropriate multi-UAVs formation shape in a specific mission to meet different and even conflicting requirements. The proposed approach introduces requirement satisfaction and spherical fuzzy analytic network process (SFANP) to improve the technique for order preference by similarity to ideal solution (TOPSIS). First, multi-UAVs capability criteria and their evaluation models are constructed. Next, performance data are transformed into requirement satisfaction of capability and unified into a same scale. Qualitative judgments are made and quantified based on spherical fuzzy sets and nonlinear transformation functions are developed for benefit, cost, and interval metrics. Then, SFANP is used to handle interrelationships among criteria and determine their global weights, which takes decision vagueness and hesitancy into account and extends decision-makers’ preference domain onto a spherical surface. Finally, alternative formation shapes are ranked by their distances to the positive and negative ideal solution according to the TOPSIS. Furthermore, a case study of 9 UAVs performing a search-attack mission is set up to illustrate the proposed approach, and a comparative analysis is conducted to verify the applicability and credibility.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136232091","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Early detection of diseases in men and women can improve treatment and reduce the risk involved in human life. Nowadays techniques which are non-invasive in nature are popularly used to detect the various types of diseases. Histopathological analysis plays a major role in finding the nature of the disease through medical images. Manual interpretation of these medical imaging takes time, is tedious, subjective, and can have human errors. It has also been discovered that the interpretation of these images varies amongst diagnostic labs. As computer power and memory capacity have increased, methodologies and medical image processing techniques have been developed to interpret and analyse these images as a substitute for human involvement. The challenge lies in devising an efficient pre-processing technique that helps in analysing, processing and preparing the medical image for further diagnostics. This research provides a hybrid technique that reduces noise in the NITFI medical image by using a 2D adaptive median filter at level 1. The edges of the filtered medical image are preserved using the modified CLAHE algorithm which preserves the local contrast of the image. Expectation Maximization (EM) algorithm extracts the ROI part of the image which helps in easy and accurate identification of the disease. All the three steps are run over the 3D image slices of a NIFTI image. The proposed method proves that it achieves close to ideal RMSE, PSNR and UQI values as well as achieves an average runtime of 37.193 seconds for EM per slice.
{"title":"Medical image segmentation using hybrid Contrast Limited Adaptive Histogram Equalization (CLAHE) and expectation maximization algorithm","authors":"K.C. Prabu Shankar, S. Prayla Shyry","doi":"10.3233/jifs-233931","DOIUrl":"https://doi.org/10.3233/jifs-233931","url":null,"abstract":"Early detection of diseases in men and women can improve treatment and reduce the risk involved in human life. Nowadays techniques which are non-invasive in nature are popularly used to detect the various types of diseases. Histopathological analysis plays a major role in finding the nature of the disease through medical images. Manual interpretation of these medical imaging takes time, is tedious, subjective, and can have human errors. It has also been discovered that the interpretation of these images varies amongst diagnostic labs. As computer power and memory capacity have increased, methodologies and medical image processing techniques have been developed to interpret and analyse these images as a substitute for human involvement. The challenge lies in devising an efficient pre-processing technique that helps in analysing, processing and preparing the medical image for further diagnostics. This research provides a hybrid technique that reduces noise in the NITFI medical image by using a 2D adaptive median filter at level 1. The edges of the filtered medical image are preserved using the modified CLAHE algorithm which preserves the local contrast of the image. Expectation Maximization (EM) algorithm extracts the ROI part of the image which helps in easy and accurate identification of the disease. All the three steps are run over the 3D image slices of a NIFTI image. The proposed method proves that it achieves close to ideal RMSE, PSNR and UQI values as well as achieves an average runtime of 37.193 seconds for EM per slice.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136232341","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In today’s complex decision-making environment, accounting for attribute interdependencies and expert relationships is crucial. Traditional models often assume attribute independence and overlook the significant impact of expert relationships on decision outcomes. Also, amidst the dynamic and ever-changing decision-making landscape, the effect of news and real-time updates on alternative rankings is significant. In complex decision-making environments, information is constantly evolving, and staying up-to-date with the latest developments is paramount. To overcome these limitations, this study aims to develop a novel model that effectively captures attribute dependencies and incorporates the influence of social media on alternative ordering. To establish the model, the Decision-making trial and evaluation laboratory (DEMATEL) method and regression analysis are integrated to capture attribute dependencies. Furthermore, social network analysis (SNA) is employed to develop a trust propagation model for determining experts’ weights. Additionally, we present a two-stage multi-skilled and high potential multi-criteria decision-making (MCDM) framework, where the base-criterion method (BCM) is adopted to evaluate attribute weights and the well-known traditional Vlekriterijumsko KOmpromisno Rangiranje (VIKOR) method is redefined using Heronian mean (HM) operator to capture the relationships between arguments. Despite uncertainties, the proposed fuzzy-BCM-VIKOR-Heronian (F-BCM-VIKOR-H) approach enhances flexibility by addressing inconsistent data in complex decision-making problems. Similarly, certain news or future updates about any alternative or attribute can significantly affect the ranking. Acknowledging the significance of timely information, the proposed approach actively considers the effect of such news through the formation of an updated matrix. By factoring in the latest developments, we ensure that the proposed decision-making model remains relevant and adaptable, capturing the most current insights into alternative performance. To demonstrate the model’s effectiveness, we apply the proposed approach to a numerical illustration in the electronics industry, specifically for ranking cars. Sensitivity analysis evaluates the model’s stability, and comparing the results with existing approaches showcases its advantage and superiority.
{"title":"Functional dependency-based group decision-making with incomplete information under social media influence: An application to automobile","authors":"Garima Bisht, A.K. Pal","doi":"10.3233/jifs-232608","DOIUrl":"https://doi.org/10.3233/jifs-232608","url":null,"abstract":"In today’s complex decision-making environment, accounting for attribute interdependencies and expert relationships is crucial. Traditional models often assume attribute independence and overlook the significant impact of expert relationships on decision outcomes. Also, amidst the dynamic and ever-changing decision-making landscape, the effect of news and real-time updates on alternative rankings is significant. In complex decision-making environments, information is constantly evolving, and staying up-to-date with the latest developments is paramount. To overcome these limitations, this study aims to develop a novel model that effectively captures attribute dependencies and incorporates the influence of social media on alternative ordering. To establish the model, the Decision-making trial and evaluation laboratory (DEMATEL) method and regression analysis are integrated to capture attribute dependencies. Furthermore, social network analysis (SNA) is employed to develop a trust propagation model for determining experts’ weights. Additionally, we present a two-stage multi-skilled and high potential multi-criteria decision-making (MCDM) framework, where the base-criterion method (BCM) is adopted to evaluate attribute weights and the well-known traditional Vlekriterijumsko KOmpromisno Rangiranje (VIKOR) method is redefined using Heronian mean (HM) operator to capture the relationships between arguments. Despite uncertainties, the proposed fuzzy-BCM-VIKOR-Heronian (F-BCM-VIKOR-H) approach enhances flexibility by addressing inconsistent data in complex decision-making problems. Similarly, certain news or future updates about any alternative or attribute can significantly affect the ranking. Acknowledging the significance of timely information, the proposed approach actively considers the effect of such news through the formation of an updated matrix. By factoring in the latest developments, we ensure that the proposed decision-making model remains relevant and adaptable, capturing the most current insights into alternative performance. To demonstrate the model’s effectiveness, we apply the proposed approach to a numerical illustration in the electronics industry, specifically for ranking cars. Sensitivity analysis evaluates the model’s stability, and comparing the results with existing approaches showcases its advantage and superiority.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136231812","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Brain tumor is observed to be grown in irregular shape and presented deep inside the tissues that led to cancer. Human brain tumor identification and categorization are performed with high latency, but also an essential task for the medical experts. The assistance through the automated diagnosis is generally utilized for the advancement in the diagnosis ability in order to get superior accuracy in brain tumor detection. Although the researches are enhancing the brain tumor detection performance, the highly challenging is to segment the brain tumor since it has variability concerning the tumor type, contrast, image modality and also in other factors. To meet up all the challenges, a novel classification method is introduced using segmentation and machine learning approaches. Initially, the required images are collected from benchmark data sources. The input images are undergone for pre-processing stage, where it is done via “Contrast Limited Adaptive Histogram Equalization (CLAHE) and filtering methods”. Further, the pre-processed imagesare given as input to two classifier models as “Residual Network (ResNet) and Gated Recurrent Unit (GRU)”, in which the model provide the result as normal and abnormal images. In the second part, obtained abnormal image acts an input for segmentation step. In segmentation, it is needed to extract the relevant features by texture and spatial features. The resultant features are subjected for optimizing, where the optimal features are acquired through Adaptive Coyote Optimization Algorithm (ACOA). Then, the extracted features are fed into machine learning model like “Support Vector Machine (SVM), Artificial Neural Network (ANN), and Random Forest (RF)” to render the segmented image. Finally, the hybrid classification named Hybrid ResGRUis developed by integrating the ResNet and GRU, where the hyper parameters are tuned optimally using developed ACOA, thus it is used for classifying the abnormal image that belongs to benign stage or malignant stage. The experimental results are evaluated, and its performance is analyzed by various metrics. Hence, the proposed classification model ensures effective segmentation and classification performance.
{"title":"Hybrid ResGRU: Effective brain tumour classification approach using of abnormal images","authors":"Aishwarya Rajendran, Sumathi Ganesan, T.K.S. Rathis Babu","doi":"10.3233/jifs-233546","DOIUrl":"https://doi.org/10.3233/jifs-233546","url":null,"abstract":"Brain tumor is observed to be grown in irregular shape and presented deep inside the tissues that led to cancer. Human brain tumor identification and categorization are performed with high latency, but also an essential task for the medical experts. The assistance through the automated diagnosis is generally utilized for the advancement in the diagnosis ability in order to get superior accuracy in brain tumor detection. Although the researches are enhancing the brain tumor detection performance, the highly challenging is to segment the brain tumor since it has variability concerning the tumor type, contrast, image modality and also in other factors. To meet up all the challenges, a novel classification method is introduced using segmentation and machine learning approaches. Initially, the required images are collected from benchmark data sources. The input images are undergone for pre-processing stage, where it is done via “Contrast Limited Adaptive Histogram Equalization (CLAHE) and filtering methods”. Further, the pre-processed imagesare given as input to two classifier models as “Residual Network (ResNet) and Gated Recurrent Unit (GRU)”, in which the model provide the result as normal and abnormal images. In the second part, obtained abnormal image acts an input for segmentation step. In segmentation, it is needed to extract the relevant features by texture and spatial features. The resultant features are subjected for optimizing, where the optimal features are acquired through Adaptive Coyote Optimization Algorithm (ACOA). Then, the extracted features are fed into machine learning model like “Support Vector Machine (SVM), Artificial Neural Network (ANN), and Random Forest (RF)” to render the segmented image. Finally, the hybrid classification named Hybrid ResGRUis developed by integrating the ResNet and GRU, where the hyper parameters are tuned optimally using developed ACOA, thus it is used for classifying the abnormal image that belongs to benign stage or malignant stage. The experimental results are evaluated, and its performance is analyzed by various metrics. Hence, the proposed classification model ensures effective segmentation and classification performance.","PeriodicalId":54795,"journal":{"name":"Journal of Intelligent & Fuzzy Systems","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-10-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136232345","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：