Article
Computer Science, Artificial Intelligence
Muhammad Irfan Sharif, Jian Ping Li, Muhammad Attique Khan, Seifedine Kadry, Usman Tariq
Summary: Brain tumor is an active research topic in medical imaging due to its low survival rate. This study proposes an optimized deep learning system for multimodal brain tumor classification, achieving high accuracy and showing superiority over existing techniques.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Computer Science, Information Systems
Muhammad Attique Khan, Reham R. Mostafa, Yu-Dong Zhang, Jamel Baili, Majed Alhaisoni, Usman Tariq, Junaid Ali Khan, Ye Jin Kim, Jaehyuk Cha
Summary: This paper proposes a novel multimodal brain tumor classification framework based on two-way deep learning feature extraction and a hybrid feature optimization algorithm. The proposed method achieved high accuracy rates on two publicly available datasets and outperformed several recent studies.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Information Systems
Alireza Samavat, Ebrahim Khalili, Bentolhoda Ayati, Marzieh Ayati
Summary: This paper proposes a hybrid multi-input deep model using convolution neural networks (CNNs) and bidirectional Long Short-term Memory (Bi-LSTM) to recognize emotions from raw EEG signals. The method extracts temporal and frequency features and considers the spatial information of EEG, achieving significant improvement in accuracy compared to the baseline.
Article
Medicine, General & Internal
Venkatesan Rajinikanth, P. M. Durai Raj Vincent, C. N. Gnanaprakasam, Kathiravan Srinivasan, Chuan-Yu Chang
Summary: This research aims to develop an efficient deep-learning-based brain-tumor detection scheme using FLAIR- and T2-modality MRI slices. The scheme includes preprocessing, deep-feature extraction, tumor segmentation, feature optimization, and binary classification. Experimental results show that the integrated feature-based scheme achieves a classification accuracy of 99.6667% when using a support-vector-machine classifier.
Article
Computer Science, Information Systems
Dongdong Li, Yijun Zhou, Zhe Wang, Daqi Gao
Summary: Recent studies on speech signals have focused on emotional information and the importance of feature representation in speech emotion recognition (SER). Different combinations of features and models have a significant impact on SER performance, with the proposed ECFW method showing promising results in improving performance across different databases.
INFORMATION SCIENCES
(2021)
Article
Radiology, Nuclear Medicine & Medical Imaging
Yuen Teng, Chaoyue Chen, Xin Shu, Fumin Zhao, Lei Zhang, Jianguo Xu
Summary: This study aimed to optimize brain extraction models for oncological analysis and developed an nnU-Net-based deep learning model for automated brain extraction in the presence of brain tumors.
EUROPEAN RADIOLOGY
(2023)
Article
Engineering, Electrical & Electronic
Ruohong Huan, Chengxi Jiang, Luoqi Ge, Jia Shu, Ziwei Zhan, Peng Chen, Kaikai Chi, Ronghua Liang
Summary: In this paper, the optimal feature representation of human complex activities is studied, and a method for extracting multi-layer features from a hybrid CNN and BLSTM network is proposed. A new feature selection method is also introduced to generate mixed features by fusing different sources. Experimental results demonstrate that this method outperforms existing approaches in human complex activity recognition.
IEEE SENSORS JOURNAL
(2022)
Article
Computer Science, Information Systems
Petra Takacs, Levente Kovacs, Andrea Manno-Kovacs
Summary: This study introduces an improved brain tumor segmentation method utilizing visual saliency features on MRI image volumes. The novel approach combines deep learning techniques with handcrafted feature models, demonstrating enhanced segmentation performance through the use of healthy templates in the training process and fusion of saliency maps with convolutional neural networks' prediction maps.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Multidisciplinary Sciences
Cristiana Fiscone, Leonardo Rundo, Alessandra Lugaresi, David Neil Manners, Kieren Allinson, Elisa Baldin, Gianfranco Vornetti, Raffaele Lodi, Caterina Tonon, Claudia Testa, Mauro Castelli, Fulvio Zaccagna
Summary: This study explored the robustness of QSM radiomic features by varying the number of grey levels and echo times and validated the reliability of QSM-based radiomics features against these parameters. The results provide important insights for future radiomics studies using QSM in clinical applications.
SCIENTIFIC REPORTS
(2023)
Article
Computer Science, Artificial Intelligence
Yuanlun Xie, Wenhong Tian, Hengxin Zhang, Tingsong Ma
Summary: Recent studies have shown that deep learning has great potential in facial expression recognition tasks and has attracted more and more attention from researchers. Existing methods have achieved good results in laboratory settings, but face challenges when applied in wild environments with more complex and diverse facial expression images. This paper proposes a new method for facial expression recognition by extracting and fusing multi-level features.
Article
Engineering, Electrical & Electronic
Muhammad Attique Khan, Awais Khan, Majed Alhaisoni, Abdullah Alqahtani, Shtwai Alsubai, Meshal Alharbi, Nazir Ahmed Malik, Robertas Damasevicius
Summary: In the last decade, there has been a significant increase in medical cases involving brain tumors. This paper proposes an automated system for brain tumor detection and classification using a saliency map and deep learning feature optimization. The framework consists of several stages, including contrast enhancement, tumor segmentation, deep transfer learning, feature fusion, and classification. The experimental results show improved accuracy compared to other neural nets.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Article
Neurosciences
Leonard Elia van Dyck, Roland Kwitt, Sebastian Jochen Denzler, Walter Roland Gruber
Summary: Deep convolutional neural networks (DCNNs) and the ventral visual pathway exhibit vast architectural and functional similarities in object recognition tasks. However, differences in spatial priorities of information processing are not fully taken into account in comparisons between the two systems. This study compares human observers and three forward DCNNs, revealing fundamentally different resolutions in terms of behavior and activation. It also provides evidence that a DCNN with biologically plausible receptive field sizes shows higher agreement with human viewing behavior.
FRONTIERS IN NEUROSCIENCE
(2021)
Article
Neurosciences
Andrew Hoopes, Jocelyn S. Mora, Adrian V. Dalca, Bruce Fischl, Malte Hoffmann
Summary: Skull-stripping, the removal of non-brain signal from MRI data, is an integral part of neuroimage analysis. Existing methods are limited in their applicability to specific image types, but our learning-based tool, SynthStrip, can successfully generalize to a variety of real acquired brain images, regardless of their acquisition properties. We demonstrate its efficacy across diverse populations and show substantial improvements in accuracy compared to popular baselines.
Article
Engineering, Biomedical
Aparajita Nanda, Ram Chandra Barik, Sambit Bakshi
Summary: Early-stage diagnosis is crucial for curing brain tumors, and existing classification schemes using machine learning and deep learning have limitations. This paper proposes a new hybrid approach that combines saliency-K-means segmentation with social spider optimization algorithm in radial basis neural network for efficient classification. The proposed method is validated on standard datasets and achieves high accuracy.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2023)
Article
Computer Science, Interdisciplinary Applications
Erdal Ozbay, Feyza Altunbey Ozbay
Summary: This study proposes a precision medical image hashing method that addresses the issue of medical image retrieval by combining MRI images with feature fusion. Experimental results showed that the proposed method can effectively identify tumor regions and generate more accurate hash codes using three loss functions in feature fusion. It has been demonstrated that our method can effectively increase the accuracy of medical image retrieval and potentially be applied to computer-aided diagnosis systems(CADs).
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
(2023)
Article
Engineering, Electrical & Electronic
Muhammad Attique Khan, Awais Khan, Majed Alhaisoni, Abdullah Alqahtani, Shtwai Alsubai, Meshal Alharbi, Nazir Ahmed Malik, Robertas Damasevicius
Summary: In the last decade, there has been a significant increase in medical cases involving brain tumors. This paper proposes an automated system for brain tumor detection and classification using a saliency map and deep learning feature optimization. The framework consists of several stages, including contrast enhancement, tumor segmentation, deep transfer learning, feature fusion, and classification. The experimental results show improved accuracy compared to other neural nets.
INTERNATIONAL JOURNAL OF IMAGING SYSTEMS AND TECHNOLOGY
(2023)
Article
Medicine, General & Internal
Ameer Hamza, Muhammad Attique Khan, Majed Alhaisoni, Abdullah Al Hejaili, Khalid Adel Shaban, Shtwai Alsubai, Areej Alasiry, Mehrez Marzougui
Summary: In this study, we proposed a novel framework for COVID-19 classification using deep Bayesian optimization and improved canonical correlation analysis (ICCA). The framework utilized data augmentation and transfer learning to train two pre-trained deep models, and fused their features using ICCA. The fused features were further optimized and classified using an optimization algorithm and a neural network classifier, achieving very high accuracy.
Article
Chemistry, Analytical
Muhammad Asim Saleem, Ngoc Thien Le, Widhyakorn Asdornwised, Surachai Chaitusaney, Ashir Javeed, Watit Benjapolakul
Summary: Lung cancer is a leading cause of cancer-related deaths in today's world. Screening and early detection of lung nodules are crucial for improving patient outcomes. A new deep learning model, based on the sooty tern optimization algorithm, has been proposed to enhance the accuracy of diagnosing non-small cell lung cancer tumors.
Article
Chemistry, Analytical
Ahmad Almadhor, Gabriel Avelino Sampedro, Mideth Abisado, Sidra Abbas, Ye-Jin Kim, Muhammad Attique Khan, Jamel Baili, Jae-Hyuk Cha
Summary: With recent advancements in wearable technology, continuous stress monitoring through physiological factors has gained attention. Identifying stress early can improve healthcare outcomes by reducing the negative effects. However, privacy concerns limit the availability of data, making it challenging to utilize AI models in the medical industry. This research proposes a Federated Learning approach that utilizes a Deep Neural Network model to classify wearable-based electrodermal activities while ensuring patient data privacy.
Article
Green & Sustainable Science & Technology
Waqas Ahmad, Hikmat Ullah Khan, Tasswar Iqbal, Muhammad Attique Khan, Usman Tariq, Jae-hyuk Cha
Summary: With the rapid growth of user-generated content on social media, sentiment analysis has become a significant research area. In this manuscript, a technique combining variant algorithms in a parallel manner is proposed to extract advantageous informative features, and then perform sentiment classification. The proposed methodology, a combination of MC-CNN and MC-Bi-GRU, treats them equally in terms of input parameters and shares hidden layer information, making it distinctive and outperforming existing models.
Article
Chemistry, Multidisciplinary
Ashir Javeed, Muhammad Asim Saleem, Ana Luiza Dallora, Liaqat Ali, Johan Sanmartin Berglund, Peter Anderberg
Summary: Researchers developed a clinical decision support system using machine learning and data mining techniques to predict mortality in cardiac patients. They used synthetic minority oversampling technique (SMOTE) to address the problem of imbalanced dataset and optimized random forest model for classification. Experimental results showed that the proposed system outperformed other machine learning models in terms of accuracy.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Information Systems
Shah Faisal, Kashif Javed, Sara Ali, Areej Alasiry, Mehrez Marzougui, Muhammad Attique Khan, Jae-Hyuk Cha
Summary: This paper presents an automatic system for early detection and classification of citrus plant diseases based on a deep learning model. By using the latest transfer learning models, the classification accuracy is improved. The proposed CNN model outperforms other previous models in identifying and categorizing citrus plant diseases.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Information Systems
Wajiha Rahim Khan, Tahir Mustafa Madni, Uzair Iqbal Janjua, Umer Javed, Muhammad Attique Khan, Majed Alhaisoni, Usman Tariq, Jae-Hyuk Cha
Summary: This research proposes a new 3D deep learning model for brain tumor segmentation. The model utilizes lightweight feature extraction modules to improve performance without compromising contextual information or accuracy. Experimental results show that the proposed model outperforms other state-of-the-art models while having a smaller number of parameters, and can improve brain tumor segmentation accuracy, facilitating appropriate diagnosis and treatment planning for medical practitioners.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Information Systems
Qurat-ul-Ain Arshad, Mudassar Raza, Wazir Zada Khan, Ayesha Siddiqa, Abdul Muiz, Muhammad Attique Khan, Usman Tariq, Taerang Kim, Jae-Hyuk Cha
Summary: Video anomaly detection is used to identify anomalous situations in surveillance videos or images that can result in security issues. Differentiating abnormal situations from normal ones can be challenging due to variations in human activity in complex environments. This work proposes a deep learning architecture called ASRNet for deep feature extraction to improve the accuracy of detecting various anomalous image situations.
CMC-COMPUTERS MATERIALS & CONTINUA
(2023)
Article
Computer Science, Hardware & Architecture
Ayesha Kanwal, Kashif Javed, Sara Ali, Saddaf Rubab, Muhammad Attique Khan, Areej Alasiry, Mehrez Marzougui, Mohammad Shabaz
Summary: This paper presents a novel framework for early detection of autism using a semi-supervised approach that includes the ConvNeXt-T model and clustering of image embedding vectors. The framework achieves high accuracy in autism classification and can work on unlabelled datasets.
JOURNAL OF SUPERCOMPUTING
(2023)
Article
Mathematical & Computational Biology
Venkatesan Rajinikanth, Seifedine Kadry, Ramya Mohan, Arunmozhi Rama, Muhammad Attique Khan, Jungeun Kim
Summary: Cancer occurrence rates are rising, especially colorectal cancer. Early screening and treatment are crucial. This research develops a deep-learning framework to classify colon histology slides with high accuracy.
MATHEMATICAL BIOSCIENCES AND ENGINEERING
(2023)
Article
Engineering, Electrical & Electronic
Ameer Hamza, Muhammad Attique Khan, Shams Ur Rehman, Hussain Mobarak Albarakati, Roobaea Alroobaea, Abdullah M. Baqasah, Majed Alhaisoni, Anum Masood
Summary: The paper introduces a new automated technique based on the inner fusion of two deep learning models and feature selection, with the initialization of hyperparameters based on Bayesian optimization and the development of a controlled entropy feature selection technique. Experimental results on three publicly available datasets show improved accuracy compared to state-of-the-art techniques.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2023)
Article
Multidisciplinary Sciences
Muhammad Attique Khan, Yu-Dong Zhang, Majed Alhusseni, Seifedine Kadry, Shui-Hua Wang, Tanzila Saba, Tassawar Iqbal
Summary: In this paper, a method for action recognition based on the fusion of shape and deep learning features is proposed. The method consists of two steps: human extraction and action recognition. By combining entropy-controlled feature selection and parallel conditional entropy approach, the features are fused and classified, achieving a high accuracy rate.
ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Muhammad Attique Khan, Seifedine Kadry, Pritee Parwekar, Robertas Damasevicius, Asif Mehmood, Junaid Ali Khan, Syed Rameez Naqvi
Summary: Human gait analysis is an important topic in computer vision with various applications. However, the variability in patients' clothes, viewing angle, and carrying conditions can affect system performance. To enhance accuracy, this study proposes a deep learning feature aggregation method applied in gait recognition. The results demonstrate that the proposed method achieves accuracy beyond 96% and outperforms other classifiers.
COMPLEX & INTELLIGENT SYSTEMS
(2023)
Article
Computer Science, Interdisciplinary Applications
Zainab Akhtar, Jong Weon Lee, Muhammad Attique Khan, Muhammad Sharif, Sajid Ali Khan, Naveed Riaz
Summary: This paper presents an automated OCR technique based on multi-properties features fusion and selection. The features are fused using serial formulation and selected using partial least square method with an entropy fitness function. The method achieves high accuracy in testing and outperforms existing techniques.
JOURNAL OF ENTERPRISE INFORMATION MANAGEMENT
(2023)