Browsing by Author "Sujin Butdisuwan"

Now showing 1 - 2 of 2
  • Default Image
    Item
    Brain Cancer Tumor Detection by U-Net Deep Learning Algorithm from MRI Images
    (Institute of Electrical and Electronics Engineers Inc., 2024) Utpal Chandra Das; Watit Benjapolakul; Manoj Gupta; Timporn Vitoonpong; Pannee Suanpang; Chanyanan Somthawinpongsai; Sujin Butdisuwan; Aziz Nanthaamornphong; U.C. Das; Center of Excellence in Artificial Intelligence, Machine Learning and Smart Grid Technology, Department of Electrical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok, 10330, Thailand; email: dasutpal999@gmail.com
    This research looks at the genomic subtypes of low-grade glioma tumors and their shape characteristics by deep learning magnetic resonance image (MRI) segmentation. We analyzed preoperative imaging and genetic data from 110 patients with low-grade glioma from the Cancer Genome Atlas. Three shape features were recovered to quantify the two- and three-dimensional aspects of the malignancies. Based on gene expression, DNA copy number, IDH mutation, 1p/19q co-deletion, DNA methylation, and microRNA, previously identified clusters were found in genomic data. We used the exact trait test to investigate the connection between chromosomal clusters and imaging traits. Our findings show a significant correlation between the margin fluctuation-bounding ellipsoid volume ratio and the RNA Seq clusters. Furthermore, a correlation was discovered between RNA-seq clusters and angular standard deviation. The U-net deep learning algorithm demonstrated a test accuracy of 94\% and a mean Dice coefficient of 90\%. These findings suggest that tumor shape characteristics derived from MRI can be projected through genomic subtypes in lower-grade gliomas. © 2024 IEEE.
  • Default Image
    Item
    SPSO-EFVM: A Particle Swarm Optimization- Based Ensemble Fusion Voting Model for Sentence-Level Sentiment Analysis
    (Institute of Electrical and Electronics Engineers Inc., 2024) Dimple Tiwari; Bharti Nagpal; Bhoopesh Singh Bhati; Manoj Gupta; Pannee Suanpang; Sujin Butdisuwan; Aziz Nanthaamornphong; M. Gupta; Guru Ghasidas Vishwavidyalaya Bilaspur, Department of Electrical Engineering, Bilaspur, Koni, Chhattisgarh, 495009, India; email: manojgupta35@yahoo.co.in; A. Nanthaamornphong; Prince of Songkla University, College of Computing, Phuket, 83120, Thailand; email: aziz.n@phuket.psu.ac.th
    Sentiment analysis has received incremental growth in recent years for emerging applications, including human-robot integration, social platforms monitoring, and decision-support systems. Several neural or transformer model-based solutions have been provided in the field of sentiment analysis that relies on the decision of a single classifier or neural model. These are erroneous to encode contextual information into appropriate dialogues and increase extra computational cost and time. Hence, we proposed a compact and parameter-effective Particle Swarm Optimization-based Ensemble Fusion Voting Model (PSO-EFVM) that exploited the combined properties of four ensemble techniques, namely Adaptive-Boost, Gradient-Boost, Random-Forest, and Extremely-Randomized Tree with Particle Swarm Optimization (PSO)-based hyperparameter selection. The proposed model is investigated on five cross-domain datasets after applying the foremost initialization and feature extraction using Information Gain (IG). It employs adaptive and gradient learning to incorporate the automatic attribute selection with the arbitrary loss function optimization. In short, a generalized two-block composite classifier is designed to perform context compositionality and sentiment classification. A population-based meta-heuristic optimization PSO is applied to each base ensemble learner that calculates weights for the best parameter selection. Comprehensive investigations of different domains reveal the superiority of the proposed PSO-EFVM over established baselines and the latest state-of-the-art models. © 2013 IEEE.