Article
Computer Science, Artificial Intelligence
Steve Tsham Mpinda Ataky, Alessandro Lameiras Koerich
Summary: Texture descriptors are widely used in medical image analysis, especially in histopathologic images (HIs), due to the variability of texture and tissue appearance caused by staining irregularities. However, extracting texture features in a discriminant way is challenging because of the non-deterministic complex system formed by the intrinsic properties of such images. This paper proposes a novel method that quantifies these properties using ecological diversity measures and discrete wavelet transform, which outperforms state-of-the-art shallow and deep methods according to experimental results on two HI datasets.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Engineering, Mechanical
Xiaoqiang Zhang, Zhiwei Liu, Xiaochang Yang
Summary: In order to improve the security of image encryption, a two-dimensional Fibonacci combined with cubic and sine map (2D-FCSM) and a pseudo-wavelet transform (PWT) are designed. The proposed image encryption algorithm using 2D-FCSM and PWT shows desirable encryption effect, strong security, and high encryption efficiency.
NONLINEAR DYNAMICS
(2023)
Article
Computer Science, Information Systems
Hiren Mewada
Summary: Cardiovascular disease is a common and threatening illness nowadays. Computer-aided diagnostic (CAD) can diagnose cardiovascular disease by identifying anomalies in an electrocardiogram (ECG). However, this traditional diagnostic approach is inefficient and requires extensive analysis and medical knowledge for accurate diagnosis. Deep learning can aid in the timely detection of anomalies. Since ECG is a multi-tone signal, CNN-based temporal features are inadequate for classification, and integrating multi-spectral information with temporal features can enhance classification accuracy. This paper proposes an automated CAD system using a 2D deep-convolution network (CNN) for ECG classification, which integrates wavelet-based spectral features with temporal features. The proposed CNN reshapes the 1D ECG into a 2D image and uses a wavelet-encoded 2D CNN to classify the images. The evaluation of the proposed model using MIT-BIH datasets demonstrates its superior performance with a highest accuracy of 99.52% and 95.64% F1 score. The study highlights that the proposed model eliminates the need for complex preprocessing steps such as signal cleaning, RR peak detection, and waveform cropping, and the integration of wavelet-based multi-spectral features improves classification accuracy.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Mathematics, Applied
Hakim Monaim, Said Fahlaoui
Summary: In this paper, a new type of orthogonal plane-split wave-packet transform that combines windowed and wavelet transforms is proposed. Fundamental properties and estimates are derived and analyzed for the given transform.
ADVANCES IN APPLIED CLIFFORD ALGEBRAS
(2023)
Article
Computer Science, Information Systems
Xiangyu Zhao, Peng Huang, Xiangbo Shu
Summary: This paper investigates the issues in feature learning methods based on CNN and proposes a new module based on wavelet attention for image classification. Experimental results demonstrate significant improvements in accuracy using this approach.
MULTIMEDIA SYSTEMS
(2022)
Article
Computer Science, Information Systems
Saroj S. Shivagunde, V. Vijaya Saradhi
Summary: With the advancement of technology, the size of high-definition images has grown exponentially, making traditional 1D methods less suitable for handling such data. To address this issue, we propose a generalized method called 2D Multi-view Discriminant Analysis (2DMvDA), which directly uses 2D image matrices for classification, resulting in a significant reduction in data size.
INFORMATION SCIENCES
(2022)
Article
Chemistry, Physical
Gyana Ranjan Sahoo, Aritro Sinha Roy, Madhur Srivastava
Summary: Two-dimensional electron spin resonance (2D ESR) spectroscopy is a unique experimental technique for studying protein structure and dynamics. This study proposes a time-frequency analysis method that resolves the limitation of traditional methods in resolving overlapping peaks in 2D ESR by decomposing the signal and identifying frequency peaks.
JOURNAL OF PHYSICAL CHEMISTRY A
(2023)
Article
Medicine, Research & Experimental
Jackson Rodrigues, Ashwini Amin, Chandavalli Ramappa Raghushaker, Subhash Chandra, Manjunath B. Joshi, Keerthana Prasad, Sharada Rai, Subramanya G. Nayak, Satadru Ray, Krishna Kishore Mahato
Summary: This study assessed tumor progression in athymic nude mice using Photoacoustic spectroscopy-based machine learning tools. Progressive tumors were classified with 99.5% accuracy using multi-class Support Vector Machine (SVM) algorithms. Serum metabolomic levels during tumor progression complemented photoacoustic spectral features, depicting breast cancer pathophysiology.
LABORATORY INVESTIGATION
(2021)
Article
Computer Science, Artificial Intelligence
Mohammed Bahoura, Hassan Ezzaidi, Jean-Francois Methot
Summary: This paper presents a new approach for near-perfect image reconstruction in online/real-time image analysis/synthesis applications. By equalizing filter delays, the peak signal-to-noise ratios of the reconstructed images have been significantly improved.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Review
Medicine, General & Internal
Manoj Diwakar, Prabhishek Singh, Vinayakumar Ravi, Ankur Maurya
Summary: Today, medical images are crucial for obtaining relevant clinical information, but their quality needs analysis and improvement. Various factors affect the quality of medical images, and multi-modality image fusion is beneficial for obtaining clinically relevant information. However, there are numerous techniques for multi-modality image fusion in the literature, each with its own assumptions, merits, and barriers. This paper critically analyzes non-conventional methods of multi-modality image fusion, providing assistance for researchers in understanding and choosing appropriate approaches.
Article
Energy & Fuels
Shida Chen, Pengcheng Liu, Dazhen Tang, Shu Tao, Taiyuan Zhang
Summary: The study utilized wavelet transform to enhance the resolution of well logging curves for better coal texture prediction. A quantitative coal texture identification model was established using linear discrimination analysis, leading to a significant improvement in recognition accuracy.
INTERNATIONAL JOURNAL OF COAL GEOLOGY
(2021)
Article
Engineering, Electrical & Electronic
Vinh Hao Le, Hui Ma, Chandima Ekanayake, Tapan Saha
Summary: This paper proposes an improved methodology for detecting and classifying multiple power quality disturbance (PQD) events, which has been evaluated and verified using synthesized signals and real-life disturbance recordings. The results show that the proposed method achieves high classification accuracy and computational efficiency.
IET GENERATION TRANSMISSION & DISTRIBUTION
(2022)
Article
Spectroscopy
Yang Bao-hua, Gao Yuan, Wang Meng-xuan, Qi Lin, Ning Jing-ming
Summary: This study proposed a method of combining spectral and spatial features to improve the accuracy of tea polyphenol estimation. By extracting wavelet coefficients and gray level co-occurrence matrix texture features, a model for estimating the content of polyphenols in yellow tea was constructed, which effectively improved the prediction accuracy.
SPECTROSCOPY AND SPECTRAL ANALYSIS
(2021)
Article
Computer Science, Information Systems
Lamiaa Abdel-Hamid
Summary: Chest computer tomography (CT) is a convenient and efficient tool for COVID-19 diagnosis. The use of wavelet and contourlet transforms allows for the analysis of features that can capture disease symptoms in chest CT images. The study showed that features extracted from the first level contourlet transform led to reliable COVID-19 classification results with reduced computational complexity.
MULTIMEDIA TOOLS AND APPLICATIONS
(2023)
Article
Computer Science, Artificial Intelligence
Tiecheng Song, Liangliang Xin, Chenqiang Gao, Tianqi Zhang, Yao Huang
Summary: This paper introduces a novel feature representation method called quaternionic extended local binary pattern (QxLBP) with adaptive structural pyramid pooling (ASPP) for color images. Through this method, it is able to handle color image information and achieve state-of-the-art performance.
PATTERN RECOGNITION
(2021)
