Can Magnetic Resonance Radiomics Analysis Discriminate Parotid Gland Tumors? A Pilot Study

Using Machine Learning on MRI Radiomics to Diagnose Parotid Tumours Before Comparing Performance with Radiologists: A Pilot Study

The parotid glands are the largest of the major salivary glands. They can harbour both benign and malignant tumours. Preoperative work-up relies on MR images and fine needle aspiration biopsy, but these diagnostic tools have low sensitivity and specificity, often leading to surgery for diagnostic purposes. The aim of this paper is (1) to develop a machine learning algorithm based on MR images characteristics to automatically classify parotid gland tumours and (2) compare its results with the diagnoses of junior and senior radiologists in order to evaluate its utility in routine practice. While automatic algorithms applied to parotid tumours classification have been developed in the past, we believe that our study is one of the first to leverage four different MRI sequences and propose a comparison with clinicians. In this study, we leverage data coming from a cohort of 134 patients treated for benign or malignant parotid tumours. Using radiomics extracted from the MR images of the gland, we train a random forest and a logistic regression to predict the corresponding histopathological subtypes. On the test set, the best results are given by the random forest: we obtain a 0.720 accuracy, a 0.860 specificity, and a 0.720 sensitivity over all histopathological subtypes, with an average AUC of 0.838. When considering the discrimination between benign and malignant tumours, the algorithm results in a 0.760 accuracy and a 0.769 AUC, both on test set. Moreover, the clinical experiment shows that our model helps to improve diagnostic abilities of junior radiologists as their sensitivity and accuracy raised by 6 % when using our proposed method. This algorithm may be useful for training of physicians. Radiomics with a machine learning algorithm may help improve discrimination between benign and malignant parotid tumours, decreasing the need for diagnostic surgery. Further studies are warranted to validate our algorithm for routine use.

Assessment of synthetic MRI to distinguish Warthin's tumor from pleomorphic adenoma in the parotid gland: comparison of two methods of positioning the region of interest for synthetic relaxometry measurement

Purpose

To assess the diagnostic potential of the synthetic MRI (SyMRI) for differentiating Warthin's tumors (WT) from pleomorphic adenomas (PA).

Materials and methods

Forty-nine individuals with parotid gland tumors (PA, n = 23; WT, n = 26) were recruited. Using two distinct regions of interest (ROI), SyMRI quantitative parameters of lesions were calculated, including mean and standard deviation (T1, T2, PD, T1sd, T2sd, and PDsd). Meanwhile, T1ratio, T2ratio, and PDratio (lesion/masseter muscle) were calculated based on the mean SyMRI quantitative parameters of masseter muscle (T1, T2, PD). Using the independent samples t test, we compared PA and WT parameters, while comparing the areas under the curve (AUC) using the DeLong's test. A multi-parameter SyMRI model was constructed using logistic regression analysis.

Results

In PA, the T1, T1sd, T2, PD, T1ratio, T2ratio, and PDratio derived from full and partial lesion ROIs were significantly higher than in WT. According to the receiver operating curve analysis, the AUC of the quantitative parameters derived from full-lesion and partial-lesion ROIs ranged from 0.722 to 0.983 for differentiating PA from WT. T1 values derived from partial-lesion ROI delineation demonstrated the best diagnostic performance among all single parameters, achieving an AUC of 0.983. Using 1322 ms as a cutoff value, the sensitivity, specificity, and accuracy were 88.46%, 100% and 93.88%, respectively.

Conclusion

The SyMRI-derived quantitative parameters demonstrated excellent performance for discriminating PA from WT in the parotid gland.

Diagnostic performance of MRI-based radiomics models using machine learning approaches for the triple classification of parotid tumors

Rationale and objectives

Preoperative differentiation of malignant tumors (MT), pleomorphic adenomas (PA), and other benign tumors of the parotid gland is critical to clinical strategy, this study aimed to develop and validate a T2-weighted image (T2WI) based radiomics model through machine learning approaches for the triple classification of parotid gland tumors.

Materials and methods

We retrospectively enrolled 147 patients from January 2010 to July 2022. T2WIs were used to extract radiomics features. Max-Relevance and Min-Redundancy (mRMR) and Extreme Gradient Boosting (XGBoost) algorithms were used to select features. Using a 5-fold cross-validation strategy, radiomics models were constructed using a Support Vector Machine (SVM), Logistic Regression (LR), and k-Nearest Neighbor (KNN) for the triple classification of parotid tumors. The three models were evaluated and compared using the receiver operator characteristic (ROC) curve, sensitivity, specificity, and accuracy.

Results

A total of 1057 radiomics features were extracted, and 8 features were selected to developed the radiomics model, including First-order Median, First-order Skewness, First-order Minimum, Original_shape_Flatness, Glcm Inverse Variance, Glcm Inverse Variance, Glszm Low Gray Level Zone Emphasis, and Glszm Small Area Low Gray Level Emphasis. The mean area under the curves (AUCs) for the radiomics models in training and validation sets through LR, SVM and KNN were 0.85 and 0.80, 0.85 and 0.80 and 0.83 and 0.79, respectively.

Conclusion

The T2WI-based radiomics models through LR, SVM and KNN demonstrated good performance in the triple classification of parotid tumors.

Prediction of malignant risk stratification model for parotid gland nodules based on clinical and conventional ultrasound features: construction and validation

Background

Ultrasound is widely used in the examination of the parotid gland, but no single ultrasound feature has demonstrated satisfactory diagnostic performance in predicting the nature of parotid nodules. Unlike the established and widely used grading systems for breast and thyroid nodules, a universally adopted and clinically accepted risk stratification system for malignancy in parotid gland nodules remains absent at present. This study aims to establish a malignant risk stratification model for parotid nodules by analyzing patients' clinical features and conventional ultrasound image characteristics.

Methods

In this study, clinical data and ultrasound images of 736 patients with parotid nodules were retrospectively analyzed. Pathological results served as the gold standard, and the patients were randomly divided into training and validation groups in a 7:3 ratio. Clinical and ultrasound features of parotid nodules in the training group were compared. Multifactor logistic regression analysis was employed to screen for risk factors of malignant nodules and quantify scores. The probability of malignant risk was assessed and classified into five grades (Grade 1, normal parotid; Grade 2, definitive benign; Grade 3, possibly benign; Grade 4, suspicious malignant; Grade 5, high probability of malignancy). The diagnostic performance of the model was assessed by using calibration curves, receiver operating characteristic curves, decision curves, and clinical impact curves.

Results

Facial symptoms, unclear margin, irregular shape, microcalcification, and abnormal cervical lymph nodes were independent risk factors for malignant parotid nodules. The area under the curve of the model was 0.850 [95% confidence interval (CI): 0.816-0.879] in the training group and 0.846 (95% CI: 0.791-0.891) in the validation group.

Conclusions

The malignancy risk stratification model based on clinical and ultrasound image features has a good differentiation between benign and malignant parotid nodules, which is helpful for diagnosis and guiding clinical treatment.

Performance of radiomics in the differential diagnosis of parotid tumors: a systematic review

Purpose

A systematic review and meta-analysis were conducted to evaluate the diagnostic precision of radiomics in the differential diagnosis of parotid tumors, considering the increasing utilization of radiomics in tumor diagnosis. Although some researchers have attempted to apply radiomics in this context, there is ongoing debate regarding its accuracy.

Methods

Databases of PubMed, Cochrane, EMBASE, and Web of Science up to May 29, 2024 were systematically searched. The quality of included primary studies was assessed using the Radiomics Quality Score (RQS) checklist. The meta-analysis was performed utilizing a bivariate mixed-effects model.

Results

A total of 39 primary studies were incorporated. The machine learning model relying on MRI radiomics for diagnosis malignant tumors of the parotid gland, demonstrated a sensitivity of 0.80 [95% CI: 0.74, 0.86], SROC of 0.89 [95% CI: 0.27-0.99] in the validation set. The machine learning model based on MRI radiomics for diagnosis malignant tumors of the parotid gland, exhibited a sensitivity of 0.83[95% CI: 0.76, 0.88], SROC of 0.89 [95% CI: 0.17-1.00] in the validation set. The models also demonstrated high predictive accuracy for benign lesions.

Conclusion

There is great potential for radiomics-based models to improve the accuracy of diagnosing benign and malignant tumors of the parotid gland. To further enhance this potential, future studies should consider implementing standardized radiomics-based features, adopting more robust feature selection methods, and utilizing advanced model development tools. These measures can significantly improve the diagnostic accuracy of artificial intelligence algorithms in distinguishing between benign and malignant tumors of the parotid gland.

Systematic review registration

https://www.crd.york.ac.uk/prospero/, identifier CRD42023434931.

Different MRI-based radiomics models for differentiating misdiagnosed or ambiguous pleomorphic adenoma and Warthin tumor of the parotid gland: a multicenter study

Purpose

To evaluate the effectiveness of MRI-based radiomics models in distinguishing between Warthin tumors (WT) and misdiagnosed or ambiguous pleomorphic adenoma (PA).

Methods

Data of patients with PA and WT from two centers were collected. MR images were used to extract radiomic features. The optimal radiomics model was found by running nine machine learning algorithms after feature reduction and selection. To create a clinical model, univariate logistic regression (LR) analysis and multivariate LR were used. The independent clinical predictors and radiomics were combined to create a nomogram. Two integrated models were constructed by the ensemble and stacking algorithms respectively based on the clinical model and the optimal radiomics model. The models' performance was evaluated using the area under the curve (AUC).

Results

There were 149 patients included in all. Gender, age, and smoking of patients were independent clinical predictors. With the greatest average AUC (0.896) and accuracy (0.839) in validation groups, the LR model was the optimal radiomics model. In the average validation group, the radiomics model based on LR did not have a higher AUC (0.795) than the clinical model (AUC = 0.909). The nomogram (AUC = 0.953) outperformed the radiomics model in terms of discrimination performance. The nomogram in the average validation group had a highest AUC than the stacking model (0.914) or ensemble model (0.798).

Conclusion

Misdiagnosed or ambiguous PA and WT can be non-invasively distinguished using MRI-based radiomics models. The nomogram exhibited excellent and stable diagnostic performance. In daily work, it is necessary to combine with clinical parameters for distinguishing between PA and WT.

Deep learning for the automatic detection and segmentation of parotid gland tumors on MRI

OBJECTIVES

Parotid gland tumors (PGTs) often occur as incidental findings on magnetic resonance images (MRI) that may be overlooked. This study aimed to construct and validate a deep learning model to automatically identify parotid glands (PGs) with a PGT from normal PGs, and in those with a PGT to segment the tumor.

MATERIALS AND METHODS

The nnUNet combined with a PG-specific post-processing procedure was used to develop the deep learning model trained on T1-weighed images (T1WI) in 311 patients (180 PGs with tumors and 442 normal PGs) and fat-suppressed (FS)-T2WI in 257 patients (125 PGs with tumors and 389 normal PGs), for detecting and segmenting PGTs with five-fold cross-validation. Additional validation set separated by time, comprising T1WI in 34 and FS-T2WI in 41 patients, was used to validate the model performance.

RESULTS AND CONCLUSION

To identify PGs with tumors from normal PGs, using combined T1WI and FS-T2WI, the deep learning model achieved an accuracy, sensitivity and specificity of 98.2% (497/506), 100% (119/119) and 97.7% (378/387), respectively, in the cross-validation set and 98.5% (67/68), 100% (20/20) and 97.9% (47/48), respectively, in the validation set. For patients with PGTs, automatic segmentation of PGTs on T1WI and FS-T2WI achieved mean dice coefficients of 86.1% and 84.2%, respectively, in the cross-validation set, and of 85.9% and 81.0%, respectively, in the validation set. The proposed deep learning model may assist the detection and segmentation of PGTs and, by acting as a second pair of eyes, ensure that incidentally detected PGTs on MRI are not missed.

Multimodality radiomics for tumor prognosis in nasopharyngeal carcinoma

BACKGROUND

The prognosis of nasopharyngeal carcinoma (NPC) is challenging due to late-stage identification and frequently undetectable Epstein-Barr virus (EBV) DNA. Incorporating radiomic features, which quantify tumor characteristics from imaging, may enhance prognosis assessment.

PURPOSE

To investigate the predictive power of radiomic features on overall survival (OS), progression-free survival (PFS), and distant metastasis-free survival (DMFS) in NPC.

MATERIALS AND METHODS

A retrospective analysis of 183 NPC patients treated with chemoradiotherapy from 2010 to 2019 was conducted. All patients were followed for at least three years. The pretreatment CT images with contrast medium, MR images (T1W and T2W), as well as gross tumor volume (GTV) contours, were used to extract radiomic features using PyRadiomics v.2.0. Robust and efficient radiomic features were chosen using the intraclass correlation test and univariate Cox proportional hazard regression analysis. They were then combined with clinical data including age, gender, tumor stage, and EBV DNA level for prognostic evaluation using Cox proportional hazard regression models with recursive feature elimination (RFE) and were optimized using 20 repetitions of a five-fold cross-validation scheme.

RESULTS

Integrating radiomics with clinical data significantly enhanced the predictive power, yielding a C-index of 0.788 ± 0.066 to 0.848 ± 0.079 for the combined model versus 0.745 ± 0.082 to 0.766 ± 0.083 for clinical data alone (p<0.05). Multimodality radiomics combined with clinical data offered the highest performance. Despite the absence of EBV DNA, radiomics integration significantly improved survival predictions (C-index ranging from 0.770 ± 0.070 to 0.831 ± 0.083 in combined model versus 0.727 ± 0.084 to 0.734 ± 0.088 in clinical model, p<0.05).

CONCLUSIONS

The combination of multimodality radiomic features from CT and MR images could offer superior predictive performance for OS, PFS, and DMFS compared to relying on conventional clinical data alone.

Characterization of parotid gland tumors: Whole-tumor histogram analysis of diffusion weighted imaging, diffusion kurtosis imaging, and intravoxel incoherent motion - A pilot study

PURPOSE

To investigate the diagnostic performance of histogram features of diffusion parameters in characterizating parotid gland tumors.

METHOD

From December 2018 to January 2023, patients who underwent diffusion weighted imaging (DWI), diffusion kurtosis imaging (DKI), and intravoxel incoherent motion (IVIM) were consecutively enrolled in this retrospective study. The histogram features of diffusion parameters, including apparent diffusion coefficient (ADC), diffusion coefficient (Dk), diffusion kurtosis (K), pure diffusion coefficient (D), pseudo-diffusion coefficient (DP), and perfusion fraction (FP) were analyzed. The Mann-Whitney U test was used for comparison between benign parotid gland tumors (BPGTs) and malignant parotid gland tumors (MPGTs). Receiver operating characteristic curve and logistic regression analysis were used to identify the differential diagnostic performance. The Spearman's correlation coefficient was used to analyze the correlation between diffusion parameters and Ki-67 labeling index.

RESULTS

For diffusion MRI, twenty-three histogram features of diffusion parameters showed significant differences between BPGTs and MPGTs (all P < 0.05). Compared with the DWI model, the IVIM model and combined model had better diagnostic specificity (58 %, 94 %, and 88 %, respectively; both corrected P < 0.001) and accuracy (64 %, 89 %, and 86 %, respectively; both corrected P = 0.006). The combined model was superior to the single DWI model with improved IDI (IDI improvement 0.25). Significant correlations were found between Ki-67 and ADCmean, Dkmean, Kmean, and Dmean (r = -0.57 to 0.53; all P < 0.05).

CONCLUSIONS

Whole-tumor histogram analysis of IVIM and combined diffusion model could further improve the diagnostic performance for differentiating BPGTs from MPGTs.

Radiomics Analysis in Characterization of Salivary Gland Tumors on MRI: A Systematic Review

Radiomics analysis can potentially characterize salivary gland tumors (SGTs) on magnetic resonance imaging (MRI). The procedures for radiomics analysis were various, and no consistent performances were reported. This review evaluated the methodologies and performances of studies using radiomics analysis to characterize SGTs on MRI. We systematically reviewed studies published until July 2023, which employed radiomics analysis to characterize SGTs on MRI. In total, 14 of 98 studies were eligible. Each study examined 23-334 benign and 8-56 malignant SGTs. Least absolute shrinkage and selection operator (LASSO) was the most common feature selection method (in eight studies). Eleven studies confirmed the stability of selected features using cross-validation or bootstrap. Nine classifiers were used to build models that achieved area under the curves (AUCs) of 0.74 to 1.00 for characterizing benign and malignant SGTs and 0.80 to 0.96 for characterizing pleomorphic adenomas and Warthin's tumors. Performances were validated using cross-validation, internal, and external datasets in four, six, and two studies, respectively. No single feature consistently appeared in the final models across the studies. No standardized procedure was used for radiomics analysis in characterizing SGTs on MRIs, and various models were proposed. The need for a standard procedure for radiomics analysis is emphasized.

Semiquantitative magnetic resonance imaging parameters for differentiating parotid pleomorphic adenoma from Warthin tumor

Background

Accurately distinguishing between pleomorphic adenoma (PA) and Warthin tumor (WT) is beneficial for their respective management. Preoperative magnetic resonance imaging (MRI) can provide valuable information due to its excellent soft tissue contrast. This study explored the value of semiquantitative contrast-enhanced MRI parameters in the differential diagnosis of PA and WT.

Methods

Data from 106 patients, 62 with PA and 44 with WT (confirmed by histopathology) were retrospectively and consecutively analyzed. The tumor-to-spinal cord contrast ratios (TSc-CR) based on the mean, maximum, and minimum signal intensity (T1-mean TSc-CR, T1-max TSc-CR, and T1-min TSc-CR, respectively) in the early and delayed phases were calculated on contrast-enhanced T1-weighted images as semiquantitative parameters, and then compared between PA and WT. Receiver operating characteristic (ROC) curve analysis and areas under the curve (AUCs) were used to determine the performance of these parameters in the differential diagnosis of PA from WT.

Results

Except T1-min TSc-CR in the early phase, all semiquantitative MRI parameters differed significantly between PA and WT (all P<0.05). T1-max TSc-CR showed higher sensitivity {70.45% [95% confidence interval (CI): 0.548-0.832]} and specificity [70.97% (95% CI: 0.581-0.818)] and had a higher AUC [0.707 (95% CI: 0.610-0.791)] in the early phase when using a cutoff value of 1.89. T1-max TSc-CR showed higher sensitivity [88.64% (95% CI: 0.754-0.962)], specificity [72.58% (95% CI: 0.598-0.831)], and AUC [0.854 (95% CI: 0.772-0.915)] in the delayed phase when using a cutoff value of 2.33. The sensitivity, specificity, and AUC were improved to 90.91% (95% CI: 0.783-0.975), 93.55% (95% CI: 0.843-0.982), and 0.960 (95% CI: 0.903-0.988), respectively, after combination of all semiquantitative parameters in the early and delayed phases. The two radiologists had excellent interobserver agreement on TSc-CRs [all interclass correlation coefficient (ICC) >0.75].

Conclusions

Semiquantitative parameters using TSc-CR are valuable in distinguishing PA from WT, and a combination of these parameters can improve the differential diagnostic efficiency.

The Role of an MRI-Based Radiomic Signature in Predicting Malignancy of Parotid Gland Tumors

The aim of this study was to assess the ability of MRI radiomic features to differentiate between benign parotid gland tumors (BPGT) and malignant parotid gland tumors (MPGT). This retrospective study included 93 patients who underwent MRI examinations of the head and neck region (78 patients presenting unique PGT, while 15 patients presented double PGT). A total of 108 PGT with histological confirmation were eligible for the radiomic analysis and were assigned to a training group (n = 83; 58 BPGT; 25 MPGT) and a testing group (n = 25; 16 BPGT; 9 MPGT). The radiomic features were extracted from 3D segmentations of the PGT on the T2-weighted and fat-saturated, contrast-enhanced T1-weighted images. Following feature reduction techniques, including LASSO regression analysis, a radiomic signature (RS) was built with five radiomic features. The RS presented a good diagnostic performance in differentiating between PGT, achieving an area under the curve (AUC) of 0.852 (p < 0.001) in the training set and 0.786 (p = 0.017) in the testing set. In both datasets, the RS proved to have lower values in the BPGT group as compared to MPGT group (p < 0.001 and p = 0.023, respectively). The multivariate analysis revealed that RS was independently associated with PGT malignancy, together with the ill-defined margin pattern (p = 0.031, p = 0.001, respectively). The complex model, using clinical data, MRI features and the RS, presented a higher diagnostic performance (AUC of 0.976) in comparison to the RS alone. MRI-based radiomic features could be considered potential additional imaging biomarkers able to discriminate between benign and malignant parotid gland tumors.

CT-based radiomics with various classifiers for histological differentiation of parotid gland tumors

Objective

This study assessed whether radiomics features could stratify parotid gland tumours accurately based on only noncontrast CT images and validated the best classifier of different radiomics models.

Methods

In this single-centre study, we retrospectively recruited 249 patients with a diagnosis of pleomorphic adenoma (PA), Warthin tumour (WT), basal cell adenoma (BCA) or malignant parotid gland tumours (MPGTs) from June 2020 to August 2022. Each patient was randomly classified into training and testing cohorts at a ratio of 7:3, and then, pairwise comparisons in different parotid tumour groups were performed. CT images were transferred to 3D-Slicer software and the region of interest was manually drawn for feature extraction. Feature selection methods were performed using the intraclass correlation coefficient, t test and least absolute shrinkage and selection operator. Five common classifiers, namely, random forest (RF), support vector machine (SVM), logistic regression (LR), K-nearest neighbours (KNN) and general Bayesian network (Gnb), were selected to build different radiomics models. The receiver operating characteristic curve, area under the curve (AUC), accuracy, sensitivity, specificity and F-1 score were used to assess the prediction performances of these models. The calibration of the model was calculated by the Hosmer–Lemeshow test. DeLong’s test was utilized for comparing the AUCs.

Results

The radiomics model based on the RF, SVM, Gnb, LR, LR and RF classifiers obtained the highest AUC in differentiating PA from MPGTs, WT from MPGTs, BCA from MPGTs, PA from WT, PA from BCA, and WT from BCA, respectively. Accordingly, the AUC and the accuracy of the model for each classifier were 0.834 and 0.71, 0.893 and 0.79, 0.844 and 0.79, 0.902 and 0.88, 0.602 and 0.68, and 0.861 and 0.94, respectively.

Conclusion

Our study demonstrated that noncontrast CT-based radiomics could stratify refined pathological types of parotid tumours well but could not sufficiently differentiate PA from BCA. Different classifiers had the best diagnostic performance for different parotid tumours. Our study findings add to the current knowledge on the differential diagnosis of parotid tumours.

Analysis of radiomic features derived from post-contrast T1-weighted images and apparent diffusion coefficient (ADC) maps for breast lesion evaluation: A retrospective study

INTRODUCTION

Breast cancer is the most common malignancy among women, and its diagnosis relies on medical imaging and the invasive, uncomforted biopsy. Recent advances in quantitative imaging and specifically the application of radiomics has proved to be a very promising technique, facilitating both diagnosis and therapy. The purpose of this study is to assess radiomic features derived from post-contrast T1w Magnetic Resonance Imaging (MRI) sequences and Apparent Diffusion Coefficient (ADC) maps for the evaluation of breast pathologies.

METHODS

MRI data from 52 women were retrospectively reviewed, involving 54 breast lesions, both malignant and benign. Diffusion Weighted Imaging (DWI) was applied as a standard MRΙ protocol, including dynamic contrast-enhanced (DCE) MRΙ in all cases. All patients were examined on a 1.5T MRI scanner, and 216 features were initially extracted from DCE-MRI images. Histological analysis of the breast lesions was performed, and a comparative analysis of the results was carried out to assess the accuracy of the method.

RESULTS

Following surgery and histological analysis, 30 lesions were found to be malignant and 24 benign. Implementation of a Machine Learning (ML) classification algorithm with 5-fold cross-validation resulted in a sensitivity of 70%, specificity of 66%, Negative Predictive Value of 82% and overall accuracy of 67% in differentiating malignancy from benevolence.

CONCLUSION

Texture analysis and ML methodology based on the first post-contrast dynamic sequences and ADC maps may be employed to differentiate between malignant and benign breast lesions, offering a promising new tool for diagnostic analysis.

IMPLICATIONS FOR PRACTICE

The results of this study will enhance knowledge around application and performance of radiomics in breast MRI, thus helping MRI radiographers who use AI-enabled technologies to better delineate the pros and cons of these procedures.

Enhanced CT-based texture analysis and radiomics score for differentiation of pleomorphic adenoma, basal cell adenoma, and Warthin tumor of the parotid gland

OBJECTIVE

To evaluate the diagnostic performance of computed tomography (CT) radiomics analysis for differentiating pleomorphic adenoma (PA), Warthin tumor (WT), and basal cell adenoma (BCA).

METHODS

A total of 189 patients with PA (n = 112), WT (n = 53) and BCA (n = 24) were divided into a training set (n = 133) and a test set (n = 56). The radiomics features were extracted from plain CT and contrast-enhanced CT images. After dimensionality reduction, plain CT, multiphase-enhanced CT, integrated radiomics signature models and radiomics score (Rad-score) were established and calculated. The receiver operating characteristic (ROC) curve analysis was taken for the assessment of the model performance, and then comparison was conducted among these models. Decision curve analysis (DCA) was adopted to assess the clinical benefits of the models. Diagnostic performances including sensitivity, specificity, and accuracy of the radiologists were evaluated.

RESULTS

Seven, nine, fourteen, and fourteen optimal features were used to constructed plain scan, arterial phase, venous phase, and integrated radiomics signature models, respectively. ROC analysis showed these four models were able to differentiate PA from BCA and WT, with the area under the ROC curve (AUC) values of 0.79, 0.90, 0.87, and 0.94 in the training set, and 0.79, 0.89, 0.86, and 0.94 in the test set, respectively. The integrated model had better diagnostic performance than single-phase radiomics model, but it had similar diagnostic performance to that of the radiomics model based on the arterial phase (p > 0.05). The sensitivity, specificity, and accuracy in the diagnosis of PA were 0.86, 0.46, and 0.70 for the non-subspecialized radiologist and 0.88, 0.77, and 0.84 for the subspecialized radiologist, respectively. Six venous phase parameters were finally selected in differentiating WT from BCA. The predictive effect of the model was favorable, with AUC value of 0.95, sensitivity of 0.96, specificity of 0.83, and accuracy of 0.92. The sensitivity, specificity, and accuracy in the diagnosis between WT and BCA were 0.26, 0.87, and 0.45 for the non-subspecialized radiologist and 0.85, 0.58, and 0.77 for the subspecialized radiologist, respectively.

CONCLUSION

The CT-based radiomics analysis showed favorable predictive performance for differentiating PA, WT, and BCA, thus may be helpful in the clinical decision-making.

Value of T2-weighted-based radiomics model in distinguishing Warthin tumor from pleomorphic adenoma of the parotid

OBJECTIVES

The differentiation of Warthin tumor and pleomorphic adenoma before treatment is crucial for clinical strategies. The aim of this study was to develop and test a T2-weighted-based radiomics model for differentiating pleomorphic adenoma from Warthin tumor of the parotid gland.

METHODS

A total of 117 patients, including 61 cases of Warthin tumor and 56 cases of pleomorphic adenoma, were retrospectively enrolled from two centers between January 2010 and June 2022. The training set included 82 cases, and the validation set included 35 cases. From T2-weighted images, 971 radiomics features were extracted. Seven radiomics features remained after a two-step selection process. We used the seven radiomics features and clinical factors through multivariable logistic regression to build radiomics and clinical models, respectively. A radiomics-clinical model was also built that combined the independent clinical predictors with the radiomics features. Through ROC curves, the three models were evaluated and compared.

RESULTS

In the radiomics model, AUCs were 0.826 and 0.796 in training and validation sets, respectively. In the clinical model, the AUCs were 0.923 and 0.926 in the training and validation sets, respectively. Decision curve analysis revealed that the radiomics-clinical model had the best diagnostic performance for distinguishing Warthin tumor from pleomorphic adenoma of the parotid gland (AUC = 0.962 and 0.934 for the training and validation sets, respectively).

CONCLUSION

The radiomics-clinical model performed well in differentiating pleomorphic adenoma from Warthin tumor of the parotid gland.

KEY POINTS

• The clinical model outperformed the radiomics model in distinguishing pleomorphic adenoma from Warthin tumor of the parotid gland. • The radiomics features extracted from T2-weighted images could help differentiate pleomorphic adenoma from Warthin tumor of the parotid gland. • The radiomics-clinical model was superior to the radiomics and the clinical models for differentiating pleomorphic adenoma from Warthin tumor of the parotid gland.

The Role of Radiomics in Salivary Gland Imaging: A Systematic Review and Radiomics Quality Assessment

Background: Radiomics of salivary gland imaging can support clinical decisions in different clinical scenarios, such as tumors, radiation-induced xerostomia and sialadenitis. This review aims to evaluate the methodological quality of radiomics studies on salivary gland imaging. Material and Methods: A systematic search was performed, and the methodological quality was evaluated using the radiomics quality score (RQS). Subgroup analyses according to the first author's professional role (medical or not medical), journal type (radiological journal or other) and the year of publication (2021 or before) were performed. The correlation of RQS with the number of patients was calculated. Results: Twenty-three articles were included (mean RQS 11.34 ± 3.68). Most studies well-documented the imaging protocol (87%), while neither prospective validations nor cost-effectiveness analyses were performed. None of the included studies provided open-source data. A statistically significant difference in RQS according to the year of publication was found (p = 0.009), with papers published in 2021 having slightly higher RQSs than older ones. No differences according to journal type or the first author's professional role were demonstrated. A moderate relationship between the overall RQS and the number of patients was found. Conclusions: Radiomics application in salivary gland imaging is increasing. Although its current clinical applicability can be affected by the somewhat inadequate quality of the papers, a significant improvement in radiomics methodologies has been demonstrated in the last year.

Radiomics for Discriminating Benign and Malignant Salivary Gland Tumors; Which Radiomic Feature Categories and MRI Sequences Should Be Used?

The lack of a consistent MRI radiomic signature, partly due to the multitude of initial feature analyses, limits the widespread clinical application of radiomics for the discrimination of salivary gland tumors (SGTs). This study aimed to identify the optimal radiomics feature category and MRI sequence for characterizing SGTs, which could serve as a step towards obtaining a consensus on a radiomics signature. Preliminary radiomics models were built to discriminate malignant SGTs (n = 34) from benign SGTs (n = 57) on T1-weighted (T1WI), fat-suppressed (FS)-T2WI and contrast-enhanced (CE)-T1WI images using six feature categories. The discrimination performances of these preliminary models were evaluated using 5-fold-cross-validation with 100 repetitions and the area under the receiver operating characteristic curve (AUC). The differences between models’ performances were identified using one-way ANOVA. Results show that the best feature categories were logarithm for T1WI and CE-T1WI and exponential for FS-T2WI, with AUCs of 0.828, 0.754 and 0.819, respectively. These AUCs were higher than the AUCs obtained using all feature categories combined, which were 0.750, 0.707 and 0.774, respectively (p < 0.001). The highest AUC (0.846) was obtained using a combination of T1WI + logarithm and FS-T2WI + exponential features, which reduced the initial features by 94.0% (from 1015 × 3 to 91 × 2). CE-T1WI did not improve performance. Using one feature category rather than all feature categories combined reduced the number of initial features without compromising radiomic performance.

Magnetic resonance imaging of parotid gland tumors: a pictorial essay

Imaging of parotid gland tumors is challenging due to the wide variety of differential diagnoses. Malignant parotid tumors can have very similar features to benign ones, such as slow growth and displacement instead of infiltration of neighboring structures. Malignant and benign tumors may therefore not be clinically distinguishable. Correct characterization of parotid tumors (i.e., benign or malignant) determines preoperative treatment planning and is important in optimizing the individualized surgical plan. Magnetic resonance imaging (MRI) is the imaging modality of choice for evaluation of suspected parotid gland lesions and differentiation between benign and malignant lesions. Certain conventional MRI features can suggest whether a mass is more likely to be a benign or low-grade malignancy or a high-grade malignancy and adding diffusion-weighted imaging or advanced MRI techniques like perfusion can aid in this distinction. Morphological features seen on MRI, such as low signal on T2-w, infiltrative changes or ill-defined margins, change over time and diffusion restriction can point to the malignant nature of the lesion. MRI is useful for detection and localization of the lesion(s), and associated findings like perineural spread of tumor, lymph node involvement and infiltrative changes of the surrounding tissues. In this pictorial essay, we present selected images of a variety of benign and malignant parotid tumors and emphasize the MRI features that may be useful in their characterization.

Differentiation of Benign From Malignant Parotid Gland Tumors Using Conventional MRI Based on Radiomics Nomogram

Objectives

We aimed to develop and validate radiomic nomograms to allow preoperative differentiation between benign- and malignant parotid gland tumors (BPGT and MPGT, respectively), as well as between pleomorphic adenomas (PAs) and Warthin tumors (WTs).

Materials and Methods

This retrospective study enrolled 183 parotid gland tumors (68 PAs, 62 WTs, and 53 MPGTs) and divided them into training (n = 128) and testing (n = 55) cohorts. In total, 2553 radiomics features were extracted from fat-saturated T2-weighted images, apparent diffusion coefficient maps, and contrast-enhanced T1-weighted images to construct single-, double-, and multi-sequence combined radiomics models, respectively. The radiomics score (Rad-score) was calculated using the best radiomics model and clinical features to develop the radiomics nomogram. The receiver operating characteristic curve and area under the curve (AUC) were used to assess these models, and their performances were compared using DeLong’s test. Calibration curves and decision curve analysis were used to assess the clinical usefulness of these models.

Results

The multi-sequence combined radiomics model exhibited better differentiation performance (BPGT vs. MPGT, AUC=0.863; PA vs. MPGT, AUC=0.929; WT vs. MPGT, AUC=0.825; PA vs. WT, AUC=0.927) than the single- and double sequence radiomics models. The nomogram based on the multi-sequence combined radiomics model and clinical features attained an improved classification performance (BPGT vs. MPGT, AUC=0.907; PA vs. MPGT, AUC=0.961; WT vs. MPGT, AUC=0.879; PA vs. WT, AUC=0.967).

Conclusions

Radiomics nomogram yielded excellent diagnostic performance in differentiating BPGT from MPGT, PA from MPGT, and PA from WT.

Machine learning-based radiomics for histological classification of parotid tumors using morphological MRI: a comparative study

OBJECTIVES

To evaluate the effectiveness of machine learning models based on morphological magnetic resonance imaging (MRI) radiomics in the classification of parotid tumors.

METHODS

In total, 298 patients with parotid tumors were randomly assigned to a training and test set at a ratio of 7:3. Radiomics features were extracted from the morphological MRI images and screened using the Select K Best and LASSO algorithm. Three-step machine learning models with XGBoost, SVM, and DT algorithms were developed to classify the parotid neoplasms into four subtypes. The ROC curve was used to measure the performance in each step. Diagnostic confusion matrices of these models were calculated for the test cohort and compared with those of the radiologists.

RESULTS

Six, twelve, and eight optimal features were selected in each step of the three-step process, respectively. XGBoost produced the highest area under the curve (AUC) for all three steps in the training cohort (0.857, 0.882, and 0.908, respectively), and for the first step in the test cohort (0.826), but produced slightly lower AUCs than SVM in the latter two steps in the test cohort (0.817 vs. 0.833, and 0.789 vs. 0.821, respectively). The total accuracies of XGBoost and SVM in the confusion matrices (70.8% and 59.6%) outperformed those of DT and the radiologist (46.1% and 49.2%).

CONCLUSION

This study demonstrated that machine learning models based on morphological MRI radiomics might be an assistive tool for parotid tumor classification, especially for preliminary screening in absence of more advanced scanning sequences, such as DWI.

KEY POINTS

• Machine learning algorithms combined with morphological MRI radiomics could be useful in the preliminary classification of parotid tumors. • XGBoost algorithm performed better than SVM and DT in subtype differentiation of parotid tumors, while DT seemed to have a poor validation performance. • Using morphological MRI only, the XGBoost and SVM algorithms outperformed radiologists in the four-type classification task for parotid tumors, thus making these models a useful assistant diagnostic tool in clinical practice.

Apparent Diffusion Coefficient Map–Based Radiomics Features for Differential Diagnosis of Pleomorphic Adenomas and Warthin Tumors From Malignant Tumors

Purpose

The magnetic resonance imaging (MRI) findings may overlap due to the complex content of parotid gland tumors and the differentiation level of malignant tumor (MT); consequently, patients may undergo diagnostic lobectomy. This study assessed whether radiomics features could noninvasively stratify parotid gland tumors accurately based on apparent diffusion coefficient (ADC) maps.

Methods

This study examined diffusion-weighted imaging (DWI) obtained with echo planar imaging sequences. Eighty-eight benign tumors (BTs) [54 pleomorphic adenomas (PAs) and 34 Warthin tumors (WTs)] and 42 MTs of the parotid gland were enrolled. Each case was randomly divided into training and testing cohorts at a ratio of 7:3 and then was compared with each other, respectively. ADC maps were digitally transferred to ITK SNAP (www.itksnap.org). The region of interest (ROI) was manually drawn around the whole tumor margin on each slice of ADC maps. After feature extraction, the Synthetic Minority Oversampling TEchnique (SMOTE) was used to remove the unbalance of the training dataset. Then, we applied the normalization process to the feature matrix. To reduce the similarity of each feature pair, we calculated the Pearson correlation coefficient (PCC) value of each feature pair and eliminated one of them if the PCC value was larger than 0.95. Then, recursive feature elimination (RFE) was used to process feature selection. After that, we used linear discriminant analysis (LDA) as the classifier. Receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic performance of the ADC.

Results

The LDA model based on 13, 8, 3, and 1 features can get the highest area under the ROC curve (AUC) in differentiating BT from MT, PA from WT, PA from MT, and WT from MT on the validation dataset, respectively. Accordingly, the AUC and the accuracy of the model on the testing set achieve 0.7637 and 73.17%, 0.925 and 92.31%, 0.8077 and 75.86%, and 0.5923 and 65.22%, respectively.

Conclusion

The ADC-based radiomics features may be used to assist clinicians for differential diagnosis of PA and WT from MTs.

The Diagnostic Value of Ultrasound-Based Deep Learning in Differentiating Parotid Gland Tumors

Objectives. Evidence suggests that about 80% of all salivary gland tumors involve the parotid glands, with approximately 20% of parotid gland tumors (PGTs) being malignant. Discriminating benign and malignant parotid gland lesions preoperatively is vital for selecting the appropriate treatment strategy. This study explored the diagnostic performance of deep learning system for discriminating benign and malignant PGTs in ultrasonography images and compared it with radiologists. Methods. A total of 251 consecutive patients with surgical resection and proven parotid gland malignant or benign tumors who underwent preoperative ultrasound examinations were enrolled in this study between January 2014 and November 2020. Next, we compared the diagnostic accuracy of deep learning methods (ViT-B\16, EfficientNetB3, DenseNet121, and ResNet50) and radiologists in parotid gland tumor. In addition, the area under the curve (AUC), specificity, sensitivity, positive predictive value, and negative predictive value were calculated. Results. Among the 251 patients, 176/251 were the training set, whereas 75/251 were the validation set. Results showed that 74/251 patients had malignant tumor. Deep learning models achieved good performance in differentiating benign from malignant tumors, with the diagnostic accuracy and AUCs of ViT-B\16, EfficientNetB3, DenseNet121, and ResNet50 model being 81% and 0.81, 80% and 0.82, 77% and 0.81, and 79% and 0.80, respectively. On the other hand, the diagnostic accuracy and AUCs of radiologists were 77%-81% and 0.68-0.75, respectively. It was evident that the diagnostic accuracy of deep learning methods was higher than that of inexperienced radiologists, but there was no significant difference between deep learning methods and experienced radiologists. Conclusions. This study shows that the deep learning system can be used for diagnosing parotid tumors. The findings also suggest that the deep learning system may improve the diagnosis performance of inexperienced radiologists.

A nomogram based on clinical information, conventional ultrasound and radiomics improves prediction of malignant parotid gland lesions

Although conventional ultrasound (CUS) allows for clear detection of parotid gland lesions (PGLs), it fails to accurately provide benign-malignant differentiation due to overlapping morphological features. Radiomics is capable of processing large-quantity volume of data hidden in CUS image undiscovered by naked eyes. The aim was to explore the potential of CUS-based radiomics score (Rad-score) in distinguishing benign (BPGLs) and malignant PGLs (MPGLs). A consecutive of 281 PGLs (197 in training set and 84 in test set) with definite pathological confirmation was retrospectively enrolled. 1465 radiomics features were extracted from CUS images and Rad-score was constructed by using Least Absolute Shrinkage and Selection Operator (LASSO) algorithm. Different nomogram models, including clinic-radiomics (Clin + Rad-score), CUS-clinic (CUS + Clin) and combined CUS-clinic-radiomics (CUS + Clin + Rad-score), were built using logistic regression. The diagnostic performance of different models were calculated and compared by area under receiver operating curve (AUC) and corresponding sensitivity and specificity. Finally, 26 radiomics features were independent signatures for predicting MPGLs, with MPGLs having higher Rad-scores in both cohorts (both P < 0.05). In the test population, CUS + Clin + Rad-score obtained an excellent diagnostic result, with significantly higher AUC value (AUC = 0.91) when compared to that of CUS + Clin (AUC = 0.84) and Clin + Rad-score (AUC = 0.74), respectively (both P < 0.05). In addition, the sensitivity of this combined model was higher than that of single Rad-score model (100.00% vs. 71.43%, P = 0.031) without compromising the specificity value (82.86% vs. 88.57%, P = 0.334). The calibration curve and decision curve analysis also indicated the clinical effectiveness of the proposed combined nomogram. The combined CUS-clinic-radiomics model may help improve the discrimination of BPGLs from MPGLs.

Radiomics and deep learning approach to the differential diagnosis of parotid gland tumors

PURPOSE OF REVIEW

Advances in computer technology and growing expectations from computer-aided systems have led to the evolution of artificial intelligence into subsets, such as deep learning and radiomics, and the use of these systems is revolutionizing modern radiological diagnosis. In this review, artificial intelligence applications developed with radiomics and deep learning methods in the differential diagnosis of parotid gland tumors (PGTs) will be overviewed.

RECENT FINDINGS

The development of artificial intelligence models has opened new scenarios owing to the possibility of assessing features of medical images that usually are not evaluated by physicians. Radiomics and deep learning models come to the forefront in computer-aided diagnosis of medical images, even though their applications in the differential diagnosis of PGTs have been limited because of the scarcity of data sets related to these rare neoplasms. Nevertheless, recent studies have shown that artificial intelligence tools can classify common PGTs with reasonable accuracy.

SUMMARY

All studies aimed at the differential diagnosis of benign vs. malignant PGTs or the identification of the commonest PGT subtypes were identified, and five studies were found that focused on deep learning-based differential diagnosis of PGTs. Data sets were created in three of these studies with MRI and in two with computed tomography (CT). Additional seven studies were related to radiomics. Of these, four were on MRI-based radiomics, two on CT-based radiomics, and one compared MRI and CT-based radiomics in the same patients.

Foundations of Lesion Detection Using Machine Learning in Clinical Neuroimaging

This chapter describes technical considerations and current and future clinical applications of lesion detection using machine learning in the clinical setting. Lesion detection is central to neuroradiology and precedes all further processes which include but are not limited to lesion characterization, quantification, longitudinal disease assessment, prognosis, and prediction of treatment response. A number of machine learning algorithms focusing on lesion detection have been developed or are currently under development which may either support or extend the imaging process. Examples include machine learning applications in stroke, aneurysms, multiple sclerosis, neuro-oncology, neurodegeneration, and epilepsy.

Differentiating low and high grade mucoepidermoid carcinoma of the salivary glands using CT radiomics

Background

The purpose of this study is to determine if Haralick texture analysis on CT imaging of mucoepidermoid carcinomas (MEC) can differentiate low-grade and high-grade tumors.

Methods

A retrospective review of 18 patients with MEC of the salivary glands, corresponding CT imaging and pathology report was performed. Tumors were manually segmented and image analysis was performed to calculate radiomic features. Radiomic features were compared between low-grade and high-grade MEC. A multivariable logistic regression model and receiver operating characteristic analysis was performed.

Results

A total of 18 patients (mean age, 51, range 9-83 years, 8 men and 10 women) were included. Nine patients had low-grade pathology and nine patients had high-grade pathology. Of the 18 cases, 7 (39%) occurred in the parotid gland and 11 (61%) occurred in minor salivary glands. No individual feature was significantly different between low-grade and high-grade MEC. A logistic regression model including surface regularity, energy and information measure II of correlation was performed and was able to predict high-grade MEC accurately (sensitivity 89%, specificity 68%). The area under the receiver operating characteristic curve was 0.802.

Conclusions

High-grade MEC tend to have a low energy, high correlation texture as well as surface irregularity. Together, these three features may comprise a tumor phenotype that is able to predict high-grade pathology in MECs.

MRI-Based Radiomics to Differentiate between Benign and Malignant Parotid Tumors With External Validation

Background

The differentiation between benign and malignant parotid lesions is crucial to defining the treatment plan, which highly depends on the tumor histology. We aimed to evaluate the role of MRI-based radiomics using both T2-weighted (T2-w) images and Apparent Diffusion Coefficient (ADC) maps in the differentiation of parotid lesions, in order to develop predictive models with an external validation cohort.

Materials and Methods

A sample of 69 untreated parotid lesions was evaluated retrospectively, including 37 benign (of which 13 were Warthin’s tumors) and 32 malignant tumors. The patient population was divided into three groups: benign lesions (24 cases), Warthin’s lesions (13 cases), and malignant lesions (32 cases), which were compared in pairs. First- and second-order features were derived for each lesion. Margins and contrast enhancement patterns (CE) were qualitatively assessed. The model with the final feature set was achieved using the support vector machine binary classification algorithm.

Results

Models for discriminating between Warthin’s and malignant tumors, benign and Warthin’s tumors and benign and malignant tumors had an accuracy of 86.7%, 91.9% and 80.4%, respectively. After the feature selection process, four parameters for each model were used, including histogram-based features from ADC and T2-w images, shape-based features and types of margins and/or CE. Comparable accuracies were obtained after validation with the external cohort.

Conclusions

Radiomic analysis of ADC, T2-w images, and qualitative scores evaluating margins and CE allowed us to obtain good to excellent diagnostic accuracies in differentiating parotid lesions, which were confirmed with an external validation cohort.