Brain Tumor Detection and Categorization with Segmentation of Improved Unsupervised Clustering Approach and Machine Learning Classifier

Bioengineering, Journal Year: 2024, Volume and Issue: 11(3), P. 266 - 266

Published: March 8, 2024

There is no doubt that brain tumors are one of the leading causes death in world. A biopsy considered most important procedure cancer diagnosis, but it comes with drawbacks, including low sensitivity, risks during treatment, and a lengthy wait for results. Early identification provides patients better prognosis reduces treatment costs. The conventional methods identifying based on medical professional skills, so there possibility human error. labor-intensive nature traditional approaches makes healthcare resources expensive. variety imaging available to detect tumors, magnetic resonance (MRI) computed tomography (CT). Medical research being advanced by computer-aided diagnostic processes enable visualization. Using clustering, automatic tumor segmentation leads accurate detection risk helps effective treatment. This study proposed Fuzzy C-Means algorithm MRI images. To reduce complexity, relevant shape, texture, color features selected. improved Extreme Learning machine classifies 98.56% accuracy, 99.14% precision, 99.25% recall. classifier consistently demonstrates higher accuracy across all classes compared existing models. Specifically, model exhibits improvements ranging from 1.21% 6.23% when other consistent enhancement emphasizes robust performance classifier, suggesting its potential more reliable classification. achieved recall rates 98.47%, 98.59%, 98.74% Fig share dataset 99.42%, 99.75%, 99.28% Kaggle dataset, respectively, which surpasses competing algorithms, particularly detecting glioma grades. shows an improvement approximately 5.39%, 6.22% Despite challenges, artifacts computational study's commitment refining technique addressing limitations positions FCM as noteworthy advancement realm precise efficient identification.

Language: Английский

---

Recent Advancements and Future Prospects in Active Deep Learning for Medical Image Segmentation and Classification

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 113623 - 113652

Published: Jan. 1, 2023

Medical images are helpful for the diagnosis, treatment, and evaluation of diseases. Precise medical image segmentation improves diagnosis decision-making, aiding intelligent services better disease management recovery. Due to unique nature images, algorithms based on deep learning face problems such as sample imbalance, edge blur, false positives, negatives. In view these problems, researchers primarily improve network structure but rarely from unstructured aspect. The paper tackles challenges, accentuating limitations convolutional neural network-based methods proposing solutions reduce annotation costs, particularly in complex introduces improvement strategies solve Additionally, article latest learning-based applications analysis, covering segmentation, acquisition, enhancement, registration, classification. Moreover, provides an overview four cutting-edge models, namely (CNN), belief (DBN), stacked autoencoder (SAE), recurrent (RNN). study selection involved searching benchmark academic databases, collecting relevant literature appropriate indicator emphasizing DL-based classification approaches, evaluating performance metrics. research highlights clinicians' scholars' obstacles developing efficient accurate malignancy prognostic framework state-of-the-art deep-learning algorithms. Furthermore, future perspectives explored overcome challenges advance field analysis.

Language: Английский

---

APESTNet with Mask R-CNN for Liver Tumor Segmentation and Classification

Cancers, Journal Year: 2023, Volume and Issue: 15(2), P. 330 - 330

Published: Jan. 4, 2023

Diagnosis and treatment of hepatocellular carcinoma or metastases rely heavily on accurate segmentation classification liver tumours. However, due to the tumor's hazy borders wide range possible shapes, sizes, positions, automatic tumour remains a difficult challenge. With advancement computing, new models in artificial intelligence have evolved. Following its success Natural language processing (NLP), transformer paradigm has been adopted by computer vision (CV) community NLP. While there are already accepted approaches classifying liver, especially clinical settings, is room for terms their precision. This paper makes an effort apply novel model segmenting tumours built deep learning. In order accomplish this, created follows three-stage procedure consisting (a) pre-processing, (b) segmentation, (c) classification. first phase, collected Computed Tomography (CT) images undergo three stages including contrast improvement via histogram equalization noise reduction median filter. Next, enhanced mask region-based convolutional neural networks (Mask R-CNN) used separate from CT abdominal image. To prevent overfitting, segmented picture fed onto Enhanced Swin Transformer Network with Adversarial Propagation (APESTNet). The experimental results prove superior performance proposed perfect variety images, as well efficiency low sensitivity noise.

Language: Английский

---

Radiophysiomics: Brain Tumors Classification by Machine Learning and Physiological MRI Data

Cancers, Journal Year: 2022, Volume and Issue: 14(10), P. 2363 - 2363

Published: May 10, 2022

The precise initial characterization of contrast-enhancing brain tumors has significant consequences for clinical outcomes. Various novel neuroimaging methods have been developed to increase the specificity conventional magnetic resonance imaging (cMRI) but also increased complexity data analysis. Artificial intelligence offers new options manage this challenge in settings. Here, we investigated whether multiclass machine learning (ML) algorithms applied a high-dimensional panel radiomic features from advanced MRI (advMRI) and physiological (phyMRI; thus, radiophysiomics) could reliably classify tumors. recently phyMRI technique enables quantitative assessment microvascular architecture, neovascularization, oxygen metabolism, tissue hypoxia. A training cohort 167 patients suffering one five most common tumor entities (glioblastoma, anaplastic glioma, meningioma, primary CNS lymphoma, or metastasis), combined with nine ML algorithms, was used develop overall 135 classifiers. Multiclass classification performance using tenfold cross-validation an independent test cohort. Adaptive boosting random forest combination advMRI were superior human reading accuracy (0.875 vs. 0.850), precision (0.862 0.798), F-score (0.774 0.740), AUROC (0.886 0.813), error (5 6). radiologists, however, showed higher sensitivity (0.767 0.750) (0.925 0.902). We demonstrated that ML-based radiophysiomics be helpful routine diagnosis tumors; high expenditure time work preprocessing requires inclusion deep neural networks.

Language: Английский

---

Efficient Liver Segmentation from Computed Tomography Images Using Deep Learning

Computational Intelligence and Neuroscience, Journal Year: 2022, Volume and Issue: 2022, P. 1 - 12

Published: May 18, 2022

Segmentation of a liver in computed tomography (CT) images is an important step toward quantitative biomarkers for computer-aided decision support system and precise medical diagnosis. To overcome the difficulties that come across segmentation are affected by fuzzy boundaries, stacked autoencoder (SAE) applied to learn most discriminative features among other tissues abdominal images. In this paper, we propose patch-based deep learning method from CT using SAE. Unlike traditional machine methods, instead anticipating pixel learning, our algorithm utilizes patches representations identify area. We preprocessed whole dataset get enhanced converted each image into many overlapping patches. These given as input SAE unsupervised feature learning. Finally, learned with labels fine tuned, classification performed develop probability map supervised way. Experimental results demonstrate proposed shows satisfactory on test Our achieved 96.47% dice similarity coefficient (DSC), which better than methods same domain.

Language: Английский

---

An Attention-Preserving Network-Based Method for Assisted Segmentation of Osteosarcoma MRI Images

Mathematics, Journal Year: 2022, Volume and Issue: 10(10), P. 1665 - 1665

Published: May 12, 2022

Osteosarcoma is a malignant bone tumor that extremely dangerous to human health. Not only does it require large amount of work, also complicated task outline the lesion area in an image manually, using traditional methods. With development computer-aided diagnostic techniques, more and researchers are focusing on automatic segmentation techniques for osteosarcoma analysis. However, existing methods ignore size osteosarcomas, making difficult identify segment smaller tumors. This very detrimental early diagnosis osteosarcoma. Therefore, this paper proposes Contextual Axial-Preserving Attention Network (CaPaN)-based MRI image-assisted method detection. Based use Res2Net, parallel decoder added aggregate high-level features which effectively combines local global In addition, channel feature pyramid (CFP) axial attention (A-RA) mechanisms used. A lightweight CFP can extract mapping contextual information different sizes. A-RA uses distinguish tissues by mining, reduces computational costs thus improves generalization performance model. We conducted experiments real dataset provided Second Xiangya Affiliated Hospital results showed our proposed achieves better than alternative models. particular, shows significant advantages with respect small target segmentation. Its precision about 2% higher average values other For objects, DSC value CaPaN 0.021 commonly used U-Net method.

Language: Английский

---

COVID-19 severity detection using chest X-ray segmentation and deep learning

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Aug. 27, 2024

COVID-19 has resulted in a significant global impact on health, the economy, education, and daily life. The disease can range from mild to severe, with individuals over 65 or those underlying medical conditions being more susceptible severe illness. Early testing isolation are vital due virus's variable incubation period. Chest radiographs (CXR) have gained importance as diagnostic tool their efficiency reduced radiation exposure compared CT scans. However, sensitivity of CXR detecting may be lower. This paper introduces deep learning framework for accurate classification severity prediction using images. U-Net is used lung segmentation, achieving precision 0.9924. Classification performed Convulation-capsule network, high true positive rates 86% COVID-19, 93% pneumonia, 85% normal cases. Severity assessment employs ResNet50, VGG-16, DenseNet201, DenseNet201 showing superior accuracy. Empirical results, validated 95% confidence intervals, confirm framework's reliability robustness. integration advanced techniques radiological imaging enhances early detection assessment, improving patient management resource allocation clinical settings.

Language: Английский

---

Grey Wolf optimized SwinUNet based transformer framework for liver segmentation from CT images

Computers & Electrical Engineering, Journal Year: 2024, Volume and Issue: 117, P. 109248 - 109248

Published: April 18, 2024

Language: Английский

---

Medical image identification methods: A review

Computers in Biology and Medicine, Journal Year: 2023, Volume and Issue: 169, P. 107777 - 107777

Published: Dec. 5, 2023

Language: Английский

---

Automated liver tissues delineation techniques: A systematic survey on machine learning current trends and future orientations

Engineering Applications of Artificial Intelligence, Journal Year: 2022, Volume and Issue: 117, P. 105532 - 105532

Published: Nov. 21, 2022

Machine learning and computer vision techniques have grown rapidly in recent years due to their automation, suitability, ability generate astounding results. Hence, this paper, we survey the key studies that are published between 2014 2022, showcasing different machine algorithms researchers used segment liver, hepatic tumors, hepatic-vasculature structures. We divide surveyed based on tissue of interest (hepatic-parenchyma, hepatic-tumors, or hepatic-vessels), highlighting tackle more than one task simultaneously. Additionally, classified as either supervised unsupervised, they further partitioned if amount work falls under a certain scheme is significant. Moreover, datasets challenges found literature websites containing masks aforementioned tissues thoroughly discussed, organizers' original contributions those other researchers. Also, metrics excessively mentioned our review, stressing relevance at hand. Finally, critical future directions emphasized for innovative tackle, exposing gaps need addressing, such scarcity many vessels' segmentation challenge why absence needs be dealt with sooner later.

Language: Английский

---