Enhanced image diagnosing approach in medicine using quantum adaptive machine learning techniques

Optical and Quantum Electronics, Год журнала: 2024, Номер 56(4)

Опубликована: Янв. 30, 2024

Язык: Английский

---

Ensemble Federated Learning Approach for Diagnostics of Multi-Order Lung Cancer

Diagnostics, Год журнала: 2023, Номер 13(19), С. 3053 - 3053

Опубликована: Сен. 25, 2023

The early detection and classification of lung cancer is crucial for improving a patient's outcome. However, the traditional methods are based on single machine learning models. Hence, this limited by availability quality data at centralized computing server. In paper, we propose an ensemble Federated Learning-based approach multi-order classification. This combines multiple models trained different datasets allowing improvising accuracy generalization. Moreover, Learning enables use distributed while ensuring privacy security. We evaluate Kaggle dataset compare results with demonstrate 89.63%

Язык: Английский

---

Multiclass Paddy Disease Detection Using Filter-Based Feature Transformation Technique

IEEE Access, Год журнала: 2023, Номер 11, С. 109477 - 109487

Опубликована: Янв. 1, 2023

Pests and diseases are the big issues in paddy production they make farmers to lose around 20% of rice yield world-wide. Identification leaves at early stage through thermal image cameras will be helpful for avoiding such losses. The objective this work is implement a Modified Lemurs Optimization Algorithm as filter-based feature transformation technique enhancing accuracy detecting various machine learning techniques by processing images leaves. original altered inspiration Sine Cosine developing proposed Algorithm. Five namely blast, brown leaf spot, folder, hispa, bacterial blight considered work. A total six hundred thirty-six including healthy diseased analysed. Seven statistical features seven Box-Cox transformed extracted from each four K-Nearest Neighbor classifier, Random Forest Linear Discriminant Analysis Classifier, Histogram Gradient Boosting Classifier tested. All these classifiers provide balanced less than 65% their performance improved usage transform based on Optimization. Especially, 90% achieved using classifier.

Язык: Английский

---

A refined ResNet18 architecture with Swish activation function for Diabetic Retinopathy classification

Biomedical Signal Processing and Control, Год журнала: 2023, Номер 88, С. 105630 - 105630

Опубликована: Окт. 26, 2023

Язык: Английский

---

ResDenseNet:Hybrid Convolutional Neural Network Model for Advanced Classification of Diabetic Retinopathy(DR) in Retinal Image Analysis

International Journal of Computational and Experimental Science and Engineering, Год журнала: 2024, Номер 10(4)

Опубликована: Дек. 22, 2024

Preventing vision loss in diabetic retinopathy (DR) requires early and precise detection. Although strong feature extraction is required there class imbalance the current methods, deep learning (DL) techniques have showed promise DR classification. With components from both ResNeXt DenseNet designs, a unique DL architecture for classification proposed this work. A that integrates work.To address issues classification, method channel-wise masking with an attention mechanism. The network able to learn less frequent stages because reduces influence of majority concentrates on important features. To improve interpretability confidence model's predictions, incorporation Explainable AI (XAI) approaches also covered.Our findings show suggested approach outperforms architectures, achieving better sensitivity differentiating phases at 0.82 accuracy 0.87. This shows new has improving categorization, which could result earlier diagnoses patient outcomes.

Язык: Английский

---

An artificial intelligence-based decision support system for early and accurate diagnosis of Parkinson’s Disease

Decision Analytics Journal, Год журнала: 2023, Номер 10, С. 100381 - 100381

Опубликована: Дек. 13, 2023

People with Parkinson's Disease (PD) might struggle sadness, restlessness, or difficulty speaking, chewing, swallowing. A diagnosis can be challenging because there is no specific PD test. It diagnosed by doctors using a neurological exam and medical history. This study proposes several Machine Learning (ML) algorithms to predict PD. These ML include K-Nearest Neighbor (KNN), Random Forest (RF), Support Vector (SVM), eXtreme Gradient Boosting (XGBoost), their ensemble methods publicly available dataset 195 instances. The are used classify homogeneous XGBoost techniques reduced amount of entropy. Synthetic Minority Oversampling Technique (SMOTE) utilized handle imbalanced data, 10-fold cross-validation employed for evaluation. results show that the XGBoost-Random outperforms other 98% accuracy Matthew's correlation coefficient value 0.93.

Язык: Английский

---

Efficient segmentation of active and inactive plaques in FLAIR-images using DeepLabV3Plus SE with efficientnetb0 backbone in multiple sclerosis

Scientific Reports, Год журнала: 2024, Номер 14(1)

Опубликована: Июль 15, 2024

This research paper introduces an efficient approach for the segmentation of active and inactive plaques within Fluid-attenuated inversion recovery (FLAIR) images, employing a convolutional neural network (CNN) model known as DeepLabV3Plus SE with EfficientNetB0 backbone in Multiple sclerosis (MS), demonstrates its superior performance compared to other CNN architectures. The study encompasses various critical components, including dataset pre-processing techniques, utilization Squeeze Excitation Network (SE-Block), atrous spatial separable pyramid Block enhance capabilities. Detailed descriptions procedures, such removing cranial bone segment, image resizing, normalization, are provided. analyzed cross-sectional cohort 100 MS patients brain plaques, examining 5000 MRI slices. After filtering, 1500 slices were utilized labeling deep learning. training process adopts dice coefficient loss function utilizes Adam optimization. evaluated model's using multiple metrics, intersection over union (IOU), Dice Score, Precision, Recall, F1-Score, offers comparative analysis Results demonstrate ability proposed model, evidenced by IOU 69.87, Score 76.24, Precision 88.89, Recall 73.52, F1-Score 80.47 DeepLabV3+SE_EfficientNetB0 model. contributes advancement plaque FLAIR images compelling substantial potential medical diagnosis.

Язык: Английский

---

Redefining retinal vessel segmentation: empowering advanced fundus image analysis with the potential of GANs

Frontiers in Medicine, Год журнала: 2024, Номер 11

Опубликована: Окт. 21, 2024

Retinal vessel segmentation is a critical task in fundus image analysis, providing essential insights for diagnosing various retinal diseases. In recent years, deep learning (DL) techniques, particularly Generative Adversarial Networks (GANs), have garnered significant attention their potential to enhance medical analysis. This paper presents novel approach by harnessing the capabilities of GANs. Our method, termed GANVesselNet, employs specialized GAN architecture tailored intricacies structures. dual-path network employed, featuring an Auto Encoder-Decoder (AED) pathway and UNet-inspired pathway. unique combination enables efficiently capture multi-scale contextual information, improving accuracy segmentation. Through extensive experimentation on publicly available datasets, including STARE DRIVE, GANVesselNet demonstrates remarkable performance compared traditional methods state-of-the-art approaches. The proposed exhibits superior sensitivity (0.8174), specificity (0.9862), (0.9827) segmenting vessels dataset, achieves commendable results DRIVE dataset with (0.7834), (0.9846), (0.9709). Notably, previously unseen data, underscoring its real-world clinical applications. Furthermore, we present qualitative visualizations generated segmentations, illustrating network’s proficiency accurately delineating vessels. summary, this introduces powerful By capitalizing advanced GANs incorporating architecture, offers quantum leap accuracy, opening new avenues enhanced analysis improved decision-making.

Язык: Английский

---

Determining the Presence and Size of Shoulder Lesions in Sows Using Computer Vision

Animals, Год журнала: 2023, Номер 14(1), С. 131 - 131

Опубликована: Дек. 29, 2023

Shoulder sores predominantly arise in breeding sows and often result untimely culling. Reported prevalence rates vary significantly, spanning between 5% 50% depending upon the type of crate flooring inside a farm, animal’s body condition, or an existing injury that causes lameness. These lesions represent not only welfare concern but also have economic impact due to labor needed for treatment medication. The objective this study was evaluate use computer vision techniques detecting determining size shoulder lesions. A Microsoft Kinect V2 camera captured top-down depth RGB images farrowing crates. were collected at resolution 1920 × 1080. To ensure best view lesions, selected with lying on their right left sides all legs extended. total 824 from 70 various stages development identified annotated. Three deep learning-based object detection models, YOLOv5, YOLOv8, Faster-RCNN, pre-trained COCO ImageNet datasets, implemented localize lesion area. YOLOv5 predictor as it able detect [email protected] 0.92. estimate area, pixel segmentation carried out localized region using traditional image processing like Otsu’s binarization adaptive thresholding alongside DL-based models based U-Net architecture. In conclusion, demonstrates potential effectively assessing sows, providing promising avenue improving sow reducing losses.

Язык: Английский

---

Diverter transformer-based multi-encoder-multi-decoder network model for medical retinal blood vessel image segmentation

Biomedical Signal Processing and Control, Год журнала: 2024, Номер 93, С. 106132 - 106132

Опубликована: Фев. 23, 2024

Язык: Английский

---