Abstract:Purpose: To perform a radiomics analysis with comparisons of multidomain features and a variety of feature selection strategies and classifiers, with the goal of evaluating the value of quantified radiomics method for noninvasively differentiating autoimmune pancreatitis (AIP) from pancreatic ductal adenocarcinoma (PDAC) in 18 F-fluorodeoxglucose positron emission tomography/computed tomography (18 F FDG PET/CT) images. Methods: We extracted 251 expert-designed features from 2D and 3D PET/CT images of 111 patients, and recombined these features into five feature sets according to their modalities and dimensions. Among the five feature sets, the optimal one was found leveraging four feature selection strategies and four machine learning classifiers based on the area under the receiver operating characteristic curve (AUC). The feature selection strategies include spearman's rank correlation coefficient, minimum redundancy maximum relevance, support vector machine recursive feature elimination (SVM-RFE), and no feature selection, while the classifiers are random forest, adaptive boosting, support vector machine (SVM) with the Gaussian radial basis function, and SVM with the linear kernel function respectively. Based on the optimal feature set, these feature selection strategies and classifiers were comparatively studied to achieve the best differentiation. Finally, the quantified radiomics prediction model was developed based on the best combination of the feature selection strategy and classifier, and it was compared with two clinical factors based prediction models, and human doctors using nested cross-validation in terms of AUC, accuracy, sensitivity, and specificity. Results: Comparison experiments demonstrated that CT features and three-dimensional (3D) features performed better than positron emission tomography (PET) features and three-dimensional (2D) features respectively, and multidomain features were superior to single domain features. In addition, the combination of SVM-RFE feature selection strategy and Linear SVM classifier had the highest diagnostic performance (i.e., AUC = 0.93 ± 0.01, ACC = 0.85 ± 0.02, SEN = 0.86 ± 0.03, SPE = 0.84 ± 0.03). The quantified radiomics model developed is significantly superior to both human doctors and clinical factors based prediction models in terms of accuracy and specificity. Conclusions: Our preliminary results confirmed that the quantified radiomics method could aid the noninvasive differentiation of AIP and PDAC in 18 F FDG PET/CT images and the integration of multidomain features is beneficial for the differentiation.

Radiomics analysis of the optic nerve for detecting dysthyroid optic neuropathy, based on water-fat imaging

White-matter Alterations in Dysthyroid Optic Neuropathy: a Diffusion Kurtosis Imaging Study Using Tract-Based Spatial Statistics

The diagnostic value of the IDEAL-T2WI sequence in dysthyroid optic neuropathy: a quantitative analysis of the optic nerve and cerebrospinal fluid in the optic nerve sheath

Diagnosis of dysthyroid optic neuropathy: combined value of orbital MRI and intracranial visual pathway diffusion kurtosis imaging

Risk prediction of dysthyroid optic neuropathy based on CT imaging features combined the bony orbit with the soft tissue structures

Multi-Parametric Diffusion Tensor Imaging of The Optic Nerve for Detection of Dysthyroid Optic Neuropathy in Patients With Thyroid-Associated Ophthalmopathy

Three-dimensional soft tissue reconstruction and volume measurement used for the diagnosis of dysthyroid optic neuropathy

Visual function changes of dysthyroid optic neuropathy and ROC curve analysis for early diagnostic indicators

Optic nerve compression associated with visual cortex functional alteration in dysthyroid optic neuropathy: A combined orbital and brain imaging study

Functional and Morphological Brain Alterations in Dysthyroid Optic Neuropathy: A Combined Resting‐State fMRI and Voxel‐Based Morphometry Study

Reduced macular inner retinal thickness and microvascular density in the early stage of patients with dysthyroid optic neuropathy

Characteristics of optical coherence tomography image in diabetic optic neuropathy

Preliminary Diffusion‐Tensor Imaging Evidence for Trans‐Synaptic Axonal Degeneration in Dysthyroid Optic Neuropathy Due to Thyroid‐Associated Ophthalmopathy

Aberrant spontaneous brain activity in patients with thyroid-associated ophthalmopathy with and without optic neuropathy: a resting-state functional MRI study

Radiomic Evaluations of the Diagnostic Performance of DM, DBT, DCE MRI, DWI, and Their Combination for the Diagnosisof Breast Cancer

Graph theoretical analysis and independent component analysis of diabetic optic neuropathy: A resting‐state functional magnetic resonance imaging study

Reduced Contrast Sensitivity Function Correlated with Superficial Retinal Capillary Plexus Impairment in Early Stage of Dysthyroid Optic Neuropathy

Radiomics: a novel feature extraction method for brain neuron degeneration disease using 18 F-FDG PET imaging and its implementation for Alzheimer's disease and mild cognitive impairment.

Quantitative radiomic biomarkers for discrimination between neuromyelitis optica spectrum disorder and multiple sclerosis

Radiomics analysis for the differentiation of autoimmune pancreatitis and pancreatic ductal adenocarcinoma in 18 F-FDG PET/CT

Dysthyroid optic neuropathy: Demographics, risk factors, investigations, and management outcomes