Article
Computer Science, Artificial Intelligence
Jeroen Bertels, David Robben, Dirk Vandermeulen, Paul Suetens
Summary: This study discusses the method of deriving volume estimates from uncertain or ambiguous segmentations. It identifies that soft Dice optimization may introduce volume bias in tasks with high inherent uncertainty. The study suggests a closer volume analysis and optional recalibration for better results.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Computer Science, Artificial Intelligence
Sujeet More, Jimmy Singla
Summary: The novel architecture called Discrete-MultiResUNet, combining discrete wavelet transform with MultiResUNet, is applied for knee tissue feature extraction and segmentation. This hybrid model efficiently captures significant features from knee MRI images and demonstrates better segmentation performance compared to baseline models, achieving an accuracy of 96.77% and a dice coefficient of 98%.
JOURNAL OF INTELLIGENT & FUZZY SYSTEMS
(2021)
Article
Biology
Pankaj K. Jain, Neeraj Sharma, Argiris A. Giannopoulos, Luca Saba, Andrew Nicolaides, Jasjit S. Suri
Summary: The study utilized deep learning models for automated carotid plaque segmentation, showing comparable performance and completing segmentation in less than 1 second. Data demonstrated that the SegNet-UNet hybrid architecture performed the best among the models.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Article
Remote Sensing
Hongzhang Xu, Hongjie He, Ying Zhang, Lingfei Ma, Jonathan Li
Summary: This study compares the performance differences of 12 commonly used loss functions in road segmentation tasks in remote sensing imagery. It is found that the region-based loss function type generally performs better than the distribution-based one in terms of F1, IoU, and road segmentation maps, while the compound loss function type is comparable to the region-based one. This paper aims to provide suggestions for choosing the loss function that best suits the purposes of road segmentation-related studies.
INTERNATIONAL JOURNAL OF APPLIED EARTH OBSERVATION AND GEOINFORMATION
(2023)
Article
Computer Science, Artificial Intelligence
Guoping Xu, Wentao Liao, Xuan Zhang, Chang Li, Xinwei He, Xinglong Wu
Summary: To address the issue of loss of spatial information in semantic segmentation tasks, a simple yet effective pooling operation called Haar Wavelet-based Downsampling (HWD) module is introduced. This module can easily integrate into CNNs and enhance the performance of semantic segmentation models. The core idea is to reduce the spatial resolution of feature maps while preserving as much information as possible by applying Haar wavelet transform. A novel metric called feature entropy index (FEI) is proposed to measure the degree of information uncertainty after downsampling in CNNs.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
Mengyang Zhao, Aadarsh Jha, Quan Liu, Bryan A. Millis, Anita Mahadevan-Jansen, Le Lu, Bennett A. Landman, Matthew J. Tyska, Yuankai Huo
Summary: Our study proposes a Faster Mean-shift algorithm to address the computational bottleneck of embedding based cell segmentation and tracking. Through embedding simulation and empirical validation in four ISBI cell tracking challenges, the algorithm achieved a 7-10 times speedup compared to the state-of-the-art embedding-based cell instance segmentation and tracking algorithm.
MEDICAL IMAGE ANALYSIS
(2021)
Article
Computer Science, Artificial Intelligence
Tsz Ching Ng, Siu Kai Choy, Shu Yan Lam, Kwok Wai Yu
Summary: This article presents a multi-phase image segmentation methodology based on fuzzy superpixel decomposition, aggregation, and merging. The proposed method achieves more accurate segmentation results through hierarchical aggregation of superpixels and multidimensional scaling. Comparative experiments demonstrate the superior performance of the proposed method compared to existing approaches.
PATTERN RECOGNITION
(2023)
Article
Environmental Sciences
Zhenyu Zhang, Jian Wang, Zhiyuan Li, Youlong Zhao, Ruisheng Wang, Ayman Habib
Summary: This paper proposes an optimization method for individual tree segmentation (ITS) based on Gaussian mixture model for airborne LiDAR data. It effectively addresses the accuracy issue and under-segmentation of individual trees in high-density multistoried mixed forest areas. Experimental results demonstrate that the proposed method outperforms the traditional mean shift algorithm in terms of accuracy and under-segmentation.
Article
Engineering, Multidisciplinary
Erik Cuevas, Hector Becerra, Alberto Luque, Mohamed Abd Elaziz
Summary: This paper introduces a new competitive segmentation algorithm for grayscale images, utilizing a two-dimensional feature map incorporating grayscale value and local variance, and employing a more accurate approach with the Epanechnikov kernel function to reduce computational cost. Experimental results show that the proposed method produces segmented images with approximately 50% better visual perception quality compared to its competitors, while being approximately 1.8-2 times faster.
APPLIED MATHEMATICAL MODELLING
(2021)
Article
Computer Science, Information Systems
Hanhoon Park
Summary: MeanShift++ is an improved clustering algorithm that can partition a digital image faster and with higher accuracy compared to MeanShift. It minimizes computational redundancy by introducing a hash table and a speedup factor to reduce the number of iterations required for convergence.
Article
Public, Environmental & Occupational Health
Guosheng Shen, Xiaodong Jin, Chao Sun, Qiang Li
Summary: In this study, an automatic organ segmentation technique based on deep learning convolutional neural network was developed to accurately and rapidly segment human organs for radiation therapy treatment planning.
FRONTIERS IN PUBLIC HEALTH
(2022)
Article
Mathematics, Applied
Iulia Martina Bulai, Sandra Saliani
Summary: This study extends classical wavelet, wavelet packets, and time-frequency dictionaries to the graph setting, aiming to obtain atoms that are jointly localized in both the vertex and graph spectral domain. A new method is proposed to generate a complete dictionary of wavelet packet frames defined in the graph spectral domain for representing signals on weighted graphs.
APPLIED AND COMPUTATIONAL HARMONIC ANALYSIS
(2023)
Article
Computer Science, Artificial Intelligence
Moritz Wolter, Felix Blanke, Raoul Heese, Jochen Garcke
Summary: As neural networks are able to generate realistic artificial images, it becomes crucial to develop methods to identify and analyze these images. This study proposes a novel approach based on wavelet-packet representation for synthesized fake image analysis and detection. The results show that the proposed method achieves competitive performance on small network sizes.
Article
Telecommunications
C. H. Kizil, C. Diou, C. Tanougast
Summary: This paper introduces an asynchronous Impulse Radio-Ultra Wide Band (IR-UWB) transmission solution for wireless sensor networks, utilizing the orthogonality between wavelet packets and finding the unique combination of shifted primitive packets to ensure receiver immunity to time-shift effects.
ANNALS OF TELECOMMUNICATIONS
(2021)
Article
Materials Science, Textiles
Charles Kumah, Ning Zhang, Rafiu King Raji, Zhongjian Li, Ruru Pan
Summary: This study proposes an unsupervised approach of segmenting printed fabrics using the classical mean shift algorithm by inputting enhanced ranges of parameters based on different number of clusters in the printed fabrics. The experimental results demonstrate that the proposed method is suitable for segmenting printed fabrics with less complexity in clustering.
JOURNAL OF THE TEXTILE INSTITUTE
(2022)
Review
Biochemistry & Molecular Biology
Humaira, Sayyad Ali Raza Bukhari, Hafiz Abdullah Shakir, Muhammad Khan, Shagufta Saeed, Irfan Ahmad, Muhammad Irfan
Summary: Currently, nanobiotechnology primarily focuses on the safe and eco-friendly synthesis of biocompatible metal oxide nanoparticles, with biosynthesized cerium oxide nanoparticles attracting attention in medical science due to their unique surface chemistry and dual oxidation state as excellent antioxidants and free-radical scavengers. Plant extracts are widely used for the biosynthesis of CeO(2)NPs, while other biological sources such as marine oyster shell extract, egg-white, and biopolymers have also been successfully employed. This review highlights the recent progress in biosynthesis of CeO(2)NPs and their medical use as biocompatible agents for anticancer, antibacterial, antifungal, antioxidant, antidiabetic, and wound healing purposes, as well as the prospects of developing novel products in the medical sector using biogenic CeO(2)NPs.
CURRENT PHARMACEUTICAL BIOTECHNOLOGY
(2023)
Article
Medicine, General & Internal
Kiran Iqbal Masood, Shayan Shahid, Kauser Jabeen, Joveria Farooqi, Sabeika Raza Kerawala, Muhammad Irfan
Summary: This study aimed to assess the diagnostic accuracy of different cut-off values of pleural fluid adenosine deaminase levels as a diagnostic method for tuberculous pleural effusion. The results showed that the cut-off value of 30 U/L had the highest sensitivity (71.7%) and negative predictive value (87.4%), while the cut-off value of 50 U/L had the highest specificity (89.9%), and the cut-off value of 40 U/L had the highest positive predictive value (68.9%).
JOURNAL OF THE PAKISTAN MEDICAL ASSOCIATION
(2023)
Article
Computer Science, Information Systems
Zaheer Alam, Malak Adnan Khan, Zain Ahmad Khan, Waleed Ahmad, Imran Khan, Qudrat Khan, Muhammad Irfan, Grzegorz Nowakowski
Summary: This article presents a sliding mode observer-based fault diagnosis scheme for detecting sensor faults in a standalone PV system. The proposed scheme uses residual formation for fault detection based on a defined threshold. Additionally, the sliding mode observer reduces the number of sensors required in the PV system, serving as software redundancy in fault-tolerant control.
Article
Biochemistry & Molecular Biology
Iqra Jawad, Husam Bin Tawseen, Muhammad Irfan, Waqar Ahmad, Mujtaba Hassan, Fazal Sattar, Fazli Rabbi Awan, Shazia Khaliq, Nasrin Akhtar, Kalsoom Akhtar, Munir Ahmad Anwar, Nayla Munawar
Summary: Microbial exopolysaccharides (EPSs) have been found to improve biochemical parameters and modulate microbiomics and metabolomics in mice models. Inulin and dextran-type EPSs showed hypocholesterolemic effects and controlled weight gain, while also increasing the population of beneficial bacteria and inhibiting the growth of enteropathogens.
INTERNATIONAL JOURNAL OF MOLECULAR SCIENCES
(2023)
Article
Biochemical Research Methods
Ozana Almeida Lessa, Fabiane Neves Silva, Iasnaia Maria de Carvalho Tavares, Igor Carvalho Fontes Sampaio, Adriana Bispo Pimentel, Selma Gomes Ferreira Leite, Melissa Limoeiro Estrada Gutarra, Lucas Galhardo Pimenta Tienne, Muhammad Irfan, Muhammad Bilal, Paulo Neilson Marques dos Anjos, Luiz Carlos Salay, Marcelo Franco
Summary: This study reveals the structural changes caused by Penicillium roquefort in the lignocellulosic matrix, providing important information for understanding the changes after fungal growth. The results are crucial for proposing the total use of residual solid after fermentation and reducing the lack of information in the literature.
PREPARATIVE BIOCHEMISTRY & BIOTECHNOLOGY
(2023)
Review
Engineering, Chemical
Mudasar Zafar, Hamzah Sakidin, Mikhail Sheremet, Iskandar B. B. Dzulkarnain, Abida Hussain, Roslinda Nazar, Javed Akbar Khan, Muhammad Irfan, Zafar Said, Farkhanda Afzal, Abdullah Al-Yaari
Summary: The rapid changes in nanotechnology have led to significant advancements in the study of nanofluids, which are colloid suspensions of nanoparticles in base fluids known for their remarkable heat transfer ability. Mathematical models and varied geometries have been used to enhance heat transfer rates in nanofluids. This article provides an overview of the Tiwari and Das nanofluid models, studying the effects of different geometries, nanoparticles, and their physical properties on heat transfer.
Article
Environmental Studies
Shiyong Zheng, Muhammad Irfan, Fengyi Ai, Mamdouh Abdulaziz Saleh Al-Faryan
Summary: Recently, carbon emissions and ecological footprint have been criticized for being used as substitutes for assessing pollution in some empirical studies. This study differs from previous research by utilizing load capacity factor (LF) as a measure of ecological quality. LF takes into account both the demand and supply sides of nature, providing a comprehensive evaluation of ecological quality through the analysis of ecological footprint and biocapacity.
Article
Medicine, General & Internal
Maham Vaqar, Ayesha Sharif, Nousheen Iqbal, Muhammad Irfan
Summary: Bilateral spontaneous pneumothorax can be the initial manifestation of aggressive cutaneous angiosarcoma and frequently leads to respiratory failure. Early recognition is essential to prevent delay in diagnosis and management.
JOURNAL OF MEDICAL CASE REPORTS
(2023)
Article
Materials Science, Multidisciplinary
Hafiza Mehwish Rasheed, Abdul Rauf, Muhammad Arif, Ayesha Mohyuddin, Muhammad Javid, Sohail Nadeem, Afifa Yousuf, Muhammad Irfan, Shah Muhammad Haroon, Hamid Raza, Sobhy M. Ibrahim, Shams Ul Mahmood
Summary: Water contamination is an increasing problem due to the versatile applications of heavy metal ions, but the selective removal of these ions from wastewater remains an issue. To address this, a ZnO-PVA-coated membrane was developed using a sol-gel process and bio-inspired Azadirachta indica (neem) extract. The membrane successfully removed Pb+2 ions from polluted water solution, with a removal efficiency of 93.99% at pH=4 and a maximum adsorption capacity of 218.47 mg/g.
Article
Environmental Sciences
Naila Nureen, Da Liu, Muhammad Irfan, Robert Sroufe
Summary: The spontaneous organizational citizenship behavior towards environment (OCBE) of employees in the workplace is crucial to businesses' green development and low-carbon transition. This study investigates the relationship between green supply chain management, green culture, top management commitment, OCBE, and firm performance. The results show that green supply chain management indirectly affects firm performance through green culture and top management commitment. The study also finds that OCBE has a direct effect on firm performance and a positive moderating effect on the association between green supply chain management and firm performance.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2023)
Article
Environmental Sciences
Naila Nureen, Huaping Sun, Muhammad Irfan, Alina Cristina Nuta, Maida Malik
Summary: The increasing importance of green supply chain management in developing countries' manufacturing sector is driven by environmental deterioration and the need for competitiveness and sustainability. This research examines the impact of green supply chain management practices on firm performance, considering the moderating effect of collaborative capability and the mediating influence of eco-technological innovation and environmental strategy. The findings indicate that green supply chain management indirectly affects firm performance, with positive relationships observed with environmental strategy and eco-technological innovation. The study also highlights the significant mediating role of environmental strategy and eco-technological innovation, as well as the positive moderating effect of collaborative capability on the relationship between green supply chain management and firm performance.
ENVIRONMENTAL SCIENCE AND POLLUTION RESEARCH
(2023)
Article
Computer Science, Artificial Intelligence
Adam Glowacz
Summary: This paper presents a technique for analyzing thermal images of minigrinders. Different states of minigrinders were analyzed and two methods for computing essential areas of thermal images were proposed. The analysis was verified and found to be effective for fault diagnosis with 100% recognition efficiency.
EXPERT SYSTEMS WITH APPLICATIONS
(2024)
Article
Chemistry, Multidisciplinary
Mubbashar Nazeer, Muhammad Irfan, Farooq Hussain, Imran Siddique
Summary: This article investigates the application of gold nanoparticles in cancer treatment. The study finds that the small size and high atomic number of gold nanoparticles can generate heat to destroy tumors. Through mathematical modeling and graphical analysis, the study determines the impact of associated parameters on treatment outcomes, and finds that gold nanoparticles can effectively enhance temperature distribution and destroy cancer cells.
JOURNAL OF COMPUTATIONAL BIOPHYSICS AND CHEMISTRY
(2023)
Article
Energy & Fuels
Shagufta Saeed, Tahir Mehmood, Muhammad Irfan
Summary: This study aimed to produce alginate through solid-state fermentation using apple peels as a substrate. By optimizing the cultural parameters, a high yield of alginate was achieved. The results showed that the yield of alginate reached 180.64 mg/gds with a purity of 97%. This indicates that cheap fruit by-products can be efficiently utilized for alginate production.
BIOMASS CONVERSION AND BIOREFINERY
(2023)