Enhanced Bone Cancer Diagnosis through Deep Learning on Medical Imagery

International Journal of Computational and Experimental Science and Engineering, Journal Year: 2025, Volume and Issue: 11(1)

Published: Feb. 12, 2025

Bone cancer, especially osteosarcoma, is an aggressive tumor with a highly complex histopathologic appearance that imposes considerable diagnostic difficulties. Although practical and efficient, traditional methods current deep learning models have class imbalance, fused pixel intensity distributions, tissue heterogeneity hinder efficiency. These problems emphasize the demand of more sophisticated frameworks specifically address distinct properties bone cancer histopathology images. To overcome these shortcomings, in this study proposes framework, IBCDNet, to alleviate limitations. Inspired by cutting-edge improvements architecture (e.g., like attention, residual connections, proposed Intelligent Learning-Based Cancer Detection (ILB-BCD) algorithm), framework combines different features from both public private datasets efficient way. This allows for strong feature extraction, better imbalanced data, thus precise classification. The model obtains state-of-the-art results 98.39% on Osteosarcoma Tumor Assessment Dataset, outperforming powerful baseline ResNet50, DenseNet121, InceptionV3. further affirms its robustness respective precision (97.8%), recall (98.1%), F1-score (98.0%) which shows remarkable improvement We present cost-effective scalable real-world clinical applications assist pathologists early detection accurate diagnosis cancer. Those important gaps identified addressed research contribute progress towards AI-driven healthcare global goals medicine enhanced patient outcomes.

Language: Английский

---

Advanced Ensemble Classifier Techniques for Predicting Tumor Viability in Osteosarcoma Histological Slide Images

Applied Data Science and Analysis, Journal Year: 2024, Volume and Issue: 2024, P. 52 - 68

Published: May 29, 2024

Background: Osteosarcoma is considered as the primary malignant tumor of bone, emanating from primitive mesenchymal cells that form osteoid or immature bone. Accurate diagnosis and classification play a key role in management planning to achieve improved patient outcomes. Machine learning techniques may be used augment surpass existing conventional methods towards an analysis medical data. Methods: In present study, combination feature selection was development predictive models osteosarcoma cases. The include L1 Regularization (Lasso), Recursive Feature Elimination (RFE), SelectKBest, Tree-based Importance, while following were applied: Voting Classifier, Decision Tree, Naive Bayes, Multi-Layer Perceptron, Random Forest, Logistic Regression, AdaBoost, Gradient Boosting. Some model assessment done by combining metrics such accuracy, precision, recall, F1 score, AUC, V score. Results: Tree-Based Importance for Classifier with Tree proved giving higher performance compared all other combinations, where combinations helped correct positive instances wonderful minimization false positives. Other also gave significant performances but slightly less effective, example, RFE Classifier. Conclusion: This work presents strong evidence advanced machine ensemble classifiers robust can result overall improvement diagnostic accuracy robustness osteosarcoma. Research on class imbalance computational efficiency will its future research priority.

Language: Английский

---

EffiSegNet: Gastrointestinal Polyp Segmentation through a Pre-Trained EfficientNet-based Network with a Simplified Decoder

Published: July 15, 2024

This work introduces EffiSegNet, a novel segmentation framework leveraging transfer learning with pre-trained Convolutional Neural Network (CNN) classifier as its backbone. Deviating from traditional architectures symmetric U-shape, EffiSegNet simplifies the decoder and utilizes full-scale feature fusion to minimize computational cost number of parameters. We evaluated our model on gastrointestinal polyp task using publicly available Kvasir-SEG dataset, achieving state-of-the-art results. Specifically, EffiSegNet-B4 network variant achieved an F 1 score 0.9552, mean Dice (mDice) 0.9483, Intersection over Union (mIoU) 0.9056, Precision 0.9679, Recall 0.9429 backbone - best knowledge, highest reported scores in literature for this dataset. Additional training scratch also demonstrated exceptional performance compared previous work, 0.9286, mDice 0.9207, mIoU 0.8668, 0.9311 0.9262. These results underscore importance well-designed encoder image networks effectiveness approaches.

Language: Английский

---

A Novel Hybrid Approach for Classifying Osteosarcoma Using Deep Feature Extraction and Multilayer Perceptron

Diagnostics, Journal Year: 2023, Volume and Issue: 13(12), P. 2106 - 2106

Published: June 18, 2023

Osteosarcoma is the most common type of bone cancer that tends to occur in teenagers and young adults. Due crowded context, inter-class similarity, variation, noise H&E-stained (hematoxylin eosin stain) histology tissue, pathologists frequently face difficulty osteosarcoma tumor classification. In this paper, we introduced a hybrid framework for improving efficiency three types (nontumor, necrosis, viable tumor) classification by merging different CNN-based architectures with multilayer perceptron (MLP) algorithm on WSI (whole slide images) dataset. We performed various kinds preprocessing images. Then, five pre-trained CNN models were trained multiple parameter settings extract insightful features via transfer learning, where convolution combined pooling was utilized as feature extractor. For selection, decision tree-based RFE designed recursively eliminate less significant improve model generalization performance accurate prediction. Here, tree used an estimator select features. Finally, modified MLP classifier employed classify binary multiclass under five-fold CV assess robustness our proposed model. Moreover, selection criteria analyzed optimal one based their execution time accuracy. The achieved accuracy 95.2% 99.4% Experimental findings indicate significantly outperforms existing methods; therefore, could be applicable support doctors diagnosis clinics. addition, integrated into web application using FastAPI provide real-time

Language: Английский

---

AI-Based Bone Cancer Detection Using Image Processing and CNN

Published: Jan. 1, 2025

Language: Английский

---

Comparative Analysis of Deep Neural Networks for Automated Ulcerative Colitis Severity Assessment

Bioengineering, Journal Year: 2025, Volume and Issue: 12(4), P. 413 - 413

Published: April 13, 2025

Background: Ulcerative colitis (UC) is a chronic inflammatory bowel disease characterized by continuous inflammation of the colon and rectum. Accurate assessment essential for effective treatment, with endoscopic evaluation, particularly Mayo Endoscopic Score (MES), serving as key diagnostic tool. However, MES measurement can be subjective inconsistent, leading to variability in treatment decisions. Deep learning approaches have shown promise providing more objective standardized assessments UC severity. Methods: This study utilized publicly available images patients analyze compare performance state-of-the-art deep neural networks automated classification. Several architectures were tested determine most model grading The F1 score, accuracy, recall, precision calculated all models, statistical analysis was conducted verify statistically significant differences between networks. Results: VGG19 found best-performing network, achieving QWK score 0.876 macro-averaged 0.7528 across classes. among top-performing models very small suggesting that selection should depend on specific deployment requirements. Conclusions: demonstrates multiple network could automate severity Simpler achieve competitive results larger challenging assumption necessarily provide better clinical outcomes.

Language: Английский

---

Advanced hybrid deep learning model for enhanced evaluation of osteosarcoma histopathology images

Frontiers in Medicine, Journal Year: 2025, Volume and Issue: 12

Published: April 16, 2025

Recent advances in machine learning are transforming medical image analysis, particularly cancer detection and classification. Techniques such as deep learning, especially convolutional neural networks (CNNs) vision transformers (ViTs), now enabling the precise analysis of complex histopathological images, automating detection, enhancing classification accuracy across various types. This study focuses on osteosarcoma (OS), most common bone children adolescents, which affects long bones arms legs. Early accurate OS is essential for improving patient outcomes reducing mortality. However, increasing prevalence demand personalized treatments create challenges achieving diagnoses customized therapies. We propose a novel hybrid model that combines (CNN) (ViT) to improve diagnostic using hematoxylin eosin (H&E) stained images. The CNN extracts local features, while ViT captures global patterns from These features combined classified Multi-Layer Perceptron (MLP) into four categories: non-tumor (NT), non-viable tumor (NVT), viable (VT), ratio (NVR). Using Cancer Imaging Archive (TCIA) dataset, achieved an 99.08%, precision 99.10%, recall 99.28%, F1-score 99.23%. first successful four-class this setting new benchmark research offering promising potential future advancements.

Language: Английский

---

Implementation of artificial intelligence in detection, classification, and prognostication of osteosarcoma utilizing different assessment techniques: a systematic review

Intelligence-Based Medicine, Journal Year: 2025, Volume and Issue: unknown, P. 100250 - 100250

Published: May 1, 2025

Language: Английский

---

A survey on deep learning and machine learning techniques over histopathology image based Osteosarcoma Detection

Multimedia Tools and Applications, Journal Year: 2024, Volume and Issue: unknown

Published: June 24, 2024

Language: Английский

---

A comparative assessment of machine learning models and algorithms for osteosarcoma cancer detection and classification

Healthcare Analytics, Journal Year: 2025, Volume and Issue: unknown, P. 100380 - 100380

Published: Jan. 1, 2025

Language: Английский

---