Blood cancer: Advances in diagnostic research

Elsevier eBooks, Journal Year: 2025, Volume and Issue: unknown, P. 291 - 320

Published: Jan. 1, 2025

Language: Английский

---

VisTA: vision transformer-attention enhanced CNN ensemble for optimized classification of acute lymphoblastic leukemia benign and progressive malignant stages

International Journal of Information Technology, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 5, 2024

Language: Английский

---

Developing an efficient VGG19-based model and transfer learning for detecting acute lymphoblastic leukemia (ALL)

Published: June 8, 2023

Acute lymphoblastic leukemia (ALL) is a form of blood cancer that affects the lymphoid cells, leading to excessive proliferation immature lymphocytes. A pathologist typically examines bone marrow recognize specific type cells present. However, This time-honoured approach takes lot effort and time may not always yield accurate results due variations in specialist expertise. As result, there need for automated methods can increase efficiency accuracy identifying cells. Deep learning techniques have shown promise this regard, as they analyze images make predictions about their type. In our study, we utilized VGG19 convolutional neural network (CNN) model from ALL-IDB-1 dataset ALL. Our demonstrate remarkable rate 99.49%, indicating proposed outperformed other tested models simplicity performance. These findings suggest machine deep offer an effective way streamline identification improve patient outcome.

Language: Английский

---

Detection of chronic lymphocytic leukemia using Deep Neural Eagle Perch Fuzzy Segmentation – A novel comparative approach

Biomedical Signal Processing and Control, Journal Year: 2023, Volume and Issue: 90, P. 105905 - 105905

Published: Dec. 30, 2023

Language: Английский

---

Multihead Neural Network for Multiple Segmented Images-Based Diagnosis of Thyroid-Associated Orbitopathy Activity

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 43862 - 43873

Published: Jan. 1, 2024

Thyroid-associated orbitopathy is an autoimmune disease that causes changes in various structures close to the eye. Medical images, such as three-dimensional computed tomography scans, can be used by medical experts diagnose thyroid-associated orbitopathy. Meanwhile, image segmentation has been widely imaging owing its significant impact on improving model performance filtering out unnecessary pixel values. In this study, a neural network specialized processing multiple segmented images was proposed evaluate thyroid activity, focusing fact extracted from orbital scans. The consists of convolutional embedding heads, group squeeze-and-excitation block, and classifier stage. Our empirical study shows outperforms four baseline models activity dataset obtained cohort 1,068 patients at Chung-Ang University Hospital between January 2008 October 2019. achieved average area under receiver operating characteristic curve 0.800, accuracy 0.721, F1 score 0.416, sensitivity 0.728, specificity 0.720 across 50 replicate experiments. source code for available https://github.com/tkdgur658/MultiheadGroupSENet.

Language: Английский

---

Design of an Efficient Prediction Model for Early Parkinson’s Disease Diagnosis

IEEE Access, Journal Year: 2024, Volume and Issue: 12, P. 137295 - 137309

Published: Jan. 1, 2024

Parkinson's Disease (PD) is a long-lasting and progressive brain disorder that disrupts the body's nervous system pathways. This disruption leads to various issues with movement control, leading symptoms, including tremors, stiffness, difficulty coordination. In early stages of this condition, patients struggle speak also slowly. Dysphonia, speech impairment or alteration in speech, experienced by 70 90 percent an indication disease. Hence, voice can be vital modality for stage PD diagnosis. literature, Machine Learning models are implemented diagnosis based on data. However, like class imbalance, feature selection, interpretable prediction analysis not addressed effectively. Moreover, accurate trustworthiness results essential providing better healthcare services. Here, we propose enhanced Interpretable Feature Ranking XGBoost (IFRX) model predict early-stage The proposed addresses above-mentioned effectively provides performance. Using model, eight classifiers Among these classifiers, approach shows performance accuracy 96.61%.

Language: Английский

---

The evaluation of denoising techniques on microscopic blood smear images of Acute Lymphoblastic Leukemia (ALL)

Open Access Research Journal of Engineering and Technology, Journal Year: 2023, Volume and Issue: 4(2), P. 024 - 035

Published: June 6, 2023

Preprocessing is the first step in image processing for any digital before it goes further step. Denoising techniques are one of important used preprocessing. The digitized microscopic blood smear contains unwanted noise due to poor illumination, electronic interference, different variation lighting condition etc. these images without filtering techniques, can produce inaccurate results such as segmentation, feature extraction and classification. So, necessary preprocess with proper each type In this paper, we have tried evaluate types on Acute Lymphoblastic Leukemia (ALL) removal noise. We reviewed many research work which various

Language: Английский

---

Morphological diagnosis of hematologic malignancy using feature fusion-based deep convolutional neural network

Scientific Reports, Journal Year: 2023, Volume and Issue: 13(1)

Published: Oct. 9, 2023

Leukemia is a cancer of white blood cells characterized by immature lymphocytes. Due to cancer, many people die every year. Hence, the early detection these blast necessary for avoiding cancer. A novel deep convolutional neural network (CNN) 3SNet that has depth-wise convolution blocks reduce computation costs been developed aid diagnosis leukemia cells. The proposed method includes three inputs CNN model. These are grayscale and their corresponding histogram gradient (HOG) local binary pattern (LBP) images. HOG image finds shape, LBP describes leukaemia cell's texture pattern. suggested model was trained tested with images from AML-Cytomorphology_LMU dataset. mean average precision (MAP) cell less than 100 in dataset 84%, whereas more 93.83%. In addition, ROC curve area 98%. This confirmed could be an adjunct tool provide second opinion doctor.

Language: Английский

---

A Novel Approach for Leukemia Classification using Multi-Neural Networks

Published: April 17, 2024

Leukemia is a type of cancer that originates in the bone marrow and affects blood-forming cells. These abnormal cells, typically white blood multiply uncontrollably, hindering production normal Because its various genetic molecular properties, leukemia, complex heterogeneous group malignancies, provides major hurdles correct subtype categorization. Traditional categorization approaches often fail to reflect complexities leukemia subgroups. In this paper, research offer Multi-Neural Network (MNN), ground-breaking strategy for addressing these difficulties by exploiting hierarchical information merging specialized neural networks. The dataset was collected from Kaggle repository. After collecting use Non adaptive threshold Image denoising. denoising, Adam optimization algorithm process. HOG feature selection. proposed MNN architecture made up unique collection networks, each adapted certain level subtypes. An Improved Convolutional Neural (CNN), DenseNet, an improved VGG19 are among networks painstakingly developed extract distinguishing cell pictures, enhancing classification accuracy. This uses cross-entropy loss function combination with approach improve performance our even more. improves training process, allowing discover patterns correlations dataset.

Language: Английский

---

Leukemia Detection and Classification Based on Machine Learning and CNN: A Review

Indonesian Journal of Computer Science, Journal Year: 2024, Volume and Issue: 13(3)

Published: June 15, 2024

Advancements in data mining methods have significantly improved disease diagnosis, particularly the realm of leukemia detection. Leukemia, a complex cancer affecting white blood cells, poses significant challenges diagnosis and management due to its diverse manifestations. Various machine learning algorithms, including Convolutional Neural Networks (CNNs), Support Vector Machines (SVMs), Random Forests (RF), Decision Trees (DTs), K-Nearest Neighbors (K-NN), Logistic regression (LR) Naïve Bayes (NB) classifiers, been employed accurately classify cases based on datasets image analyses. This paper provides comprehensive overview comparison these classification techniques, highlighting their effectiveness diagnosing different subtypes. Additionally, discusses methodology findings several studies focusing detection, emphasizing significance enhancing diagnostic accuracy treatment planning. Furthermore, it explores future directions leveraging for need standardized datasets, algorithm refinement, integration with clinical personalized strategies.

Language: Английский

---