Dual-mode ultrasound radiomics and intrinsic imaging phenotypes for diagnosis of lymph node lesions

Classification of superficial suspected lymph nodes: non-invasive radiomic model based on multiphase contrast-enhanced ultrasound for therapeutic options of lymphadenopathy

Background

Accurate determination of the types of lymphadenopathy is of great importance in disease diagnosis and treatment and is usually confirmed by pathological findings. Radiomics is a non-invasive tool that can extract quantitative information from medical images. Our study was designed to develop a non-invasive radiomic approach based on multiphase contrast-enhanced ultrasound (CEUS) images for the classification of different types of lymphadenopathy.

Methods

A total of 426 patients with superficial suspected lymph nodes (LNs) from three centres were grouped into a training cohort (n=190), an internal testing cohort (n=127), and an external testing cohort (n=109). The radiomic features were extracted from the prevascular phase, vascular phase, and postvascular phase of the CEUS images. Model 1 (the conventional feature model), model 2 (the multiphase radiomics model), and model 3 (the combined feature model) were established for lymphadenopathy classification. The area under the curve (AUC) and confusion matrix were used to evaluate the performance of the three models. The usefulness of the models was assessed in different threshold probabilities by decision curve analysis.

Results

There were 139 patients (32.6%) with benign LNs, 110 patients (25.8%) with lymphoma, and 177 patients (41.5%) with metastatic LNs in our population. Finally, twenty features were selected to construct the radiomics models for these three types of lymphadenopathy. Model 2 integrating multiphase images of the CEUS yielded the AUCs of 0.838, 0.739, and 0.733 in the training cohort, internal testing cohort, and external testing cohort, respectively. After the combination of conventional features and radiomic features, the AUCs of model 3 improved to 0.943, 0.823 and 0.785 in the training cohort, internal testing cohort, and external testing cohort. Besides, model 3 had an accuracy of 81.05%, sensitivity of 80%, and specificity of 90.43% in the training cohort. Model performance was further confirmed in the internal testing cohort and external testing cohort.

Conclusions

We constructed a combined feature model using a series of CEUS images for the classification of the lymphadenopathies. For patients with superficial suspected LNs, this model can help clinicians make a decision on the LN type noninvasively and choose appropriate treatments.

Application and prospects of AI-based radiomics in ultrasound diagnosis

Artificial intelligence (AI)-based radiomics has attracted considerable research attention in the field of medical imaging, including ultrasound diagnosis. Ultrasound imaging has unique advantages such as high temporal resolution, low cost, and no radiation exposure. This renders it a preferred imaging modality for several clinical scenarios. This review includes a detailed introduction to imaging modalities, including Brightness-mode ultrasound, color Doppler flow imaging, ultrasound elastography, contrast-enhanced ultrasound, and multi-modal fusion analysis. It provides an overview of the current status and prospects of AI-based radiomics in ultrasound diagnosis, highlighting the application of AI-based radiomics to static ultrasound images, dynamic ultrasound videos, and multi-modal ultrasound fusion analysis.

MSMFN: An Ultrasound Based Multi-Step Modality Fusion Network for Identifying the Histologic Subtypes of Metastatic Cervical Lymphadenopathy

Identifying squamous cell carcinoma and adenocarcinoma subtypes of metastatic cervical lymphadenopathy (CLA) is critical for localizing the primary lesion and initiating timely therapy. B-mode ultrasound (BUS), color Doppler flow imaging (CDFI), ultrasound elastography (UE) and dynamic contrast-enhanced ultrasound provide effective tools for identification but synthesis of modality information is a challenge for clinicians. Therefore, based on deep learning, rationally fusing these modalities with clinical information to personalize the classification of metastatic CLA requires new explorations. In this paper, we propose Multi-step Modality Fusion Network (MSMFN) for multi-modal ultrasound fusion to identify histological subtypes of metastatic CLA. MSMFN can mine the unique features of each modality and fuse them in a hierarchical three-step process. Specifically, first, under the guidance of high-level BUS semantic feature maps, information in CDFI and UE is extracted by modality interaction, and the static imaging feature vector is obtained. Then, a self-supervised feature orthogonalization loss is introduced to help learn modality heterogeneity features while maintaining maximal task-consistent category distinguishability of modalities. Finally, six encoded clinical information are utilized to avoid prediction bias and improve prediction ability further. Our three-fold cross-validation experiments demonstrate that our method surpasses clinicians and other multi-modal fusion methods with an accuracy of 80.06%, a true-positive rate of 81.81%, and a true-negative rate of 80.00%. Our network provides a multi-modal ultrasound fusion framework that considers prior clinical knowledge and modality-specific characteristics. Our code will be available at: https://github.com/RichardSunnyMeng/MSMFN.

Intelligent diagnosis of left ventricular hypertrophy using transthoracic echocardiography videos

PURPOSE

Left ventricular hypertrophy (LVH) is an independent risk factor for cardiovascular events and mortality. Pathological LVH can be caused by various diseases. In this study, we explored the possibility of using time and frequency domain analysis of myocardial radiomics features for patients with LVH in differentiating hypertrophic cardiomyopathy (HCM), hypertensive heart disease (HHD) and uremic cardiomyopathy (UCM) based on transthoracic echocardiography (TTE). This was the first study to explore TTE myocardial time and frequency domain analyses for multiple LVH etiology differentiation.

MATERIALS AND METHODS

We proposed an artificially intelligent diagnosis system based on radiomics techniques for differentiating HCM, HHD and UCM on TTE videos of the apical four-chamber view, which mainly included interventricular septum (IVS) segmentation, feature extraction and classification. We used two independent cohorts, one with 150 patients, including 50 HHD, 50 HCM and 50 UCM, for segmentation training and testing, and another with 149 patients (namely the main cohort), including 50 HHD, 46 HCM and 53 UCM, for classification training and testing after segmentation and feature extraction. Firstly, the U-Net, Residual U-Net (ResUNet) and nnU-Net were trained and tested to segment the IVS on TTE still images in the first cohort. Then the trained model with the best segmentation performance was further used for IVS prediction of ordered TTE images in video sequences in the main cohort. The post-processing was used to eliminate the noisy debris by selecting the maximum connected region and smoothing the edges of the predicted IVS region. Secondly, static radiomics features were extracted from the IVS of ordered TTE images in each video sequence, and subsequently the time and frequency domain features were further extracted from each time series of a static radiomics feature in the video sequence. Finally, the point-wise gated Boltzmann machine (PGBM) was used to learn and fuse the time and frequency domain features, and the support vector machine was used to classify the learned features for LVH diagnosis. The classification was performed with five-fold cross validation.

RESULTS

The ResUNet showed the best segmentation performance, with Dice coefficient, sensitivity, specificity and accuracy of 0.817, 76.3%, 99.6% and 98.6%, respectively. With post-processing, the Dice coefficient, sensitivity, specificity and accuracy of the ResUNet were further improved to 0.839, 77.0%, 99.8%, and 98.8%, respectively. The classification areas under the receiver operating characteristic curves (AUCs) were 0.838 ± 0.049 for HHD vs. HCM, 0.868 ± 0.042 for HCM vs. UCM and 0.701 ± 0.140 for HHD vs. UCM.

CONCLUSION

In this work, we proposed an intelligent identification system for LVH etiology classification based on routine TTE video images with good diagnostic performance. This deep learning method is feasible in automatic TTE images interpretation and expected to assist clinicians in detecting the primary cause of LVH.

Exploratory study on the predictive value of ultrasound radiomics for cervical tuberculous lymphadenitis

PURPOSE

To investigate the predictive values of gray-scale ultrasound (G-US) and strain elastic ultrasound (SE-US) radiomic features for cervical tuberculous lymphadenitis (CTL).

MATERIAL AND METHODS

The G-US and SE-US images of 147 patients with pathologically confirmed CTL and 69 non-CTL patients were retrospectively analyzed. A total of 851 imaging features were extracted. The patients were divided into the training set and test set in 7:3 ratio. In the training set, the minimum redundancy maximum relevance (mRMR) and the least absolute shrinkage and selection operator (LASSO) were used for feature selection and modeling. The diagnostic power of G-US and SE-US ultrasound radiomics in identifying CTL was evaluated in test set.

RESULTS

The G-US and SE-US have finally selected 10 and 14 features, respectively. In the G-US group, the diagnostic sensitivity, specificity and accuracy of the training set were 69.7%, 85.7% and 70.0%, respectively, and those values in the test set were 81.3%, 70.0% and 86.4%, respectively. The SE-US group had a sensitivity of 71.7%, a specificity of 81.6%, and an accuracy of 67.0% in the training set, and those parameters in the test set were 81.0%, 75.0%, and 83.7%, respectively. In the G-US group, the positive and negative predictive value of the training set were 0.519 and 0.901, respectively, and those values in the test set were 0.700 and 0.864, respectively. The SE-US group had a positive predictive value of 0.541, and a negative predictive value of 0.885 in the training set, and those parameters in the test set were 0.682 and 0.878, respectively. By Delong test, G-US and SE-US groups showed no significant differences in diagnostic performance between the training and test sets.

CONCLUSIONS

The ultrasound radiomic features of G-US and SE-US exhibited certain predictive potential in detecting CTL, providing a new non-invasive method for clinicians to more accurately evaluate patients with CTL.

Application of Machine Learning Methods to Improve the Performance of Ultrasound in Head and Neck Oncology: A Literature Review

Simple Summary

Ultrasound (US) is a non-invasive imaging method that is routinely utilized in head and neck cancer patients to assess the anatomic extent of tumors, nodal and non-nodal neck masses and for imaging the salivary glands. In this review, we summarize the present evidence on whether the application of machine learning (ML) methods can potentially improve the performance of US in head and neck cancer patients. We found that published clinical literature on ML methods applied to US datasets was limited but showed evidence of improved diagnostic and prognostic performance. However, a majority of these studies were based on retrospective evaluation and conducted at a single center with a limited number of datasets. The conduct of multi-center studies could help better validate the performance of ML-based US radiomics and facilitate the integration of these approaches into routine clinical practice.

Abstract

Radiomics is a rapidly growing area of research within radiology that involves the extraction and modeling of high-dimensional quantitative imaging features using machine learning/artificial intelligence (ML/AI) methods. In this review, we describe the published clinical evidence on the application of ML methods to improve the performance of ultrasound (US) in head and neck oncology. A systematic search of electronic databases (MEDLINE, PubMed, clinicaltrials.gov) was conducted according to Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Of 15,080 initial articles identified, 34 studies were selected for in-depth analysis. Twenty-five out of 34 studies (74%) focused on the diagnostic application of US radiomics while 6 (18%) studies focused on response assessment and 3 (8%) studies utilized US radiomics for modeling normal tissue toxicity. Support vector machine (SVM) was the most commonly employed ML method (47%) followed by multivariate logistic regression (24%) and k-nearest neighbor analysis (21%). Only 11/34 (~32%) of the studies included an independent validation set. A majority of studies were retrospective in nature (76%) and based on single-center evaluation (85%) with variable numbers of patients (12–1609) and imaging datasets (32–1624). Despite these limitations, the application of ML methods resulted in improved diagnostic and prognostic performance of US highlighting the potential clinical utility of this approach.