Article
Computer Science, Interdisciplinary Applications
Amrita Naik, Damodar Reddy Edla, Ramesh Dharavath
Summary: The accurate detection of malignant lung tumors on CT scans is crucial for early diagnosis of lung cancer. Combining CNN features with hand-crafted texture features like fractal and GLCM features improves detection accuracy. Replacing the softmax layer with a support vector machine classifier reduces overfitting, leading to better performance. Experiment results indicate that the combination of texture features with deep features yields higher accuracy and performance metrics.
Article
Forestry
Xi Pan, Kang Li, Zhangjing Chen, Zhong Yang
Summary: A new method utilizing near-infrared spectra and texture features has been proposed for accurate and rapid wood identification. The study shows that the identification accuracy of this method can reach 99.43%, with short-wavelength pre-processed NIR bands achieving high identification accuracy as well.
Article
Computer Science, Artificial Intelligence
Sana Ullah Khan, Naveed Islam, Zahoor Jan, Khalid Haseeb, Syed Inayat Ali Shah, Muhammad Hanif
Summary: In this article, a machine learning-based approach is proposed for automatic segmentation and classification of malignant cells in breast cytology images, achieving high accuracy of 96.3% through experiments.
NEURAL COMPUTING & APPLICATIONS
(2022)
Article
Engineering, Biomedical
Meng-Jia Lian, Chih-Ling Huang, Tzer-Min Lee
Summary: Oral cancer is a major cause of cancer-related deaths worldwide, with a low 5-year survival rate. Computer-aided detection methods, including GLCM feature extraction and SLPP reflex images, offer a promising approach for improved classification and diagnosis of oral cancer. The combined system can effectively differentiate between cancerous and normal cells, as well as different types of oral cancer.
LASERS IN MEDICAL SCIENCE
(2022)
Article
Automation & Control Systems
Opoku Eric, Rose-Mary Owusuaa Mensah Gyening, Obed Appiah, Kate Takyi, Peter Appiahene
Summary: The Cut-Test technique visually inspects the interior features of cocoa beans for classification, but it has limitations due to subjectivity and natural variations in perception. Machine Learning techniques have been proposed to address these challenges, and in this study, a novel image representation method and a simplified Neural Network model are proposed to improve classification performance. Experimental results show that the proposed model outperforms other ML models and demonstrates better generalization.
ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE
(2023)
Article
Geochemistry & Geophysics
Xiao Zhang, Xin Su, Qiangqiang Yuan, Qing Wang
Summary: The proposed STGCNet method achieves success in SAR image change detection by effectively mining spatial-temporal information and introducing 3-D-GLCM auxiliary features for speckle-robust results.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Article
Engineering, Multidisciplinary
C. Emmy Prema, S. Suresh, M. Navaneetha Krishnan, N. Leema
Summary: Smoke detection is crucial for preventing fire events. This article introduces a new smoke recognition method, which reduces false alarms by segmenting smoke-colored regions based on color and extracting distinctive texture attributes. Experimental results demonstrate that this method outperforms traditional methods in terms of detection accuracy and processing time.
Article
Clinical Neurology
Christos Moschovos, Georgios Tsivgoulis, Apostolia Ghika, Eleni Bakola, Marianna Papadopoulou, Panagiotis Zis, Vasiliki Zouvelou, Stavroula Salakou, Georgia Papagiannopoulou, Vassiliki Kotsali-Peteinelli, Elisabeth Chroni, Andreas Kyrozis
Summary: The study assessed the ability of image analysis measures to quantify echotexture changes in the median nerve for diagnostic purposes in carpal tunnel syndrome (CTS). The results showed that image analysis measures were as effective as, or even better than, subjective visual analysis. This suggests that image analysis can be a reliable and complementary tool for diagnosing CTS and may be particularly useful in older patients.
CLINICAL NEUROPHYSIOLOGY
(2023)
Article
Computer Science, Artificial Intelligence
Naveed Iqbal, Rafia Mumtaz, Uferah Shafi, Syed Mohammad Hassan Zaidi
Summary: The research uses optical images collected by drones for crop classification at different phenological stages. Results show that ML algorithms perform better on GLCM features compared to gray scale images.
PEERJ COMPUTER SCIENCE
(2021)
Article
Engineering, Electrical & Electronic
Guanglin Li, Bin Li, Shunquan Tan, Guoping Qiu
Summary: We used deep convolutional neural network architecture and co-occurrence matrix to learn deep co-occurrence features. These features represent the statistics of pixel co-occurrences, overcoming the black box nature of traditional deep representation learning and solving the computational difficulty of the matrix. We proposed a parametric co-occurrence matrix model and developed approaches to decompose the model into linear and nonlinear operations, making it easily implementable and capable of learning arbitrary shape DCOFs.
IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS FOR VIDEO TECHNOLOGY
(2023)
Article
Medicine, General & Internal
Iga Wawrzyk-Bochenek, Mansur Rahnama, Slawomir Wilczynski, Anna Wawrzyk
Summary: The study aimed to quantitatively assess the effectiveness of microneedle mesotherapy in reducing skin discoloration using the gray-level co-occurrence matrix (GLCM) method. The skin of 12 women aged 29 to 68 was examined, and microneedle mesotherapy using a preparation containing 12% ascorbic acid was performed. The effectiveness of the treatment was quantified using image analysis and processing methods.
JOURNAL OF CLINICAL MEDICINE
(2023)
Article
Biotechnology & Applied Microbiology
Igor V. Pantic, Jelena Cumic, Svetlana Valjarevic, Adeeba Shakeel, Xinyu Wang, Hema Vurivi, Sayel Daoud, Vincent Chan, Georg A. Petroianu, Meklit G. Shibru, Zehara M. Ali, Dejan Nesic, Ahmed E. Salih, Haider Butt, Peter R. Corridon
Summary: Decellularized corneas provide a promising and sustainable source for replacement grafts, reducing the risk of immune rejection. Evaluating the quality of decellularized extracellular matrix is challenging due to the lack of consensus on assessment criteria. This study developed a computational method using machine learning algorithms to assess decellularization efficiency and identified regions of interest in acellularized corneal tissue with high accuracy, providing a platform for evaluating functional changes in decellularized scaffolds.
FRONTIERS IN BIOENGINEERING AND BIOTECHNOLOGY
(2023)
Article
Multidisciplinary Sciences
Yong Liu, Qiran Li, Boxue Du, Masoud Farzaneh
Summary: This study focused on the feature extraction and classification of surface discharges of ice-covered insulator strings during the alternating current flashover process. The GLCM method was used to extract four parameters of discharge image features, revealing different stages of flashover risk.
SCIENTIFIC REPORTS
(2021)
Article
Chemistry, Multidisciplinary
Jiayan Chen, Zhenyu Zhang, Lu Wang, Qiang Wang
Summary: In this study, a method for evaluating the surface stress of metal gaskets based on gray-level co-occurrence matrix theory is proposed. The analysis is conducted using finite element calculations, and the best parameters for the gasket testing rig are obtained.
APPLIED SCIENCES-BASEL
(2022)
Article
Engineering, Biomedical
Rehan Ahmad, Basant K. Mohanty
Summary: Early detection of chronic kidney disease (CKD) stages is crucial for human healthcare, and an automatic method based on ultrasound (US) imaging can provide first-hand information on kidney health efficiently. Extracting local features from US images and selecting a classifier are key to improving the accuracy of CKD stage detection. The proposed scheme using a fourteen-point feature vector demonstrates higher accuracy in identifying CKD stages compared to existing schemes.
BIOMEDICAL SIGNAL PROCESSING AND CONTROL
(2021)