https://journal.ijisnt.com/index.php/public_html/issue/feed 2025-03-15T10:02:33+00:00 Prof. Dr. Muhibul Haque Bhuyan muhibulhb@aiub.edu Open Journal Systems <div class="flex flex-grow flex-col gap-3 max-w-full"> <div class="min-h-[20px] flex flex-col items-start gap-3 whitespace-pre-wrap break-words overflow-x-auto" data-message-author-role="assistant" data-message-id="c2e15a35-7b5f-45ce-9d9e-61b4b0ac23e4"> <div class="markdown prose w-full break-words dark:prose-invert light"> <p>Welcome to the <strong>International Journal of Imminent Science &amp; Technology</strong>, where cutting-edge research meets rigorous peer review. Our journal stands proudly at the forefront of scientific innovation, serving as a dedicated platform for scholars, researchers, and scientists from around the globe to publish their groundbreaking work.</p> <p>With our meticulous peer review system, we uphold the highest standards of quality and integrity in every published paper, ensuring that our readers receive the most reliable and cutting-edge research findings. At Imminent Science &amp; Technology, we cultivate a collaborative environment that nurtures the exchange of ideas and knowledge, propelling advancements in various fields of science and technology.</p> </div> </div> </div> https://journal.ijisnt.com/index.php/public_html/article/view/30 2025-03-15T10:02:33+00:00 Roni Acharjee roniach019@gmail.com Abu Sayed Sikder PM21496@student.uniten.edu.my Hridoy paul Gupi hpg.828@gmail.com Sayeda Samina Hussain syedasaminahussain552@gmail.com

<p>Accurate classification of blood cancer subtypes, such as Acute Myeloid Leukemia (AML) and Acute Lymphoblastic Leukemia (ALL), is crucial for personalized treatment strategies. This study employs a quantitative methodology to classify blood cancer subtypes using gene expression data from 72 patients with 7,129 distinct gene expressions. Advanced preprocessing techniques, including Principal Component Analysis (PCA) and Synthetic Minority Oversampling Technique (SMOTE), were applied to handle high dimensionality and class imbalance. The dataset was split into 80% training and 20% testing sets. We evaluated ML algorithms such as Support Vector Machine (SVM), Logistic Regression, Random Forest, and K-Nearest Neighbor (KNN), alongside DL architectures like Convolutional Neural Networks (CNNs) and a hybrid CNN-LSTM model. Performance was assessed using metrics like accuracy, precision, recall, F1-score, and AUC-ROC. SVM and Logistic Regression achieved 100% accuracy, while the CNN-LSTM model achieved 99.1% accuracy, demonstrating superior performance in capturing complex gene expression patterns.</p> <p>External validation on The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) datasets confirmed the models' robustness, with slight performance drops due to dataset variability. Biological interpretation using Gene Ontology (GO) enrichment analysis identified known biomarkers (e.g., FLT3 for AML and PAX5 for ALL) and potential novel biomarkers (e.g., GATA2 and RUNX1). A comparative analysis with state-of-the-art methods, including SVM with Recursive Feature Elimination (RFE) and XGBoost, showed that the proposed models consistently outperformed existing techniques. This study highlights the potential of ML and DL in blood cancer classification, offering a foundation for automated diagnostic systems that enhance clinical decision-making and personalized treatment strategies. The findings contribute to advancing personalized medicine and improving patient outcomes.</p>

2025-03-15T00:00:00+00:00 Copyright (c) 2025 International Journal of Imminent Science & Technology.