Automatic Kidney Stone Composition Analysis Method Based on Dual-energy CT

Current Medical Imaging Formerly Current Medical Imaging Reviews, Год журнала: 2023, Номер 20

Опубликована: Сен. 11, 2023

The composition of kidney stones is related to the hardness stones. Knowing before surgery can help plan laser power and operation time percutaneous nephroscopic surgery. Moreover, patients be treated with medications if stone compounded by uric acid treatment, which relieve pain However, although literature generally reports analysis method base on dual-energy CT images, accuracy these methods not enough; they need manual delineation location, cannot analyze mixed stones.This study aimed overcome problem identifying composition; we an accurate stones.In this paper, proposed automatic algorithm based a image. first segmented mask deep learning model, then analyzed each machine model.The experimental results indicate that segment accurately (AUC=0.96) predict (mean Acc=0.86, mean Se=0.75, Sp=0.9, F1=0.75, AUC=0.83, MR (Exact match ratio)=0.6).The location stones, guide its treatment. Experimental show weighting strategy improve segmentation performance. In addition, multi-label classification model precisely, including

Язык: Английский

---

ECG-based convolutional neural network in pediatric obstructive sleep apnea diagnosis

Computers in Biology and Medicine, Год журнала: 2023, Номер 167, С. 107628 - 107628

Опубликована: Окт. 24, 2023

Obstructive sleep apnea (OSA) is a prevalent respiratory condition in children and characterized by partial or complete obstruction of the upper airway during sleep. The events OSA induce transient alterations cardiovascular system that ultimately can lead to increased risk affected children. Therefore, timely accurate diagnosis utmost importance. However, polysomnography (PSG), standard diagnostic test for pediatric OSA, complex, uncomfortable, costly, relatively inaccessible, particularly low-resource environments, thereby resulting substantial underdiagnosis. Here, we propose novel deep-learning approach simplify using raw electrocardiogram tracing (ECG). Specifically, new convolutional neural network (CNN)-based regression model was implemented automatically predict estimating its severity based on apnea-hypopnea index (AHI) deriving 4 categories. For this purpose, overnight ECGs from 1,610 PSG recordings obtained Childhood Adenotonsillectomy Trial (CHAT) database were used. randomly divided into approximately 60%, 20%, 20% training, validation, testing, respectively. performance proposed CNN largely outperformed most previous algorithms relied ECG-derived features (4-class Cohen's kappa coefficient 0.373 versus 0.166). AHI cutoff values 1, 5, 10 events/hour, binary classification achieved sensitivities 84.19%, 76.67%, 53.66%; specificities 46.15%, 91.39%, 98.06%; accuracies 75.92%, 86.96%, 91.97%, be readily identified our model, which provides simpler, faster, more accessible clinical practice.

Язык: Английский

---

Cancer detection and classification using a simplified binary state vector machine

Medical & Biological Engineering & Computing, Год журнала: 2024, Номер 62(5), С. 1491 - 1501

Опубликована: Фев. 1, 2024

Язык: Английский

---

A Novel Web Framework for Cervical Cancer Detection System: A Machine Learning Breakthrough

IEEE Access, Год журнала: 2024, Номер 12, С. 41542 - 41556

Опубликована: Янв. 1, 2024

Cervical cancer, the second most prevalent cancer among women worldwide, is primarily attributed to human papillomavirus (HPV).Despite advances in healthcare, it remains a significant cause of mortality across diverse regions, surpassing other hereditary cancers.Early detection pivotal, as survival rates exceed 90% when disease identified its early stages.In response this critical need, we introduce WFC2DS (Web Framework for Cancer Detection System), novel expert web system specifically designed revolutionize cervical diagnosis.WFC2DS integrates sophisticated ensemble machine learning classification algorithms, including Artificial Neural Network (ANN), AdaBoost, K-Nearest Neighbor (KNN), Random Forest Classifier (RFC), Support Vector Machine (SVM), and Decision Tree (DT).This approach enables comprehensive analysis large dataset comprising information from 858 patients with 36 attributes, primary objective being using last attribute, Biopsy, target variable.Our evaluation criteria encompass accuracy, specificity, sensitivity, F1 score.Among RFC DT emerge promising, demonstrating exceptional performance an accuracy 98.1% score 0.98.AdaBoost shows 97.4% 0.98, ANN attains 97.7% 0.96, SVM achieves 96.2% KNN reaches 90.6% 0.91.This research significantly contributes reducing global burden emphasizing transformative advancements women's healthcare.WFC2DS, cutting-edge techniques, not only improves diagnosis but also enhances overall healthcare landscape worldwide.

Язык: Английский

---

Detection and prediction of diabetes using effective biomarkers

Computer Methods in Biomechanics and Biomedical Engineering Imaging & Visualization, Год журнала: 2023, Номер 12(1)

Опубликована: Окт. 5, 2023

ABSTRACTDiabetes is a prevalent and costly condition, with early diagnosis pivotal in mitigating ‎its progression complications. The diagnostic process often contends data ‎ambiguity decision uncertainty, adding complexity to achieving definitive ‎outcomes. This study addresses the diabetes challenge through mining ‎and machine learning techniques. It involves training various ‎algorithms conducting statistical analysis on dataset comprising 520 patients, ‎encompassing both normal diabetic cases, discern influential features.‎ Incorporating 17 features as classifier inputs, this research evaluates ‎performance using four reputable techniques: support vector (SVM), random ‎forest (RF), multi-layer perceptron (MLP), k-nearest neighbor (kNN). outcomes ‎underscore SVM model's superior performance, boasting accuracy, specificity, ‎sensitivity values of 98.78±1.96%, 99.28±1.63%, 97.32±2.45%, ‎respectively, across 50 iterations. findings establish preferred method ‎for diagnosis.‎ highlights efficacy models ‎diabetes diagnosis. While these methods exhibit respectable predictive their ‎integration physician's assessment promises even better patient outcomes.‎KEYWORDS: Data miningdiabetesSVMdetectionprediction Abbreviations ANN=Artificial Neural NetworkAUC=Area under CurveCDC=Centers for Disease ControlCPCSSN=Canadian Primary Care Sentinel Surveillance NetworkDT=Decision TreeFN=False NegativeFP=False PositivekNN=k Nearest NeighborLDA=Linear Discrimination AnalysisLR=Logistic RegressionML=Machine LearningMLP=Multi-Layer PerceptronNB=Naive BayesianPIDD=Pima Indians Diabetes DatasetRF=Random ForestROC=Receiver Operating CharacteristicSVM=Support Vector MachineTN=True NegativeTP=True PositiveUKPDS=UK Prospective StudyDisclosure statementNo potential conflict interest was reported by author(s)Authors' contributionsAll authors evenly contributed whole work. All read approved final manuscript.Availability materialsThe used paper cited throughout paper.Ethical approvalThis article does not contain any studies human participants performed authors.Additional informationFundingNo source funding work.Notes contributorsMohammad Ehsan FarnoodianMohammad Farnoodian received B.S. degree biomedical engineering-‎‎bioelectric from Tehran Medical Science, Islamic Azad University, Tehran, Iran, earned his M.S. engineering-bioelectric Science ‎Research branch, 2023. He passionately ‎dedicated examination interpretation data, particularly ‎the context disease prediction detection. His academic pursuits involve in-‎depth exploration intricacies, specific focus ‎employing data-driven approaches anticipation identification.‎Mohammad Karimi MoridaniMohammad Moridani BS electrical engineering-‎Electronic 2006, he obtained MS Ph.D. degrees ‎engineering-bioelectric 2008 2015, respectively. Currently, serves an ‎assistant professor engineering department at ‎Science, University Iran. focuses ‎biomedical signal image processing, nonlinear time series analysis, ‎cognitive science, applications ranging ECG, HRV, EEG ‎signal processing detection epileptic seizure ‎prediction, pattern recognition, facial beauty ‎watermarking, more. driven passion contribute meaningfully scientific community employs methodologies address ‎critical challenges healthcare related fields.‎Hanieh MokhberHanieh Mokhber ‎from science. Her scholarly endeavors ‎involve meticulous complexities ‎with unwavering emphasis harnessing ‎to anticipate identify diseases.‎

Язык: Английский

---