Hippocampal functional imaging-derived radiomics features for diagnosing cognitively impaired patients with Parkinson’s disease

BMC Neuroscience, Journal Year: 2025, Volume and Issue: 26(1)

Published: March 28, 2025

Abstract Purpose The aim of this retrospective study was to investigate whether radiomics features derived from hippocampal functional imaging can effectively differentiate cognitively impaired patients preserved with Parkinson’s disease (PD). Methods included a total 89 clinically definite PD patients, comprising 55 who werecognitively and 34 were preserved. All participants underwent magnetic resonance clinical assessments. Preprocessed data utilized derive the amplitude low-frequency fluctuations (ALFF), regional homogeneity (ReHo), voxel-mirrored homotopic connectivity (VMHC), degree centrality (DC). A standardized set subsequently extracted bilateral hippocampi, resulting in 819 features. Following feature selection, score (rad-score) logistic regression (LR) models trained. Additionally, Shapley additive explanations (SHAP) algorithm employed elucidate interpret predictions made by LR models. Finally, relationships between scores measures explored assess their significance. Results rad-score constructed using combination wavelet ReHo VMHC exhibited superior classification efficiency. These demonstrated exceptional accuracy, sensitivity, specificity distinguishing (CI-PD) (CP-PD) values 0.889, 0.900, 0.882, respectively. Furthermore, SHAP indicated that critical for classifying CI-PD patients. Importantly, our findings revealed significant associations on Hamilton Anxiety Scale, Non-Motor Symptom Montreal Cognitive Assessment ( P < 0.05, Bonferroni correction). Conclusions Our novel model model, which utilize imaging, have value diagnosing CI-PDpatients. enhance accuracy efficiency MRI diagnosis, suggesting potential applications. Clinical trial number Not applicable.

Language: Английский

---

Advancing ischemic stroke diagnosis and clinical outcome prediction using improved ensemble techniques in DSC-PWI radiomics

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: Nov. 11, 2024

Ischemic stroke is a leading global cause of death and disability expected to rise in the future. The present diagnostic techniques, like CT MRI, have some limitations distinguishing acute from chronic ischemia early detection. This study investigates function ensemble models based on dynamic radiomics features (DRF) susceptibility contrast perfusion-weighted imaging (DSC-PWI) ischemic diagnosis, neurological impairment assessment, modified Rankin Scale (mRS) outcome prediction). DRF extracted 3D images, are selected, dimensionality reduced. After that, applied. Two model structures were developed: voting classifier with 6 bagging classifiers stacking 4 classifiers. evaluated three core tasks. Stacking_ens_LR performed best for detection, LR Bagging NIH Stroke (NIHSS) prediction, NB prediction. These outcomes illustrate strength models. work showcases role management process.

Language: Английский

---

Radiomics and radiogenomics in oral oncology: Harnessing imaging data for precision medicine

Oral Oncology Reports, Journal Year: 2024, Volume and Issue: 10, P. 100444 - 100444

Published: April 27, 2024

• Radiomics analyzes image features, enhancing oral tumor characterization Radiogenomics correlates imaging with genetics, unveiling genetic influence on traits. Imaging guides personalized cancer treatment, tailoring protocols to characteristics

Language: Английский

---

Radiomics in Oncology and Radiology: Statistical Significance Versus Clinical Significance

Clinical Oncology, Journal Year: 2024, Volume and Issue: 36(9), P. e342 - e342

Published: May 8, 2024

Language: Английский

---

Predicting the invasiveness of ground-glass opacity predominant lung adenocarcinoma with clinical stage Ia: a CT-based semantic and radiomics analysis

Journal of Thoracic Disease, Journal Year: 2024, Volume and Issue: 16(10), P. 6713 - 6726

Published: Oct. 1, 2024

Limited surgery is deemed advantageous due to its potential minimize damage and preserve a greater extent of functional lung tissue, contingent upon the invasiveness adenocarcinoma (ADC). The aim this study was non-invasively predict ground-glass opacity (GGO) predominant nodules presented on preoperative computed tomography (CT) ADC patients with clinical stage Ia. We constructed primary cohort comprising 437 Ia from Tianjin Medical University Cancer Institute Hospital utilized data 135 General for validation. Radiomics features were extracted by PyRadiomics software screened spearman correlation analysis, minimum redundancy maximum relevance least absolute shrinkage selection operator (LASSO) regression analysis. radiomics score (Rad-score) formula then created linearly combining selected features, using their coefficients as weights. Univariate analysis followed multivariable logistic performed estimate independent predictors. An initial univariate Area under curve (AUC) calculated after model established through visual nomogram external Three hundred seventy-four pathologically confirmed invasive (65.4%), three predictors identified: consolidation diameter (P=0.02), texture (P=0.042) Rad-score (P<0.001). combined showed good calibration an AUC 0.911 [95% confidence interval (CI): 0.872, 0.951], compared 0.883 (95% CI: 0.849, 0.932; DeLong's test P=0.16) 0.842 0.801, 0.896; P<0.001) when or CT semantic used alone. Combined prediction accuracy validation group 0.865 0.816, 0.908), which reasonable. Our has provided non-invasive tool based characteristics that can accurately assess quantitative risk associated GGO in

Language: Английский

---

Study of prediction model for high-grade meningioma using fractal geometry combined with radiological features

Journal of Neuro-Oncology, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 4, 2024

Language: Английский

---