Automatic and Early Detection of Parkinson’s Disease by Analyzing Acoustic Signals Using Classification Algorithms Based on Recursive Feature Elimination Method

Diagnostics, Journal Year: 2023, Volume and Issue: 13(11), P. 1924 - 1924

Published: May 31, 2023

Parkinson's disease (PD) is a neurodegenerative condition generated by the dysfunction of brain cells and their 60-80% inability to produce dopamine, an organic chemical responsible for controlling person's movement. This causes PD symptoms appear. Diagnosis involves many physical psychological tests specialist examinations patient's nervous system, which several issues. The methodology method early diagnosis based on analysing voice disorders. extracts set features from recording voice. Then machine-learning (ML) methods are used analyse diagnose recorded distinguish cases healthy ones. paper proposes novel techniques optimize evaluating selected hyperparameter tuning ML algorithms diagnosing dataset was balanced synthetic minority oversampling technique (SMOTE) were arranged according contribution target characteristic recursive feature elimination (RFE) algorithm. We applied two algorithms, t-distributed stochastic neighbour embedding (t-SNE) principal component analysis (PCA), reduce dimensions dataset. Both t-SNE PCA finally fed resulting into classifiers support-vector machine (SVM), K-nearest neighbours (KNN), decision tree (DT), random forest (RF), multilayer perception (MLP). Experimental results proved that proposed superior existing studies in RF with algorithm yielded accuracy 97%, precision 96.50%, recall 94%, F1-score 95%. In addition, MLP 98%, 97.66%, 96%, 96.66%.

Language: Английский

---

Effective Early Detection of Epileptic Seizures through EEG Signals Using Classification Algorithms Based on t-Distributed Stochastic Neighbor Embedding and K-Means

Diagnostics, Journal Year: 2023, Volume and Issue: 13(11), P. 1957 - 1957

Published: June 3, 2023

Epilepsy is a neurological disorder in the activity of brain cells that leads to seizures. An electroencephalogram (EEG) can detect seizures as it contains physiological information neural brain. However, visual examination EEG by experts time consuming, and their diagnoses may even contradict each other. Thus, an automated computer-aided diagnosis for diagnostics necessary. Therefore, this paper proposes effective approach early detection epilepsy. The proposed involves extraction important features classification. First, signal components are decomposed extract via discrete wavelet transform (DWT) method. Principal component analysis (PCA) t-distributed stochastic neighbor embedding (t-SNE) algorithm were applied reduce dimensions focus on most features. Subsequently, K-means clustering + PCA t-SNE used divide dataset into subgroups representative extracted from these steps fed extreme gradient boosting, K-nearest neighbors (K-NN), decision tree (DT), random forest (RF) multilayer perceptron (MLP) classifiers. experimental results demonstrated provides superior those existing studies. During testing phase, RF classifier with DWT achieved accuracy 97.96%, precision 99.1%, recall 94.41% F1 score 97.41%. Moreover, attained 98.09%, 93.9% 96.21%. In comparison, MLP reached 98.98%, 99.16%, 95.69% 97.4%.

Language: Английский

---

DSCNet: Deep Skip Connections-Based Dense Network for ALL Diagnosis Using Peripheral Blood Smear Images

Diagnostics, Journal Year: 2023, Volume and Issue: 13(17), P. 2752 - 2752

Published: Aug. 24, 2023

Acute lymphoblastic leukemia (ALL) is a life-threatening hematological malignancy that requires early and accurate diagnosis for effective treatment. However, the manual of ALL time-consuming can delay critical treatment decisions. To address this challenge, researchers have turned to advanced technologies such as deep learning (DL) models. These models leverage power artificial intelligence analyze complex patterns features in medical images data, enabling faster more ALL. existing DL-based suffers from various challenges, computational complexity, sensitivity hyperparameters, difficulties with noisy or low-quality input images. these issues, paper, we propose novel Deep Skip Connections-Based Dense Network (DSCNet) tailored using peripheral blood smear The DSCNet architecture integrates skip connections, custom image filtering, Kullback–Leibler (KL) divergence loss, dropout regularization enhance its performance generalization abilities. leverages connections vanishing gradient problem capture long-range dependencies, while filtering enhances relevant data. KL loss serves optimization objective, predictions. Dropout employed prevent overfitting during training, promoting robust feature representations. experiments conducted on an augmented dataset highlight effectiveness DSCNet. proposed outperforms competing methods, showcasing significant enhancements accuracy, sensitivity, specificity, F-score, area under curve (AUC), achieving increases 1.25%, 1.32%, 1.12%, 1.24%, 1.23%, respectively. approach demonstrates potential tool diagnosis, applications clinical settings improve patient outcomes advance detection research.

Language: Английский

---

A2M-LEUK: attention-augmented algorithm for blood cancer detection in children

Neural Computing and Applications, Journal Year: 2023, Volume and Issue: 35(24), P. 18059 - 18071

Published: June 1, 2023

Abstract Leukemia is a malignancy that affects the blood and bone marrow. Its detection classification are conventionally done through labor-intensive specialized methods. The diagnosis of cancer in children critical task requires high precision accuracy. This study proposes novel approach utilizing attention mechanism-based machine learning conjunction with image processing techniques for precise leukemia cells. proposed attention-augmented algorithm (A2M-LEUK) an innovative leverages mechanisms to improve children. A2M-LEUK was evaluated on dataset cell images achieved remarkable performance metrics: Precision = 99.97%, Recall 100.00%, F1-score 99.98%, Accuracy 99.98%. These results indicate accuracy sensitivity identifying categorizing leukemia, its potential reduce workload medical professionals leukemia. method provides promising accurate efficient cells, which could potentially treatment Overall, improves reduces professionals.

Language: Английский

---

Acute lymphocytic leukemia detection and subtype classification via extended wavelet pooling based-CNNs and statistical-texture features

Image and Vision Computing, Journal Year: 2024, Volume and Issue: 147, P. 105064 - 105064

Published: May 3, 2024

Language: Английский

---

Multiscale adaptive and attention-dilated convolutional neural network for efficient leukemia detection model with multiscale trans-res-Unet3+ -based segmentation network

Biomedical Signal Processing and Control, Journal Year: 2024, Volume and Issue: 90, P. 105847 - 105847

Published: Jan. 4, 2024

Language: Английский

---

Utilizing Deep Feature Fusion for Automatic Leukemia Classification: An Internet of Medical Things-Enabled Deep Learning Framework

Sensors, Journal Year: 2024, Volume and Issue: 24(13), P. 4420 - 4420

Published: July 8, 2024

Acute lymphoblastic leukemia, commonly referred to as ALL, is a type of cancer that can affect both the blood and bone marrow. The process diagnosis difficult one since it often calls for specialist testing, such tests, marrow aspiration, biopsy, all which are highly time-consuming expensive. It essential obtain an early ALL in order start therapy timely suitable manner. In recent medical diagnostics, substantial progress has been achieved through integration artificial intelligence (AI) Internet Things (IoT) devices. Our proposal introduces new AI-based Medical (IoMT) framework designed automatically identify leukemia from peripheral smear (PBS) images. this study, we present novel deep learning-based fusion model detect types leukemia. system seamlessly delivers diagnostic reports centralized database, inclusive patient-specific After collecting samples hospital, PBS images transmitted cloud server WiFi-enabled microscopic device. server, capable classifying configured. trained using dataset including 6512 original segmented 89 individuals. Two input channels used purpose feature extraction model. These include VGG16 responsible extracting features images, whereas DenseNet-121 two output merged together, dense layers categorization suggested obtains accuracy 99.89%, precision 99.80%, recall 99.72%, places excellent position proposed outperformed several state-of-the-art Convolutional Neural Network (CNN) models terms performance. Consequently, potential save lives effort. For more comprehensive simulation entire methodology, web application (Beta Version) developed study. This determine presence or absence findings study hold significant biomedical research, particularly enhancing computer-aided detection.

Language: Английский

---

A review on Leukemia Cancer detection and classification: Integrating classical approaches to advanced AI techniques

AIP conference proceedings, Journal Year: 2025, Volume and Issue: 3255, P. 030006 - 030006

Published: Jan. 1, 2025

Language: Английский

---

Detecting Acute Lymphocytic Leukemia in Individual Blood Cell Smear Images

Engineering Technology & Applied Science Research, Journal Year: 2025, Volume and Issue: 15(1), P. 19167 - 19173

Published: Feb. 2, 2025

Acute Lymphocytic Leukemia (ALL) is a form of blood cancer that mainly affects lymphocytes and white cells. The severity this varies progresses quickly, requiring immediate intensive treatment making quick accurate diagnosis essential. This study presents diagnostic model for the ALL using deep learning. YOLOv8 achieved 95% accuracy when trained on C-NMC dataset 94% ALL-IDB2 while maintaining generalization. outperformed other models such as SVM, ResNet-50, hybrid integrates ResNet-50 with SVM classifier, DenseNet121. YOLOv8, its strong architecture, can efficiently extract intricate patterns from medical imaging data diagnose ALL. proposed potentially reduce pathologist workloads improve patient diagnosis. research contributes to field by providing reliable tool automated leukemia detection, paving way further advances in image analysis.

Language: Английский

---

An Efficient System for Detection and Classification of Acute Lymphoblastic Leukemia Using Semi-Supervised Segmentation Technique

International Research Journal of Multidisciplinary Technovation, Journal Year: 2025, Volume and Issue: unknown, P. 121 - 134

Published: March 24, 2025

Acute lymphoblastic leukemia (ALL), sometimes referred to as hematopoietic cancer or blood cancer, is a group of cancers that impact lymphocytes, which are white cells. Improving patient outcomes and developing efficient treatment plans depend on early precise diagnosis. The lack labelled data makes it difficult segment lymphoblast cells from microscopic images. Our research aimed achieve unsupervised approach for accurate segmentation blasted lymphocyte cells, thereby improving the overall performance ALL detection classification into its subtypes L1, L2 L3. proposed method employs k-means segmentation, where parameter k tuned, optimal value determined based quality. For better performance, generated segments evaluated against ground truth image Structural Similarity Index Measure (SSIM), Dice similarity coefficient (DSC) Intersection over union (IoU). algorithm iterates different values k, assesses quality, selects with highest evaluation score. Customized convolutional neural networks employed categorization. augmentation technique has been applied expand amount training in order enhance model efficiency. ALL-IDB dataset used assess model's experimental results showed suggested can identify cell an accuracy 99%. We succeeded detecting acute 100% accuracy. not only enhances significantly but also determines clusters (k) more effective segmentation.

Language: Английский

---