Artificial Intelligence in Nuclear Cardiology: An Update and Future Trends

Seminars in Nuclear Medicine, Journal Year: 2024, Volume and Issue: 54(5), P. 648 - 657

Published: March 22, 2024

Language: Английский

---

AI in SPECT Imaging: Opportunities and Challenges

Seminars in Nuclear Medicine, Journal Year: 2025, Volume and Issue: unknown

Published: April 1, 2025

Language: Английский

---

Time-to-event overall survival prediction in glioblastoma multiforme patients using magnetic resonance imaging radiomics

La radiologia medica, Journal Year: 2023, Volume and Issue: 128(12), P. 1521 - 1534

Published: Sept. 26, 2023

Abstract Purpose Glioblastoma Multiforme (GBM) represents the predominant aggressive primary tumor of brain with short overall survival (OS) time. We aim to assess potential radiomic features in predicting time-to-event OS patients GBM using machine learning (ML) algorithms. Materials and methods One hundred nineteen GBM, who had T1-weighted contrast-enhanced T2-FLAIR MRI sequences, along clinical data time, were enrolled. Image preprocessing included 64 bin discretization, Laplacian Gaussian (LOG) filters three Sigma values eight variations Wavelet Transform. Images then segmented, followed by extraction 1212 features. Seven feature selection (FS) six ML algorithms utilized. The combination preprocessing, FS, (12 × 7 6 = 504 models) was evaluated multivariate analysis. Results Our analysis showed that best prognostic FS/ML combinations are Mutual Information (MI)/Cox Boost, MI/Generalized Linear Model Boosting (GLMB) Network (GLMN), all which done via LOG (Sigma 1 mm) method (C-index 0.77). filter mm method, MI, GLMB GLMN achieved significantly higher C-indices than other (all p < 0.05, mean 0.65, 0.70, 0.64, respectively). Conclusion capable MRI-based radiomics variables might appear promising assisting clinicians prediction GBM. Further research is needed establish applicability management clinic.

Language: Английский

---

Machine learning-based diagnosis and risk classification of coronary artery disease using myocardial perfusion imaging SPECT: A radiomics study

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

Published: Sept. 10, 2023

Abstract This study aimed to investigate the diagnostic performance of machine learning-based radiomics analysis diagnose coronary artery disease status and risk from rest/stress Myocardial Perfusion Imaging (MPI) single-photon emission computed tomography (SPECT). A total 395 patients suspicious who underwent 2-day stress-rest protocol MPI SPECT were enrolled in this study. The left ventricle myocardium, excluding cardiac cavity, was manually delineated on rest stress images define a volume interest. Added clinical features (age, sex, family history, diabetes status, smoking, ejection fraction), 118 features, extracted establish different feature sets, including Rest-, Stress-, Delta-, Combined-radiomics (all together) sets. data randomly divided into 80% 20% subsets for training testing, respectively. classifiers built combinations three selections, nine learning algorithms evaluated two tasks, 1) normal/abnormal (no CAD vs. CAD) classification, 2) low-risk/high-risk classification. Different metrics, area under ROC curve (AUC), accuracy (ACC), sensitivity (SEN), specificity (SPE), reported models’ evaluation. Overall, models Stress set (compared other sets), second task 1 models) revealed better performance. Stress-mRMR-KNN (feature set-feature selection-classifier) reached highest with AUC, ACC, SEN, SPE equal 0.61, 0.63, 0.64, 0.6, Stress-Boruta-GB model achieved 2 0.79, 0.76, 0.75, Diabetes family, dependence count non-uniformity normalized, NGLDM which is representative region interest most frequently selected promising results classification using radiomics. proposed are helpful alleviate labor-intensive interpretation process regarding can potentially expedite process.

Language: Английский

---

Machine Learning and Deep Learning Applications in Magnetic Particle Imaging

Journal of Magnetic Resonance Imaging, Journal Year: 2024, Volume and Issue: unknown

Published: Feb. 15, 2024

In recent years, magnetic particle imaging (MPI) has emerged as a promising technique depicting high sensitivity and spatial resolution. It originated in the early 2000s where it proposed new approach to challenge low resolution achieved by using relaxometry order measure fields. MPI presents 2D 3D images with temporal resolution, non-ionizing radiation, optimal visual contrast due its lack of background tissue signal. Traditionally, were reconstructed conversion signal from induced voltage generating system matrix X-space based methods. Because image reconstruction analyses play an integral role obtaining precise information signals, newer artificial intelligence-based methods are continuously being researched developed upon. this work, we summarize review significance employment machine learning deep models for applications potential they hold future. LEVEL OF EVIDENCE: 5 TECHNICAL EFFICACY: Stage 1.

Language: Английский

---

ELTIRADS framework for thyroid nodule classification integrating elastography, TIRADS, and radiomics with interpretable machine learning

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: March 13, 2025

Early detection of malignant thyroid nodules is crucial for effective treatment, but traditional diagnostic methods face challenges such as variability in expert opinions and limited integration advanced imaging techniques. This prospective cohort study investigates a novel multimodal approach, integrating with machine learning We studied 181 patients who underwent fine-needle aspiration (FNA) biopsy, each contributing one nodule, resulting total our analysis. Data collection included sex, age, ultrasound imaging, which incorporated elastography. Features extracted from these images Thyroid Imaging Reporting System (TIRADS) scores, elastography parameters, radiomic features. The pathological results based on the FNA provided by pathologists, served gold standard nodule classification. Our methodology, termed ELTIRADS, combines features interpretable Performance evaluation showed that Support Vector Machine (SVM) classifier using TIRADS, data, achieved high accuracy (0.92), sensitivity (0.89), specificity (0.94), precision F1 score (0.89). To enhance interpretability, we used hierarchical clustering, shapley additive explanations (SHAP), partial dependence plots (PDP). combined approach holds promise enhancing malignancy detection, thereby to advancements personalized medicine field cancer research.

Language: Английский

---

Left Ventricular Myocardial Dysfunction Evaluation in Thalassemia Patients Using Echocardiographic Radiomic Features and Machine Learning Algorithms

Journal of Digital Imaging, Journal Year: 2023, Volume and Issue: 36(6), P. 2494 - 2506

Published: Sept. 21, 2023

Abstract Heart failure caused by iron deposits in the myocardium is primary cause of mortality beta-thalassemia major patients. Cardiac magnetic resonance imaging (CMRI) T2* screening technique used to detect myocardial overload, but inherently bears some limitations. In this study, we aimed differentiate patients with overload from those without (detected T2*CMRI) based on radiomic features extracted echocardiography images and machine learning (ML) normal left ventricular ejection fraction (LVEF > 55%) echocardiography. Out 91 cases, 44 thalassemia LVEF (> ≤ 20 ms 47 people 55% as control group were included study. Radiomic for each end-systolic (ES) end-diastolic (ED) image. Then, three feature selection (FS) methods six different classifiers used. The models evaluated using various metrics, including area under ROC curve (AUC), accuracy (ACC), sensitivity (SEN), specificity (SPE). Maximum relevance-minimum redundancy-eXtreme gradient boosting (MRMR-XGB) (AUC = 0.73, ACC SPE SEN 0.73), ANOVA-MLP 0.69, 0.56, 0.83), recursive elimination-K-nearest neighbors (RFE-KNN) 0.65, 0.64, 0.65) best ED, ES, ED&ES datasets. Using echocardiographic ML, it feasible predict cardiac problems overload.

Language: Английский

---

Interpretation of SPECT Wall Motion with Deep Learning

Journal of Nuclear Cardiology, Journal Year: 2024, Volume and Issue: 37, P. 101881 - 101881

Published: May 7, 2024

We sought to develop a novel deep learning (DL) workflow interpret single-photon emission computed tomography (SPECT) wall motion.

Language: Английский

---

Machine learning based on SPECT/CT to differentiate bone metastasis and benign bone lesions in lung malignancy patients

Medical Physics, Journal Year: 2023, Volume and Issue: 51(4), P. 2578 - 2588

Published: Nov. 15, 2023

Abstract Background Bone metastasis is a common event in lung cancer progression. Early diagnosis of malignant tumor with bone crucial for selecting effective treatment strategies. However, 14.3% patients are still difficult to diagnose after SPECT/CT examination. Purpose Machine learning analysis [ 99m Tc]‐methylene diphosphate ( Tc‐MDP) scans distinguish metastases from benign lesions cancer. Methods One hundred forty‐one (69 and 72 lesions) were randomly assigned the training group or testing 7:3 ratio. Lesions manually delineated using ITK‐SNAP, 944 radiomics features extracted SPECT CT images. The least absolute shrinkage selection operator (LASSO) regression was used select set, single/bimodal models established based on support vector machine (SVM). To further optimize model, best bimodal combined clinical establish an integrated Radiomics‐clinical model. diagnostic performance evaluated receiver operating characteristic (ROC) curve confusion matrix, differences between Delong test. Results optimal model comprised structural modality (CT) metabolic (SPECT), area under (AUC) 0.919 0.907 respectively. which SPECT, CT, two features, exhibited satisfactory differentiation AUC 0.939 0.925, Conclusions can effectively differentiate lesions. demonstrated performance.

Language: Английский

---

Prediction of Ablation Rate for High-Intensity Focused Ultrasound Therapy of Adenomyosis in MR Images Based on Multi-model Fusion

Deleted Journal, Journal Year: 2024, Volume and Issue: unknown

Published: March 5, 2024

This study aimed to develop a model based on radiomics and deep learning features predict the ablation rate in patients with adenomyosis undergoing high-intensity focused ultrasound (HIFU) therapy. A total of 119 who received HIFU therapy were retrospectively analyzed. Participants included training testing queues 7:3 ratio. Radiomics extracted from T2-weighted imaging (T2WI) images, VGG-19 was used extract advanced features. An ensemble multi-model fusion for predicting efficacy proposed, which consists four base classifiers evaluated using accuracy, precision, recall, F-score, area under receiver operating characteristic curve (AUC). The predictive performance combined combining outperformed feature models alone, accuracy 0.848 0.814 test sets, AUC 0.916 0.861, respectively. Compared that make up model, also exhibited better prediction performance. incorporating both had certain value could help select would benefit

Language: Английский

---