Can we use radiomics in ultrasound imaging? Impact of preprocessing on feature repeatability

Enhancing the Clinical Utility of Radiomics: Addressing the Challenges of Repeatability and Reproducibility in CT and MRI

Radiomics, which integrates the comprehensive characterization of imaging phenotypes with machine learning algorithms, is increasingly recognized for its potential in the diagnosis and prognosis of oncological conditions. However, the repeatability and reproducibility of radiomic features are critical challenges that hinder their widespread clinical adoption. This review aims to address the paucity of discussion regarding the factors that influence the reproducibility and repeatability of radiomic features and their subsequent impact on the application of radiomic models. We provide a synthesis of the literature on the repeatability and reproducibility of CT/MR-based radiomic features, examining sources of variation, the number of reproducible features, and the availability of individual feature repeatability indices. We differentiate sources of variation into random effects, which are challenging to control but can be quantified through simulation methods such as perturbation, and biases, which arise from scanner variability and inter-reader differences and can significantly affect the generalizability of radiomic model performance in diverse settings. Four suggestions for repeatability and reproducibility studies are suggested: (1) detailed reporting of variation sources, (2) transparent disclosure of calculation parameters, (3) careful selection of suitable reliability indices, and (4) comprehensive reporting of reliability metrics. This review underscores the importance of random effects in feature selection and harmonizing biases between development and clinical application settings to facilitate the successful translation of radiomic models from research to clinical practice.

Involution Transformer Based U-Net for Landmark Detection in Ultrasound Images for Diagnosis of Infantile DDH

The B-mode ultrasound based computer-aided diagnosis (CAD) has demonstrated its effectiveness for diagnosis of Developmental Dysplasia of the Hip (DDH) in infants, which can conduct the Graf's method by detecting landmarks in hip ultrasound images. However, it is still necessary to explore more valuable information around these landmarks to enhance feature representation for improving detection performance in the detection model. To this end, a novel Involution Transformer based U-Net (IT-UNet) network is proposed for hip landmark detection. The IT-UNet integrates the efficient involution operation into Transformer to develop an Involution Transformer module (ITM), which consists of an involution attention block and a squeeze-and-excitation involution block. The ITM can capture both the spatial-related information and long-range dependencies from hip ultrasound images to effectively improve feature representation. Moreover, an Involution Downsampling block (IDB) is developed to alleviate the issue of feature loss in the encoder modules, which combines involution and convolution for the purpose of downsampling. The experimental results on two DDH ultrasound datasets indicate that the proposed IT-UNet achieves the best landmark detection performance, indicating its potential applications.

Impact of Preprocessing Parameters in Medical Imaging-Based Radiomic Studies: A Systematic Review

BACKGROUND

Lately, radiomic studies featuring the development of a signature to use in prediction models in diagnosis or prognosis outcomes have been increasingly published. While the results are shown to be promising, these studies still have many pitfalls and limitations. One of the main issues of these studies is that radiomic features depend on how the images are preprocessed before their computation. Since, in widely known and used software for radiomic features calculation, it is possible to set these preprocessing parameters before the calculation of the radiomic feature, there are ongoing studies assessing the stability and repeatability of radiomic features to find the most suitable preprocessing parameters for every used imaging modality.

MATERIALS AND METHODS

We performed a comprehensive literature search using four electronic databases: PubMed, Cochrane Library, Embase, and Scopus. Mesh terms and free text were modeled in search strategies for databases. The inclusion criteria were studies where preprocessing parameters' influence on feature values and model predictions was addressed. Records lacking information on image acquisition parameters were excluded, and any eligible studies with full-text versions were included in the review process, while conference proceedings and monographs were disregarded. We used the QUADAS-2 (Quality Assessment of Diagnostic Accuracy Studies 2) tool to investigate the risk of bias. We synthesized our data in a table divided by the imaging modalities subgroups.

RESULTS

After applying the inclusion and exclusion criteria, we selected 43 works. This review examines the impact of preprocessing parameters on the reproducibility and reliability of radiomic features extracted from multimodality imaging (CT, MRI, CBCT, and PET/CT). Standardized preprocessing is crucial for consistent radiomic feature extraction. Key preprocessing steps include voxel resampling, normalization, and discretization, which influence feature robustness and reproducibility. In total, 44% of the included works studied the effects of an isotropic voxel resampling, and most studies opted to employ a discretization strategy. From 2021, several studies started selecting the best set of preprocessing parameters based on models' best performance. As for comparison metrics, ICC was the most used in MRI studies in 58% of the screened works.

CONCLUSIONS

From our work, we highlighted the need to harmonize the use of preprocessing parameters and their values, especially in light of future studies of prospective studies, which are still lacking in the current literature.

[Preoperative Evaluation of Cervical Lymph Node Metastasis in Patients With Hashimoto's Thyroiditis Combined With Thyroid Papillary Carcinoma Using Machine Learning and Radiomics-Based Features: A Preliminary Study]

Objective

To analyze the radiomic and clinical features extracted from 2D ultrasound images of thyroid tumors in patients with Hashimoto's thyroiditis (HT) combined with papillary thyroid carcinoma (PTC) using machine learning (ML) models, and to explore the diagnostic performance of the method in making preoperative noninvasive identification of cervical lymph node metastasis (LNM).

Methods

A total of 528 patients with HT combined with PTC were enrolled and divided into two groups based on their pathological results of the presence or absence of LNM. The groups were subsequently designated the With LNM Group and the Without LNM Group. Three ultrasound doctors independently delineated the regions of interest and extracted radiomic features. Two modes, radiomic features and radiomics-clinical features, were used to construct random forest (RF), support vector machine (SVM), LightGBM, K-nearest neighbor (KNN), and XGBoost models. The performance of these five ML models in the two modes was evaluated by the receiver operating characteristic (ROC) curves on the test dataset, and SHapley Additive exPlanations (SHAP) was used for model visualization.

Results

All five ML models showed good performance, with area under the ROC curve (AUC) ranging from 0.798 to 0.921. LightGBM and XGBoost demonstrated the best performance, outperforming the other models (P<0.05). The ML models constructed with radiomics-clinical features performed better than those constructed using only radiomic features (P<0.05). The SHAP visualization of the best-performing models indicated that the anteroposterior diameter, superoinferior diameter, original_shape_VoxelVolume, age, wavelet-LHL_firstorder_10Percentile, and left-to-right diameter had the most significant effect on the LightGBM model. On the other hand, the superoinferior diameter, anteroposterior diameter, left-to-right diameter, original_shape_VoxelVolume, original_firstorder_InterquartileRange, and age had the most significant effect on the XGBoost model.

Conclusion

ML models based on radiomics and clinical features can accurately evaluate the cervical lymph node status in patients with HT combined with PTC. Among the 5 ML models, LightGBM and XGBoost demonstrate the best evaluation performance.

Predicting Ki-67 expression levels in breast cancer using radiomics-based approaches on digital breast tomosynthesis and ultrasound

Purpose

To construct and validate radiomics models that utilize ultrasound (US) and digital breast tomosynthesis (DBT) images independently and in combination to non-invasively predict the Ki-67 status in breast cancer.

Materials and methods

149 breast cancer women who underwent DBT and US scans were retrospectively enrolled from June 2018 to August 2023 in total. Radiomics features were acquired from both the DBT and US images, then selected and reduced in dimensionality using several screening approaches. Establish radiomics models based on DBT, and US separately and combined. The area under the receiver operating characteristic curve (AUC), accuracy, specificity, and sensitivity were utilized to validate the predictive ability of the models. The decision curve analysis (DCA) was used to evaluate the clinical applicability of the models. The output of the classifier with the best AUC performance was converted into Rad-score and was regarded as Rad-Score model. A nomogram was constructed using the logistic regression method, integrating the Rad-Score and clinical factors. The model's stability was assessed through AUC, calibration curves, and DCA.

Results

Support vector machine (SVM), logistic regression (LR), and random forest (RF) were trained to establish radiomics models with the selected features, with SVM showing optimal results. The AUC values for three models (US_SVM, DBT_SVM, and merge_SVM) were 0.668, 0.704, and 0.800 respectively. The DeLong test indicated a notable disparity in the area under the curve (AUC) between merge_SVM and US_SVM (p = 0.048), while there was no substantial variability between merge_SVM and DBT_SVM (p = 0.149). The DCA curve indicates that merge_SVM is superior to unimodal models in predicting high Ki-67 level, showing more clinical values. The nomogram integrating Rad-Score with tumor size obtained the better performance in test set (AUC: 0.818) and had more clinical net.

Conclusion

The fusion radiomics model performed better in predicting the Ki-67 expression level of breast carcinoma, but the gain effect is limited; thus, DBT is preferred as a preoperative diagnosis mode when resources are limited. Nomogram offers predictive advantages over other methods and can be a valuable tool for predicting Ki-67 levels in BC.

Radiomics-Based Analysis of Intestinal Ultrasound Images for Inflammatory Bowel Disease: A Feasibility Study

Background

The increasing adoption of intestinal ultrasound (IUS) for monitoring inflammatory bowel diseases (IBD) by IBD providers has uncovered new challenges regarding standardized image interpretation and limitations as a research tool. Artificial intelligence approaches can help address these challenges. We aim to determine the feasibility of radiomic analysis of IUS images and to determine if a radiomics-based classification model can accurately differentiate between normal and abnormal IUS images. We will also compare the radiomic-based model's performance to a convolutional neural network (CNN)-based classification model to understand which method is more effective for extracting meaningful information from IUS images.

Methods

Retrospectively analyzing IUS images obtained during routine outpatient visits, we developed and tested radiomic-based and CNN-based models to distinguish between normal and abnormal images, with abnormal images defined as bowel wall thickness > 3 mm or bowel hyperemia with modified Limberg score ≥ 1 (both are surrogate markers for inflammation). Model performances were measured by area under the receiver operator curve (AUC).

Results

For this feasibility study, 125 images (33% abnormal) were analyzed. A radiomic-based model using XG boost yielded the best classifier model with average test AUC 0.98%, 93.8% sensitivity, 93.8% specificity, and 93.7% accuracy. The CNN-based classification model yielded an average testing AUC of 0.75.

Conclusions

Radiomic analysis of IUS images is feasible, and a radiomic-based classification model could accurately differentiate abnormal from normal images. Our findings establish methods to facilitate future radiomic-based IUS studies that can help standardize image interpretation and expand IUS research capabilities.

Can multi-modal radiomics using pretreatment ultrasound and tomosynthesis predict response to neoadjuvant systemic treatment in breast cancer?

OBJECTIVES

Response assessment to neoadjuvant systemic treatment (NAST) to guide individualized treatment in breast cancer is a clinical research priority. We aimed to develop an intelligent algorithm using multi-modal pretreatment ultrasound and tomosynthesis radiomics features in addition to clinical variables to predict pathologic complete response (pCR) prior to the initiation of therapy.

METHODS

We used retrospective data on patients who underwent ultrasound and tomosynthesis before starting NAST. We developed a support vector machine algorithm using pretreatment ultrasound and tomosynthesis radiomics features in addition to patient and tumor variables to predict pCR status (ypT0 and ypN0). Findings were compared to the histopathologic evaluation of the surgical specimen. The main outcome measures were area under the curve (AUC) and false-negative rate (FNR).

RESULTS

We included 720 patients, 504 in the development set and 216 in the validation set. Median age was 51.6 years and 33.6% (242 of 720) achieved pCR. The addition of radiomics features significantly improved the performance of the algorithm (AUC 0.72 to 0.81; p = 0.007). The FNR of the multi-modal radiomics and clinical algorithm was 6.7% (10 of 150 with missed residual cancer). Surface/volume ratio at tomosynthesis and peritumoral entropy characteristics at ultrasound were the most relevant radiomics. Hormonal receptors and HER-2 status were the most important clinical predictors.

CONCLUSION

A multi-modal machine learning algorithm with pretreatment clinical, ultrasound, and tomosynthesis radiomics features may aid in predicting residual cancer after NAST. Pending prospective validation, this may facilitate individually tailored NAST regimens.

CLINICAL RELEVANCE STATEMENT

Multi-modal radiomics using pretreatment ultrasound and tomosynthesis showed significant improvement in assessing response to NAST compared to an algorithm using clinical variables only. Further prospective validation of our findings seems warranted to enable individualized predictions of NAST outcomes.

KEY POINTS

• We proposed a multi-modal machine learning algorithm with pretreatment clinical, ultrasound, and tomosynthesis radiomics features to predict response to neoadjuvant breast cancer treatment. • Compared with the clinical algorithm, the AUC of this integrative algorithm is significantly higher. • Used prior to the initiative of therapy, our algorithm can identify patients who will experience pathologic complete response following neoadjuvant therapy with a high negative predictive value.

Diagnostic Performance of Dynamic Contrast-Enhanced 3T MR Imaging for Characterization of Orbital Lesions: Validation in a Large Prospective Study

BACKGROUND AND PURPOSE

Orbital lesions are rare but serious. Their characterization remains challenging. Diagnosis is based on biopsy or surgery, which implies functional risks. It is necessary to develop noninvasive diagnostic tools. The goal of this study was to evaluate the diagnostic performance of dynamic contrast-enhanced MR imaging at 3T when distinguishing malignant from benign orbital tumors on a large prospective cohort.

MATERIALS AND METHODS

This institutional review board-approved prospective single-center study enrolled participants presenting with an orbital lesion undergoing a 3T MR imaging before surgery from December 2015 to May 2021. Morphologic, diffusion-weighted, and dynamic contrast-enhanced MR images were assessed by 2 readers blinded to all data. Univariable and multivariable analyses were performed. To assess diagnostic performance, we used the following metrics: area under the curve, sensitivity, and specificity. Histologic analysis, obtained through biopsy or surgery, served as the criterion standard for determining the benign or malignant status of the tumor.

RESULTS

One hundred thirty-one subjects (66/131 [50%] women and 65/131 [50%] men; mean age, 52 [SD, 17.1] years; range, 19-88 years) were enrolled. Ninety of 131 (69%) had a benign lesion, and 41/131 (31%) had a malignant lesion. Univariable analysis showed a higher median of transfer constant from blood plasma to the interstitial environment (K trans) and of transfer constant from the interstitial environment to the blood plasma (minute-1) (Kep) and a higher interquartile range of K trans in malignant-versus-benign lesions (1.1 minute-1 versus 0.65 minute-1, P = .03; 2.1 minute-1 versus 1.1 minute-1, P = .01; 0.81 minute-1 versus 0.65 minute-1, P = .009, respectively). The best-performing multivariable model in distinguishing malignant-versus-benign lesions included parameters from dynamic contrast-enhanced imaging, ADC, and morphology and reached an area under the curve of 0.81 (95% CI, 0.67-0.96), a sensitivity of 0.82 (95% CI, 0.55-1), and a specificity of 0.81 (95% CI, 0.65-0.96).

CONCLUSIONS

Dynamic contrast-enhanced MR imaging at 3T appears valuable when characterizing orbital lesions and provides complementary information to morphologic imaging and DWI.

A cost-effectiveness analysis of an integrated clinical-radiogenomic screening program for the identification of BRCA 1/2 carriers (e-PROBE study)

Current approach to identify BRCA 1/2 carriers in the general population is ineffective as most of the carriers remain undiagnosed. Radiomics is an emerging tool for large scale quantitative analysis of features from standard diagnostic imaging and has been applied also to identify gene mutational status. The objective of this study was to evaluate the clinical and economic impact of integrating a radiogenomics model with clinical and family history data in identifying BRCA mutation carriers in the general population. This cost-effective analysis compares three different approaches to women selection for BRCA testing: established clinical criteria/family history (model 1); established clinical criteria/family history and the currently available radiogenomic model (49% sensitivity and 87% specificity) based on ultrasound images (model 2); same approach used in model 2 but simulating an improvement of the performances of the radiogenomic model (80% sensitivity and 95% specificity) (model 3). All models were trained with literature data. Direct costs were calculated according to the rates currently used in Italy. The analysis was performed simulating different scenarios on the generation of 18-year-old girls in Italy (274,000 people). The main outcome was to identify the most effective model comparing the number of years of BRCA-cancer healthy life expectancy (HLYs). An incremental cost-effectiveness ratio (ICER) was also derived to determine the cost in order to increase BRCA carriers-healthy life span by 1 year. Compared to model 1, model 2 increases the detection rate of BRCA carriers by 41.8%, reduces the rate of BRCA-related cancers by 23.7%, generating over a 62-year observation period a cost increase by 2.51 €/Year/Person. Moreover, model 3 further increases BRCA carriers detection (+ 68.3%) and decrease in BRCA-related cancers (- 38.4%) is observed compared to model 1. Model 3 increases costs by 0.7 €/Year/Person. After one generation, the estimated ICER in the general population amounts to about 3800€ and 653€ in model 2 and model 3 respectively. Model 2 has a massive effect after only one generation in detecting carriers in the general population with only a small cost increment. The clinical impact is limited mainly due to the current low acceptance rate of risk-reducing surgeries. Further multicentric studies are required before implementing the integrated clinical-radiogenomic model in clinical practice.

Predicting central lymph node metastasis in patients with papillary thyroid carcinoma based on ultrasound radiomic and morphological features analysis

OBJECTIVES

To build a combined model based on the ultrasound radiomic and morphological features, and evaluate its diagnostic performance for preoperative prediction of central lymph node metastasis (CLNM) in patients with papillary thyroid carcinoma (PTC).

METHOD

A total of 295 eligible patients, who underwent preoperative ultrasound scan and were pathologically diagnosed with unifocal PTC were included at our hospital from October 2019 to July 2022. According to ultrasound scanners, patients were divided into the training set (115 with CLNM; 97 without CLNM) and validation set (45 with CLNM; 38 without CLNM). Ultrasound radiomic, morphological, and combined models were constructed using multivariate logistic regression. The diagnostic performance was assessed by the area under the curve (AUC) of the receiver operating characteristic curve, accuracy, sensitivity, and specificity.

RESULTS

A combined model was built based on the morphology, boundary, length diameter, and radiomic score. The AUC was 0.960 (95% CI, 0.924-0.982) and 0.966 (95% CI, 0.901-0.993) in the training and validation set, respectively. Calibration curves showed good consistency between prediction and observation, and DCA demonstrated the clinical benefit of the combined model.

CONCLUSION

Based on ultrasound radiomic and morphological features, the combined model showed a good performance in predicting CLNM of patients with PTC preoperatively.

Incremental predictive utility of a radiomics signature in a nomogram for the recurrence of atrial fibrillation

Background

Recurrence of atrial fibrillation (AF) after catheter ablation (CA) remains a challenge today. Although it is believed that evaluating the structural and functional remodeling of the left atrium (LA) may be helpful in predicting AF recurrence, there is a lack of consensus on prediction accuracy. Ultrasound-based radiomics is currently receiving increasing attention because it might aid in the diagnosis and prognosis prediction of AF recurrence. However, research on LA ultrasound radiomics is limited.

Objective

We aim to investigate the incremental predictive utility of LA radiomics and construct a radiomics nomogram to preoperatively predict AF recurrence following CA.

Methods

A training cohort of 232 AF patients was designed for nomogram construction, while a validation cohort (n = 100) served as the model performance test. AF recurrence during a follow-up period of 3-12 months was defined as the endpoint. The radiomics features related to AF recurrence were extracted and selected to create the radiomics score (rad score). These rad scores, along with other morphological and functional indicators for AF recurrence, were included in the multivariate Cox analysis to establish a nomogram for the prediction of the likelihood of AF recurrence within 1 year following CA.

Results

In the training and validation cohorts, AF recurrence rates accounted for 32.3% (75/232) and 25.0% (25/100), respectively. We extracted seven types of radiomics features associated with AF recurrence from apical four-chamber view echocardiography images and established a rad score for each patient. The radiomics nomogram was built with the rad score, AF type, left atrial appendage emptying flow velocity, and peak atrial longitudinal strain. It outperformed the nomogram building without the rad score in terms of the predictive efficacy of CA outcome and showed favorable performance in both cohorts.

Conclusion

We revealed the incremental utility of a radiomics signature in the prediction of AF recurrence and preliminarily developed and validated a radiomics nomogram for identifying patients who were at high risk of post-CA recurrence, which contributed to an appropriate management strategy for AF.

Radiomics-based sialadenitis staging in contrast-enhanced computed tomography and ultrasonography: a preliminary rat model study

OBJECTIVE

The aim of this study was to measure the ability of radiomics analysis to diagnose different stages of sialadenitis, compare the diagnostic accuracy of computed tomography (CT) and ultrasonography (US), and suggest radiomics features selected through 3 machine learning algorithms that would be helpful in discriminating between stages of sialadenitis with both imaging systems.

STUDY DESIGN

Wistar rats were treated to induce acute and chronic sialadenitis in the left and right submandibular glands, respectively. Contrast-enhanced CT and US of the glands were performed, followed by extirpation and histopathologic confirmation. Radiomics feature values of the glands were obtained from all images. Based on 3 feature selection methods, an optimal feature set was defined after a comparison of the receiver operating characteristic area under the curve (AUC) of each combination of 3 deep learning algorithms and 3 classification models.

RESULTS

The attribute features for the CT model were 2 gray-level run length matrices and 2 gray-level zone length matrices. In the US model, there were 2 gray-level co-occurrence matrices and 2 gray-level zone length matrices. The most accurate diagnostic models of CT and US yielded outstanding (AUC = 1.000) and excellent (AUC = 0.879) discrimination, respectively.

CONCLUSIONS

The radiomics diagnostic model using gray-level zone length matrices-based features conferred clinically outstanding discriminating ability among stages of sialadenitis using CT and excellent discrimination with US in almost all combinations of machine learning feature selections and classification models.

Prediction of Renal Function 1 Year After Transplantation Using Machine Learning Methods Based on Ultrasound Radiomics Combined With Clinical and Imaging Features

Kidney transplantation is the most effective treatment for advanced chronic kidney disease (CKD). If the prognosis of transplantation can be predicted early after transplantation, it might improve the long-term survival of patients with transplanted kidneys. Currently, studies on the assessment and prediction of renal function by radiomics are limited. Therefore, the present study aimed to explore the value of ultrasound (US)-based imaging and radiomics features, combined with clinical features to develop and validate the models for predicting transplanted kidney function after 1 year (TKF-1Y) using different machine learning algorithms. A total of 189 patients were included and classified into the abnormal TKF-1Y group, and the normal TKF-1Y group based on their estimated glomerular filtration rate (eGFR) levels 1 year after transplantation. The radiomics features were derived from the US images of each case. Three machine learning methods were employed to establish different models for predicting TKF-1Y using selected clinical and US imaging as well as radiomics features from the training set. Two US imaging, four clinical, and six radiomics features were selected. Then, the clinical (including clinical and US image features), radiomics, and combined models were developed. The area under the curves (AUCs) of the models was 0.62 to 0.82 within the test set. Combined models showed statistically higher AUCs than the radiomics models (all p-values <.05). The prediction performance of different models was not significantly affected by the different machine learning algorithms (all p-values >.05). In conclusion, US imaging features combined with clinical features could predict TKF-1Y and yield an incremental value over radiomics features. A model integrating all available features may further improve the predictive efficacy. Different machine learning algorithms may not have a significant impact on the predictive performance of the model.

Radiomics: A review of current applications and possibilities in the assessment of tumor microenvironment

With the recent success in the application of immunotherapy for treating various advanced cancers, the tumor microenvironment has rapidly become an important field of research. The tumor microenvironment is complex and its characteristics strongly influence disease biology and potentially responses to systemic therapy. Accurate preoperative assessment of tumor microenvironment is of great significance for the formulation of an immunotherapy strategy and evaluation of patient prognosis. As a research hotspot in medical image analysis technology, radiomics has been applied in the auxiliary diagnosis of the tumor microenvironment. This article reviews the current status of radiomics in the elective application on tumor microenvironment and discusses potential prospects.

Artificial intelligence in diagnostic and interventional radiology: Where are we now?

The emergence of massively parallel yet affordable computing devices has been a game changer for research in the field of artificial intelligence (AI). In addition, dramatic investment from the web giants has fostered the development of a high-quality software stack. Going forward, the combination of faster computers with dedicated software libraries and the widespread availability of data has opened the door to more flexibility in the design of AI models. Radiomics is a process used to discover new imaging biomarkers that has multiple applications in radiology and can be used in conjunction with AI. AI can be used throughout the various processes of diagnostic imaging, including data acquisition, reconstruction, analysis and reporting. Today, the concept of "AI-augmented" radiologists is preferred to the theory of the replacement of radiologists by AI in many indications. Current evidence bolsters the assumption that AI-assisted radiologists work better and faster. Interventional radiology becomes a data-rich specialty where the entire procedure is fully recorded in a standardized DICOM format and accessible via standard picture archiving and communication systems. No other interventional specialty can bolster such readiness. In this setting, interventional radiology could lead the development of AI-powered applications in the broader interventional community. This article provides an update on the current status of radiomics and AI research, analyzes upcoming challenges and also discusses the main applications in AI in interventional radiology to help radiologists better understand and criticize articles reporting AI in medical imaging.

Artificial intelligence in adrenal imaging: A critical review of current applications

In the elective field of adrenal imaging, artificial intelligence (AI) can be used for adrenal lesion detection, characterization, hypersecreting syndrome management and patient follow-up. Although a perfect AI tool that includes all required steps from detection to analysis does not exist yet, multiple AI algorithms have been developed and tested with encouraging results. However, AI in this setting is still at an early stage. In this regard, most published studies about AI in adrenal gland imaging report preliminary results that do not have yet daily applications in clinical practice. In this review, recent developments and current results of AI in the field of adrenal imaging are presented. Limitations and future perspectives of AI are discussed.

Artificial intelligence: A review of current applications in hepatocellular carcinoma imaging

Hepatocellular carcinoma (HCC) is the most common type of primary liver cancer and currently the third-leading cause of cancer-related death worldwide. Recently, artificial intelligence (AI) has emerged as an important tool to improve clinical management of HCC, including for diagnosis, prognostication and evaluation of treatment response. Different AI approaches, such as machine learning and deep learning, are both based on the concept of developing prediction algorithms from large amounts of data, or big data. The era of digital medicine has led to a rapidly expanding amount of routinely collected health data which can be leveraged for the development of AI models. Various studies have constructed AI models by using features extracted from ultrasound imaging, computed tomography imaging and magnetic resonance imaging. Most of these models have used convolutional neural networks. These tools have shown promising results for HCC detection, characterization of liver lesions and liver/tumor segmentation. Regarding treatment, studies have outlined a role for AI in evaluation of treatment response and improvement of pre-treatment planning. Several challenges remain to fully integrate AI models in clinical practice. Future research is still needed to robustly evaluate AI algorithms in prospective trials, and improve interpretability, generalizability and transparency. If such challenges can be overcome, AI has the potential to profoundly change the management of patients with HCC. The purpose of this review was to sum up current evidence on AI approaches using imaging for the clinical management of HCC.

Endorectal ultrasound radiomics in locally advanced rectal cancer patients: despeckling and radiotherapy response prediction using machine learning

PURPOSE

The current study aimed to evaluate the association of endorectal ultrasound (EUS) radiomics features at different denoising filters based on machine learning algorithms and to predict radiotherapy response in locally advanced rectal cancer (LARC) patients.

METHODS

The EUS images of forty-three LARC patients, as a predictive biomarker for predicting the treatment response of neoadjuvant chemoradiotherapy (NCRT), were investigated. For despeckling, the EUS images were preprocessed by traditional filters (bilateral, wiener, lee, frost, median, and wavelet filters). The rectal tumors were delineated by two readers separately, and radiomics features were extracted. The least absolute shrinkage and selection operator were used for feature selection. Classifiers including logistic regression (LR), K-nearest neighbor (KNN), support vector machine (SVM), random forest, naive Bayes, and decision tree were trained using stratified fivefold cross-validation for model development. The area under the curve (AUC) of the receiver operating characteristic curve followed by accuracy, precision, sensitivity, and specificity were obtained for model performance assessment.

RESULTS

The wavelet filter had the best results with means of AUC: 0.83, accuracy: 77.41%, precision: 82.15%, and sensitivity: 79.41%. LR and SVM by having AUC: 0.71 and 0.76; accuracy: 70.0% and 71.5%; precision: 75.0% and 73.0%; sensitivity: 69.8% and 80.2%; and specificity: 70.0% and 60.9% had the highest model's performance, respectively.

CONCLUSION

This study demonstrated that the EUS-based radiomics model could serve as pretreatment biomarkers in predicting pathologic features of rectal cancer. The wavelet filter and machine learning methods (LR and SVM) had good results on the EUS images of rectal cancer.

Comparison between 7 Tesla and 3 Tesla MRI for characterizing orbital lesions

PURPOSE

Characterizing orbital lesions remains challenging with imaging. The purpose of this study was to compare 3 Tesla (T) to 7 T magnetic resonance imaging (MRI) for characterizing orbital lesions.

MATERIALS AND METHODS

This prospective single-center study enrolled participants presenting with orbital lesions from May to October 2019, who underwent both 7 T and 3 T MRI examinations. Two neuroradiologists, blinded to all data, read both datasets independently and randomly. They assessed general characteristics of each orbital lesion as well as image quality and presence of artifacts. Comparison between both datasets was made using Fisher exact test.

RESULTS

Seven patients (4 women, 3 men) with a median age of 52 years were enrolled. Orbital lesion conspicuity was better scored at 7 T compared to 3 T MRI, with 3/7 lesions (43%) scored as very conspicuous at 7 T compared to 0/7 lesion (0%) at 3 T, although the difference was not significant (P = 0.16). Delineation of lesion margins was better scored at 7 T compared to 3 T with 3/7 lesions (43%) scored as very well delineated on 7 T compared to 0/7 lesions (0%) at 3 T, although the difference was not significant (P = 0.34). Details of internal structure were better assessed at 7 T compared to 3 T, with 4/7 lesions (57%) displaying numerous internal details compared to 0/7 lesions (0%) at 3 T (P = 0.10). Internal microvessels were visible in 3/7 lesions (43%) at 7 T compared to 0/7 lesions (0%) at 3 T (P = 0.19).

CONCLUSION

Although no significant differences were found between 7 T and 3 T MRI, assumably due to a limited number of patients, our study suggests that 7 Tesla MRI might help improve the characterization of orbital lesions. However, further studies with more patients are needed.

Impact of image quality on radiomics applications

Radiomics features extracted from medical images have been widely reported to be useful in the patient specific outcome modeling for variety of assessment and prediction purposes. Successful application of radiomics features as imaging biomarkers, however, is dependent on the robustness of the approach to the variation in each step of the modeling workflow. Variation in the input image quality is one of the main sources that impacts the reproducibility of radiomics analysis when a model is applied to broader range of medical imaging data. The quality of medical image is generally affected by both the scanner related factors such as image acquisition/reconstruction settings and the patient related factors such as patient motion. This article aimed to review the published literatures in this field that reported the impact of various imaging factors on the radiomics features through the change in image quality. The literatures were categorized by different imaging modalities and also tabulated based on the imaging parameters and the class of radiomics features included in the study. Strategies for image quality standardization were discussed based on the relevant literatures and recommendations for reducing the impact of image quality variation on the radiomics in multi-institutional clinical trial were summarized at the end of this article.