Machine learning and knowledge extraction, Jul 9, 2020
This research presents a machine vision approach to detect lesions in liver ultrasound as well as... more This research presents a machine vision approach to detect lesions in liver ultrasound as well as resolving some issues in ultrasound such as artifacts, speckle noise, and blurring effect. The anisotropic diffusion is modified using the edge preservation conditions which found better than traditional ones in quantitative evolution. To dig for more potential information, a learnable super-resolution (SR) is embedded into the deep CNN. The feature is fused using Gabor Wavelet Transform (GWT) and Local Binary Pattern (LBP) with a pre-trained deep CNN model. Moreover, we propose a Bayes rule-based informative patch selection approach to reduce the processing time with the selective image patches and design an algorithm to mark the lesion region from identified ultrasound image patches. To train this model, standard data ensures promising resolution. The testing phase considers generalized data with a varying resolution and test the performance of the model. Exploring cross-validation, it finds that a 5-fold strategy can successfully eradicate the overfitting problem. Experiment data are collected using 298 consecutive ultrasounds comprising 15,296 image patches. This proposed feature fusion technique confirms satisfactory performance compared to the current relevant works with an accuracy of 98.40%.
International journal of imaging and robotics, Oct 24, 2017
This paper presents an approach to speckle noise reduction for 3D ultrasound images using bi-dime... more This paper presents an approach to speckle noise reduction for 3D ultrasound images using bi-dimensional empirical mode decomposition (BEMD). 3D ultrasound is a popular diagnostic system for assessing the progression of diseases for its diverse benefits and application. Speckle noise often obscures the fine details and degrades the spatial resolution and, contrast quality that makes the interpretation of ultrasound image more difficult. The proposed method estimates an optimum threshold value of intrinsic mode functions (IMFs) using Fisher discriminant analysis (FDA) for reducing the speckles in 3D volume of ultrasound images. FDA has applied on 2D IMFs, then explored and extended to 3D. The 3D volume rendering is performed on the basis of integrating 2D slice images that provide strong speckle reduction and edge preservation. The experiment result has compared with the several other state-of-the-art threshold methods. The proposed method is also good in edge preservation and contrast resolution.
International Journal of Signal and Imaging Systems Engineering, 2018
This paper presents an approach to speckle noise reduction for 3D ultrasound images using bi-dime... more This paper presents an approach to speckle noise reduction for 3D ultrasound images using bi-dimensional empirical mode decomposition (BEMD). 3D ultrasound is a popular diagnostic system for assessing the progression of diseases for its diverse benefits and application. Speckle noise often obscures the fine details and degrades the spatial resolution and, contrast quality that makes the interpretation of ultrasound image more difficult. The proposed method estimates an optimum threshold value of intrinsic mode functions (IMFs) using Fisher discriminant analysis (FDA) for reducing the speckles in 3D volume of ultrasound images. FDA has applied on 2D IMFs, then explored and extended to 3D. The 3D volume rendering is performed on the basis of integrating 2D slice images that provide strong speckle reduction and edge preservation. The experiment result has compared with the several other state-of-the-art threshold methods. The proposed method is also good in edge preservation and contrast resolution.
International Journal of Ambient Computing and Intelligence, Mar 18, 2022
The brain tumor is one of the most health hazard diseases across the world in recent time. The de... more The brain tumor is one of the most health hazard diseases across the world in recent time. The development of the intelligent system has extended its applications in the automated medical diagnosis domains. However, image-based medical diagnosis result strongly depends on the selection of relevant features. This research focuses on the automatic detection of brain tumors based on the concatenation of curvelet transform and convolutional neural network (CNN) features extracted from the preprocessed MRI sequence of the brain. Relevant features are selected from the feature vector using mutual information based on the minimum redundancy maximum relevance (mRMR) method. The detection is done using the ensemble classifier of the bagging method. The experiment is performed using two standard datasets of BraTS 2018 and BraTS 2019. After five-fold cross-validation, we have obtained an outperforming accuracy of 98.96%.
COVID-19 is one of the most life-threatening and dangerous diseases caused by the novel Coronavir... more COVID-19 is one of the most life-threatening and dangerous diseases caused by the novel Coronavirus, which has already afflicted a larger human community worldwide. This pandemic disease recovery is possible if detected in the early stage. We proposed an automated deep learning approach from Computed Tomography (CT) scan images to detect COVID-19 positive patients by following a fourphase paradigm for COVID-19 detection: preprocess the CT scan images; remove noise from test image by using anisotropic diffusion techniques; make a different segment for the preprocessed images; and train and test COVID-19 detection using Convolutional Neural Network (CNN) models. This study employed well-known pre-trained models, including AlexNet, ResNet50, VGG16 and VGG19 to evaluate experiments. 80% of images are used to train the network in the detection process, while the remaining 20% are used to test it. The result of the experiment evaluation confirmed that the VGG19 pre-trained CNN model achieved better accuracy (98.06%). We used 4861 real-life COVID-19 CT images for experiment purposes, including 3068 positive and 1793 negative images. These images were acquired from a hospital in Sao Paulo, Brazil and two other different data sources. Our proposed method revealed very high accuracy and, therefore, can be used as an assistant to help professionals detect COVID-19 patients accurately. COVID-19 1 , also known as the Coronavirus, was spread in Wuhan, China. Within four months of its emergence in the year 2020, the World Health Organization (WHO) declared it a pandemic 2-4. Confirmed cases and deaths are recorded as 538,246,806 and 6,327,036, respectively, and these numbers are increasing day by day and until June 2022, 230 countries are currently being affected by COVID-19 5. As early COVID-19 symptoms, sometimes patients feel difficulty in breathing and do vomiting. Sneeze and cough droplets from an infected individual can easily spread from one person to another. With the large number of patients infected by COVID-19 during the pandemic, it was impossible for health experts and the competent authorities to assure enough testing kits for each. Besides, there is a shortage of kits to find out infected people 6. If more tests occur, it gets easy to find out more COVID-19 affected people and help not spread Coronavirus. One of the most used methods for diagnosing COVID-19 instances is reverse transcription polymerase chain reaction (RT-PCR), where respiratory samples are used to perform the test. RT-PCR can provide the result to the patients at a minimum time. However, it does not take the minimum time in most cases, and it almost takes
This paper presents a computer-aided system for speckle noise analysis in ultrasound images. The ... more This paper presents a computer-aided system for speckle noise analysis in ultrasound images. The proposed system uses the combination of convolutional neural network (CNN) features and wavelet features to detect speckle noise in ultrasound images. The wavelet features are based on the covariance of the second-order statistical measures over the wavelet transform. Evaluations on standard databases show that the proposed system is gaining an accuracy of 98.30%, sensitivity 98.79%, and specificity of 98.52%. This approach is supported by a linear discriminate analysis (LDA) for characterization of object regions from noise regions. It produces a strong speckle reduction and edge preservation due to noise-free feature extraction scheme. The experimental result is compared with several other existing speckle reduction methods and it outperforms the state-of-the-art methods on the basis of contrast resolution and MSE.
COVID-19 is one of the most life-threatening and dangerous diseases caused by the novel Coronavir... more COVID-19 is one of the most life-threatening and dangerous diseases caused by the novel Coronavirus, which has already afflicted a larger human community worldwide. This pandemic disease recovery is possible if detected in the early stage. We proposed an automated deep learning approach from Computed Tomography (CT) scan images to detect COVID-19 positive patients by following a four-phase paradigm for COVID-19 detection: preprocess the CT scan images; remove noise from test image by using anisotropic diffusion techniques; make a different segment for the preprocessed images; and train and test COVID-19 detection using Convolutional Neural Network (CNN) models. This study employed well-known pre-trained models, including AlexNet, ResNet50, VGG16 and VGG19 to evaluate experiments. 80% of images are used to train the network in the detection process, while the remaining 20% are used to test it. The result of the experiment evaluation confirmed that the VGG19 pre-trained CNN model achi...
International Journal of Wireless and Microwave Technologies
In this fast-paced technological world, individuals want to access all their electronic equipment... more In this fast-paced technological world, individuals want to access all their electronic equipment remotely, which requires devices to connect over a network via the Internet. However, it raises quite a lot of critical security concerns. This paper presented a home automation security system that employs the Internet of Things (IoT) for remote access to one's home through an Android application, as well as Artificial Intelligence (AI) to ensure the home's security. Face recognition is utilized to control door entry in a highly efficient security system. In the event of a technical failure, an additional security PIN is set up that is only accessible by the owner. Although a home automation system may be used for various tasks, the cost is prohibitive for many customers. Hence, the objective of this paper is to provide a budget and user-friendly system, ensuring access to the application and home attributes by using multi-modal security. Using Haar Cascade and LBPH the system achieved 92.86% accuracy while recognizing face.
ACM Transactions on Spatial Algorithms and Systems
The rapid spreading of coronavirus (COVID-19) caused severe respiratory infections affecting the ... more The rapid spreading of coronavirus (COVID-19) caused severe respiratory infections affecting the lungs. Automatic diagnosis helps to fight against COVID-19 in community outbreaks. Medical imaging technology can reinforce disease monitoring and detection facilities with the advancement of computer vision. Unfortunately, deep learning models are facing starvation of more generalized datasets as the data repositories of COVID-19 are not rich enough to provide significant distinct features. To address the limitation, this article describes the generation of synthetic images of COVID-19 along with other chest infections with distinct features by empirical top entropy-based patch selection approach using the generative adversarial network. After that, a diagnosis is performed through a faster region-based convolutional neural network using 6,406 synthetic as well as 3,933 original chest X-ray images of different chest infections, which also addressed the data imbalance problems and not re...
International Journal of Ambient Computing and Intelligence, 2022
The brain tumor is one of the most health hazard diseases across the world in recent time. The de... more The brain tumor is one of the most health hazard diseases across the world in recent time. The development of the intelligent system has extended its applications in the automated medical diagnosis domains. However, image-based medical diagnosis result strongly depends on the selection of relevant features. This research focuses on the automatic detection of brain tumors based on the concatenation of curvelet transform and convolutional neural network (CNN) features extracted from the preprocessed MRI sequence of the brain. Relevant features are selected from the feature vector using mutual information based on the minimum redundancy maximum relevance (mRMR) method. The detection is done using the ensemble classifier of the bagging method. The experiment is performed using two standard datasets of BraTS 2018 and BraTS 2019. After five-fold cross-validation, we have obtained an outperforming accuracy of 98.96%.
Journal of King Saud University - Computer and Information Sciences, 2020
Chest X-ray image contains sufficient information that finds wide-spread applications in diverse ... more Chest X-ray image contains sufficient information that finds wide-spread applications in diverse disease diagnosis and decision making to assist the medical experts. This paper has proposed an intelligent approach to detect Covid-19 from the chest X-ray image using the hybridization of deep convolutional neural network (CNN) and discrete wavelet transform (DWT) features. At first, the X-ray image is enhanced and segmented through preprocessing tasks, and then deep CNN and DWT features are extracted. The optimum features are extracted from these hybridized features through minimum redundancy and maximum relevance (mRMR) along with recursive feature elimination (RFE). Finally, the random forest-based bagging approach is used for doing the detection task. An extensive experiment is performed, and the results confirm that our approach gives satisfactory performance compare to the existing methods with an overall accuracy of more than 98.5%.
Machine learning and knowledge extraction, Jul 9, 2020
This research presents a machine vision approach to detect lesions in liver ultrasound as well as... more This research presents a machine vision approach to detect lesions in liver ultrasound as well as resolving some issues in ultrasound such as artifacts, speckle noise, and blurring effect. The anisotropic diffusion is modified using the edge preservation conditions which found better than traditional ones in quantitative evolution. To dig for more potential information, a learnable super-resolution (SR) is embedded into the deep CNN. The feature is fused using Gabor Wavelet Transform (GWT) and Local Binary Pattern (LBP) with a pre-trained deep CNN model. Moreover, we propose a Bayes rule-based informative patch selection approach to reduce the processing time with the selective image patches and design an algorithm to mark the lesion region from identified ultrasound image patches. To train this model, standard data ensures promising resolution. The testing phase considers generalized data with a varying resolution and test the performance of the model. Exploring cross-validation, it finds that a 5-fold strategy can successfully eradicate the overfitting problem. Experiment data are collected using 298 consecutive ultrasounds comprising 15,296 image patches. This proposed feature fusion technique confirms satisfactory performance compared to the current relevant works with an accuracy of 98.40%.
International journal of imaging and robotics, Oct 24, 2017
This paper presents an approach to speckle noise reduction for 3D ultrasound images using bi-dime... more This paper presents an approach to speckle noise reduction for 3D ultrasound images using bi-dimensional empirical mode decomposition (BEMD). 3D ultrasound is a popular diagnostic system for assessing the progression of diseases for its diverse benefits and application. Speckle noise often obscures the fine details and degrades the spatial resolution and, contrast quality that makes the interpretation of ultrasound image more difficult. The proposed method estimates an optimum threshold value of intrinsic mode functions (IMFs) using Fisher discriminant analysis (FDA) for reducing the speckles in 3D volume of ultrasound images. FDA has applied on 2D IMFs, then explored and extended to 3D. The 3D volume rendering is performed on the basis of integrating 2D slice images that provide strong speckle reduction and edge preservation. The experiment result has compared with the several other state-of-the-art threshold methods. The proposed method is also good in edge preservation and contrast resolution.
International Journal of Signal and Imaging Systems Engineering, 2018
This paper presents an approach to speckle noise reduction for 3D ultrasound images using bi-dime... more This paper presents an approach to speckle noise reduction for 3D ultrasound images using bi-dimensional empirical mode decomposition (BEMD). 3D ultrasound is a popular diagnostic system for assessing the progression of diseases for its diverse benefits and application. Speckle noise often obscures the fine details and degrades the spatial resolution and, contrast quality that makes the interpretation of ultrasound image more difficult. The proposed method estimates an optimum threshold value of intrinsic mode functions (IMFs) using Fisher discriminant analysis (FDA) for reducing the speckles in 3D volume of ultrasound images. FDA has applied on 2D IMFs, then explored and extended to 3D. The 3D volume rendering is performed on the basis of integrating 2D slice images that provide strong speckle reduction and edge preservation. The experiment result has compared with the several other state-of-the-art threshold methods. The proposed method is also good in edge preservation and contrast resolution.
International Journal of Ambient Computing and Intelligence, Mar 18, 2022
The brain tumor is one of the most health hazard diseases across the world in recent time. The de... more The brain tumor is one of the most health hazard diseases across the world in recent time. The development of the intelligent system has extended its applications in the automated medical diagnosis domains. However, image-based medical diagnosis result strongly depends on the selection of relevant features. This research focuses on the automatic detection of brain tumors based on the concatenation of curvelet transform and convolutional neural network (CNN) features extracted from the preprocessed MRI sequence of the brain. Relevant features are selected from the feature vector using mutual information based on the minimum redundancy maximum relevance (mRMR) method. The detection is done using the ensemble classifier of the bagging method. The experiment is performed using two standard datasets of BraTS 2018 and BraTS 2019. After five-fold cross-validation, we have obtained an outperforming accuracy of 98.96%.
COVID-19 is one of the most life-threatening and dangerous diseases caused by the novel Coronavir... more COVID-19 is one of the most life-threatening and dangerous diseases caused by the novel Coronavirus, which has already afflicted a larger human community worldwide. This pandemic disease recovery is possible if detected in the early stage. We proposed an automated deep learning approach from Computed Tomography (CT) scan images to detect COVID-19 positive patients by following a fourphase paradigm for COVID-19 detection: preprocess the CT scan images; remove noise from test image by using anisotropic diffusion techniques; make a different segment for the preprocessed images; and train and test COVID-19 detection using Convolutional Neural Network (CNN) models. This study employed well-known pre-trained models, including AlexNet, ResNet50, VGG16 and VGG19 to evaluate experiments. 80% of images are used to train the network in the detection process, while the remaining 20% are used to test it. The result of the experiment evaluation confirmed that the VGG19 pre-trained CNN model achieved better accuracy (98.06%). We used 4861 real-life COVID-19 CT images for experiment purposes, including 3068 positive and 1793 negative images. These images were acquired from a hospital in Sao Paulo, Brazil and two other different data sources. Our proposed method revealed very high accuracy and, therefore, can be used as an assistant to help professionals detect COVID-19 patients accurately. COVID-19 1 , also known as the Coronavirus, was spread in Wuhan, China. Within four months of its emergence in the year 2020, the World Health Organization (WHO) declared it a pandemic 2-4. Confirmed cases and deaths are recorded as 538,246,806 and 6,327,036, respectively, and these numbers are increasing day by day and until June 2022, 230 countries are currently being affected by COVID-19 5. As early COVID-19 symptoms, sometimes patients feel difficulty in breathing and do vomiting. Sneeze and cough droplets from an infected individual can easily spread from one person to another. With the large number of patients infected by COVID-19 during the pandemic, it was impossible for health experts and the competent authorities to assure enough testing kits for each. Besides, there is a shortage of kits to find out infected people 6. If more tests occur, it gets easy to find out more COVID-19 affected people and help not spread Coronavirus. One of the most used methods for diagnosing COVID-19 instances is reverse transcription polymerase chain reaction (RT-PCR), where respiratory samples are used to perform the test. RT-PCR can provide the result to the patients at a minimum time. However, it does not take the minimum time in most cases, and it almost takes
This paper presents a computer-aided system for speckle noise analysis in ultrasound images. The ... more This paper presents a computer-aided system for speckle noise analysis in ultrasound images. The proposed system uses the combination of convolutional neural network (CNN) features and wavelet features to detect speckle noise in ultrasound images. The wavelet features are based on the covariance of the second-order statistical measures over the wavelet transform. Evaluations on standard databases show that the proposed system is gaining an accuracy of 98.30%, sensitivity 98.79%, and specificity of 98.52%. This approach is supported by a linear discriminate analysis (LDA) for characterization of object regions from noise regions. It produces a strong speckle reduction and edge preservation due to noise-free feature extraction scheme. The experimental result is compared with several other existing speckle reduction methods and it outperforms the state-of-the-art methods on the basis of contrast resolution and MSE.
COVID-19 is one of the most life-threatening and dangerous diseases caused by the novel Coronavir... more COVID-19 is one of the most life-threatening and dangerous diseases caused by the novel Coronavirus, which has already afflicted a larger human community worldwide. This pandemic disease recovery is possible if detected in the early stage. We proposed an automated deep learning approach from Computed Tomography (CT) scan images to detect COVID-19 positive patients by following a four-phase paradigm for COVID-19 detection: preprocess the CT scan images; remove noise from test image by using anisotropic diffusion techniques; make a different segment for the preprocessed images; and train and test COVID-19 detection using Convolutional Neural Network (CNN) models. This study employed well-known pre-trained models, including AlexNet, ResNet50, VGG16 and VGG19 to evaluate experiments. 80% of images are used to train the network in the detection process, while the remaining 20% are used to test it. The result of the experiment evaluation confirmed that the VGG19 pre-trained CNN model achi...
International Journal of Wireless and Microwave Technologies
In this fast-paced technological world, individuals want to access all their electronic equipment... more In this fast-paced technological world, individuals want to access all their electronic equipment remotely, which requires devices to connect over a network via the Internet. However, it raises quite a lot of critical security concerns. This paper presented a home automation security system that employs the Internet of Things (IoT) for remote access to one's home through an Android application, as well as Artificial Intelligence (AI) to ensure the home's security. Face recognition is utilized to control door entry in a highly efficient security system. In the event of a technical failure, an additional security PIN is set up that is only accessible by the owner. Although a home automation system may be used for various tasks, the cost is prohibitive for many customers. Hence, the objective of this paper is to provide a budget and user-friendly system, ensuring access to the application and home attributes by using multi-modal security. Using Haar Cascade and LBPH the system achieved 92.86% accuracy while recognizing face.
ACM Transactions on Spatial Algorithms and Systems
The rapid spreading of coronavirus (COVID-19) caused severe respiratory infections affecting the ... more The rapid spreading of coronavirus (COVID-19) caused severe respiratory infections affecting the lungs. Automatic diagnosis helps to fight against COVID-19 in community outbreaks. Medical imaging technology can reinforce disease monitoring and detection facilities with the advancement of computer vision. Unfortunately, deep learning models are facing starvation of more generalized datasets as the data repositories of COVID-19 are not rich enough to provide significant distinct features. To address the limitation, this article describes the generation of synthetic images of COVID-19 along with other chest infections with distinct features by empirical top entropy-based patch selection approach using the generative adversarial network. After that, a diagnosis is performed through a faster region-based convolutional neural network using 6,406 synthetic as well as 3,933 original chest X-ray images of different chest infections, which also addressed the data imbalance problems and not re...
International Journal of Ambient Computing and Intelligence, 2022
The brain tumor is one of the most health hazard diseases across the world in recent time. The de... more The brain tumor is one of the most health hazard diseases across the world in recent time. The development of the intelligent system has extended its applications in the automated medical diagnosis domains. However, image-based medical diagnosis result strongly depends on the selection of relevant features. This research focuses on the automatic detection of brain tumors based on the concatenation of curvelet transform and convolutional neural network (CNN) features extracted from the preprocessed MRI sequence of the brain. Relevant features are selected from the feature vector using mutual information based on the minimum redundancy maximum relevance (mRMR) method. The detection is done using the ensemble classifier of the bagging method. The experiment is performed using two standard datasets of BraTS 2018 and BraTS 2019. After five-fold cross-validation, we have obtained an outperforming accuracy of 98.96%.
Journal of King Saud University - Computer and Information Sciences, 2020
Chest X-ray image contains sufficient information that finds wide-spread applications in diverse ... more Chest X-ray image contains sufficient information that finds wide-spread applications in diverse disease diagnosis and decision making to assist the medical experts. This paper has proposed an intelligent approach to detect Covid-19 from the chest X-ray image using the hybridization of deep convolutional neural network (CNN) and discrete wavelet transform (DWT) features. At first, the X-ray image is enhanced and segmented through preprocessing tasks, and then deep CNN and DWT features are extracted. The optimum features are extracted from these hybridized features through minimum redundancy and maximum relevance (mRMR) along with recursive feature elimination (RFE). Finally, the random forest-based bagging approach is used for doing the detection task. An extensive experiment is performed, and the results confirm that our approach gives satisfactory performance compare to the existing methods with an overall accuracy of more than 98.5%.
Uploads
Papers by Rafid Mostafiz