CT Texture Analysis of Cervical Lymph Nodes on Contrast-Enhanced [18F] FDG-PET/CT Images to Differentiate Nodal Metastases from Reactive Lymphadenopathy in HIV-Positive Patients with Head and Neck Squamous Cell Carcinoma

Prediction of extranodal extension in head and neck squamous cell carcinoma by CT images using an evolutionary learning model

BACKGROUND

Extranodal extension (ENE) in head and neck squamous cell carcinoma (HNSCC) correlates to poor prognoses and influences treatment strategies. Deep learning may yield promising performance of predicting ENE in HNSCC but lack of transparency and interpretability. This work proposes an evolutionary learning method, called EL-ENE, to establish a more interpretable ENE prediction model for aiding clinical diagnosis.

METHODS

There were 364 HNSCC patients who underwent neck lymph node (LN) dissection with pre-operative contrast-enhanced computerized tomography images. All the 778 LNs were divided into training and test sets with the ratio 8:2. EL-ENE uses an inheritable bi-objective combinatorial genetic algorithm for optimal feature selection and parameter setting of support vector machine. The diagnostic performances of the ENE prediction model and radiologists were compared using independent test datasets.

RESULTS

The EL-ENE model achieved the test accuracy of 80.00%, sensitivity of 81.13%, and specificity of 79.44% for ENE detection. The three radiologists achieved the mean diagnostic accuracy of 70.4%, sensitivity of 75.6%, and specificity of 67.9%. The features of gray-level texture and 3D morphology of LNs played essential roles in predicting ENE.

CONCLUSIONS

The EL-ENE method provided an accurate, comprehensible, and robust model to predict ENE in HNSCC with interpretable radiomic features for expanding clinical knowledge. The proposed transparent prediction models are more trustworthy and may increase their acceptance in daily clinical practice.

Research to develop a diagnostic ultrasound nomogram to predict benign or malignant lymph nodes in HIV-infected patients

BACKGROUND

This study aimed to establish an effective ultrasound diagnostic nomogram for benign or malignant lymph nodes in HIV-infected patients.

METHODS

The nomogram is based on a retrospective study of 131 HIV-infected patients who underwent ultrasound assess at the Shanghai Public Health Clinical Center from December 2017 to July 2022. The nomogram's predictive accuracy and discriminative ability were determined by concordance index (C-index) and calibration curve analysis. A nomogram combining the lymph node US characteristics were generated based on the multivariate logistic regression results.

RESULTS

Predictors contained in the ultrasound diagnostic nomogram included age (OR 1.044 95%CI: 1.014-1.074 P = 0.004), number of enlarged lymph node regions (OR 5.445 95%CI: 1.139-26.029 P = 0.034), and color Doppler flow imaging (CDFI) grades (OR 9.614 95%CI: 1.889-48.930 P = 0.006). The model displayed good discrimination with a C (ROC) of 0.775 and good calibration.

CONCLUSIONS

The proposed nomogram may result in more-accurate diagnostic predictions for benign or malignant lymph nodes in patients with HIV infection.

Deep learning combined with radiomics for the classification of enlarged cervical lymph nodes

PURPOSE

To investigate the application of deep learning combined with traditional radiomics methods for classifying enlarged cervical lymph nodes.

METHODS

The clinical and computed tomography (CT) imaging data of 276 patients with enlarged cervical lymph nodes (150 with lymph-node metastasis, 65 with lymphoma, and 61 with benign lymphadenopathy) who were treated at the hospital from January 2015 to January 2021 were retrospectively analysed. The patients were randomly divided into a training group and a test group at a ratio of 8:2. The radiomics features were extracted using one-by-one convolution and neural network activation, filtered with the least absolute shrinkage and selection operator (LASSO) model, and used to construct a discrimination model with PyTorch. Then, the performance of the model was compared with the radiologists' diagnostic performance. The neural network model was evaluated using the area under the receiver-operator characteristic curve (AUC), and the accuracy, sensitivity, and specificity were analysed.

RESULTS

A total of 102 features, comprising five traditional radiomic features and 97 deep learning features, were selected with LASSO and used to construct a discrimination model, which achieved a total accuracy of 87.50%. The AUC value, specificity, and sensitivity were, respectively, 0.92, 92.30%, and 90.00% for metastatic lymph nodes, 0.87, 95.45%, and 83.33% for benign lymphadenopathy, and 0.88, 90.47%, and 85.71% for lymphoma. The accuracies of the radiologists' diagnoses were 62.68% and 62.68%. The diagnostic performance of the model was significantly different from that of the radiologists (p < 0.05).

CONCLUSION

CT-based deep learning combined with the traditional radiomics methods has a high diagnostic value for the classification of cervical enlarged lymph nodes.

Radiological differences in computed tomography findings and texture analysis between cystic lymph node metastases of human papillomavirus-positive oropharyngeal cancer and second branchial cysts

Purpose

The study aimed to analyse radiological differences in computed tomography (CT) findings and texture analysis between cystic lymph node metastases (CNM) in human papillomavirus (HPV)-positive oropharyngeal cancer (OPC) and second branchial cleft cysts (2^nd BC).

Material and methods

Patients with pathological evidence of CNM-HPV-OPC and 2^nd BC, who underwent contrast-enhanced CT, were retrospectively evaluated. The evaluated characteristics include age, sex, and CT findings. CT findings included the maximum and minimum transverse diameters, maximum caudal diameter, thickness of the peripheral wall, presence of internal septation, presence of surrounding fat stranding, location, and 40 texture parameters.

Results

A total of 13 patients had CNM-HPV-OPC (19 lesions), while 20 patients had 2^nd BC (20 lesions). Patients with 2^nd BC were significantly younger than those with CNM-HPV-OPC (p < 0.001). In terms of diameter, 2^nd BC lesions were significantly larger than the CNM-HPV-OPC lesions (p < 0.001). CNM-HPV OPC lesions had significantly thicker walls than 2^nd BC lesions (p < 0.001). CNM-HPV-OPC lesions had significantly higher association with internal septations than 2^nd BC lesions (p < 0.001). Second BC lesions were significantly less common at level III than CNM-HPV-OPC lesions (p = 0.047). Among the 40 texture parameters measured, 8 had significant differences (p ≤ 0.001).

Conclusions

There were significant differences in CT findings and textural parameters between CNM-HPV-OPC and 2^nd BC lesions. These results may help in differentiating one from the other.

Nodal-based radiomics analysis for identifying cervical lymph node metastasis at levels I and II in patients with oral squamous cell carcinoma using contrast-enhanced computed tomography

OBJECTIVE

Discriminating metastatic from benign cervical lymph nodes (LNs) in oral squamous cell carcinoma (OSCC) patients using pretreatment computed tomography (CT) has been controversial. This study aimed to investigate whether CT-based texture analysis with machine learning can accurately identify cervical lymph node metastasis in OSCC patients.

METHODS

Twenty-three patients (with 201 cervical LNs [150 benign, 51 metastatic] at levels I-V) who underwent preoperative contrast-enhanced CT and subsequent cervical neck dissection were enrolled. Histopathologically proven LNs were randomly divided into the training cohort (70%; n = 141, at levels I-V) and validation cohort (30%; n = 60, at level I/II). Twenty-five texture features and the nodal size of targeted LNs were analyzed on the CT scans. The nodal-based sensitivities, specificities, diagnostic accuracy rates, and the area under the curves (AUCs) of the receiver operating characteristic curves of combined features using a support vector machine (SVM) at levels I/II, I, and II were evaluated and compared with two radiologists and a dentist (readers).

RESULTS

In the validation cohort, the AUCs (0.820 at level I/II, 0.820 at level I, and 0.930 at level II, respectively) of the radiomics approach were superior to three readers (0.798-0.816, 0.773-0.798, and 0.825-0.865, respectively). The best models were more specific at levels I/II and I and accurate at each level than each of the readers (p < .05).

CONCLUSIONS

Machine learning-based analysis with contrast-enhanced CT can be used to noninvasively differentiate between benign and metastatic cervical LNs in OSCC patients.

KEY POINTS

• The best algorithm in the validation cohort can noninvasively differentiate between benign and metastatic cervical LNs at levels I/II, I, and II. • The AUCs of the model at each level were superior to those of multireaders. • Significant differences in the specificities at level I/II and I and diagnostic accuracy rates at each level between the model and multireaders were found.

Machine learning-based MRI texture analysis to predict occult lymph node metastasis in early-stage oral tongue squamous cell carcinoma

OBJECTIVES

To develop and compare several machine learning models to predict occult cervical lymph node (LN) metastasis in early-stage oral tongue squamous cell cancer (OTSCC) from preoperative MRI texture features.

MATERIALS AND METHODS

We retrospectively enrolled 116 patients with early-stage OTSCC (cT1-2N0) who had been surgically treated by tumor excision and elective neck dissection (END). For each patient, we extracted 86 texture features from T2-weighted imaging (T2WI) and contrast-enhanced T1-weighted imaging (ceT1WI), respectively. Dimension reduction was performed in three consecutive steps: reproducibility analysis, collinearity analysis, and information gain algorithm. Models were created using six machine learning methods, including logistic regression (LR), random forest (RF), naïve Bayes (NB), support vector machine (SVM), AdaBoost, and neural network (NN). Their performance was assessed using tenfold cross-validation.

RESULTS

Occult LN metastasis was pathologically detected in 42.2% (49/116) of the patients. No significant association was identified between node status and patients' gender, age, or clinical T stage. Dimension reduction steps selected 6 texture features. The NB model gave the best overall performance, which correctly classified the nodal status in 74.1% (86/116) of the carcinomas, with an AUC of 0.802.

CONCLUSION

Machine learning-based MRI texture analysis offers a feasible tool for preoperative prediction of occult cervical node metastasis in early-stage OTSCC.

KEY POINTS

• A machine learning-based MRI texture analysis approach was adopted to predict occult cervical node metastasis in early-stage OTSCC with no evidence of node involvement on conventional images. • Six texture features from T2WI and ceT1WI of preoperative MRI were selected to construct the predictive model. • After comparing six machine learning methods, naïve Bayes (NB) achieved the best performance by correctly identifying the node status in 74.1% of the patients, using tenfold cross-validation.

Is the prediction of one or two ipsilateral positive lymph nodes by computerized tomography and ultrasound reliable enough to restrict therapeutic neck dissection in oral squamous cell carcinoma (OSCC) patients?

INTRODUCTION

Proper management of the clinically involved neck in OSCC patients continues to be a matter of debate. Our aim was to analyze the accuracy of computerized tomography (CT) and ultrasound (US) in anticipating the exact location of lymph node (LN) metastases of OSCC patients across the AAO-HNS (American Academy of Otolaryngology-Head and Neck Surgery) levels ipsi- and contralaterally. Furthermore, we wanted to assess the suitability of therapeutic selective neck dissection (SND) in patients with one or two ipsilateral positive nodes upon clinical staging (cN1/cN2a and cN2b(2/x) patients).

METHODS

We prospectively analyzed the LN status of patients with primary OSCC using CT and US from 2007 to 2013. LNs were individually assigned to a map containing the AAO-HNS levels; patients bearing a single or just two ipsilateral positive nodes (designated cN1/cN2a or cN2b(2/x) patients either by CT (CT group) or US alone (US group) or in a group combining findings of CT and US (CTUS group)) received an ipsi-ND (I-V) and a contra-ND (I-IV). 78% of the LNs were sent individually for routine histopathological examination; the remaining were dissected and analyzed per neck level.

RESULTS

Upon the analysis of 1.670 LNs of 57 patients, the exact location of pathology proven LN metastases in cN1 patients was more precisely predicted by US compared to CT with confirmed findings only in levels IA, IB und IIA. Clearly decreasing the number of missed lesions, the findings in the CTUS group nearly kept the spatial reliability of the US group. The same analysis for patients with exactly two supposed ipsilateral lesions (cN2b(2/x)) yielded confirmed metastases from levels I to V for both methods individually and in combination and, therefore, render SND insufficient for these cases.

CONCLUSION

Our findings stress the importance of conducting both, CT and US, in patients with primary OSCC. Only the combination of their findings warrants the application of therapeutic SND in patients with a single ipsilateral LN metastasis (cN1/cN2a patients) but not in patients with more than one lesion upon clinical staging (≥ cN2b).

Pre-treatment MRI predictor of high-grade malignant parotid gland cancer

OBJECTIVES

We aimed to evaluate pre-treatment MRI predictors of high-grade malignant parotid gland cancer by comparing MRI findings and texture parameters between high-grade and intermediate/low-grade parotid gland cancers.

METHODS

Patients underwent a pre-treatment MRI and had a parotid gland cancer resection with pathological evaluation. Evaluation objectives included attributive factors such as age and gender, several MRI findings of T1- and T2-weighted images, post-contrast fat suppression T1-weighted images, ADC value and 40 texture parameters calculated from T2-weighted axial images. Such objects were compared between high-grade and intermediate/low-grade lesions.

RESULTS

Of the parotid gland cancers surveyed, 39 were included for analysis. Of these, 18 were high-grade lesions, 2 were intermediate-grade lesions, and 19 were low-grade lesions. The high-grade group was significantly older than the low- and intermediate-grade groups (p = 0.01). There were more males in the high-grade group than in the low- and intermediate-grade groups (p = 0.01). There were also significantly more MRI findings of neck lymph node metastases in the high-grade group than in the low- and intermediate-grade groups (p < 0.001). Other MRI findings and texture parameters did not show significant differences between the two groups (p = 0.07-1.00).

CONCLUSIONS

Morphological assessment on MRI and texture parameters alone is not sufficient to estimate the grade of parotid cancer. MRI findings of neck lymph node metastases, as well as patient characteristics such as age (older patients) and gender (male) can be suggestive of high-grade parotid gland cancer in pre-treatment evaluation.

Quantitative ultrasound delta-radiomics during radiotherapy for monitoring treatment responses in head and neck malignancies

Aim:

We investigated quantitative ultrasound (QUS) in patients with node-positive head and neck malignancies for monitoring responses to radical radiotherapy (RT).

Materials & methods:

QUS spectral and texture parameters were acquired from metastatic lymph nodes 24 h, 1 and 4 weeks after starting RT. K-nearest neighbor and naive-Bayes machine-learning classifiers were used to build prediction models for each time point. Response was detected after 3 months of RT, and patients were classified into complete and partial responders.

Results:

Single-feature naive-Bayes classification performed best with a prediction accuracy of 80, 86 and 85% at 24 h, week 1 and 4, respectively.

Conclusion:

QUS-radiomics can predict RT response at 3 months as early as 24 h with reasonable accuracy, which further improves into 1 week of treatment.

Patients with head and neck cancer are often treated with radiation, which usually spans over 6–7 weeks. The response is usually measured 3 months after treatment completion. In this study, we had performed ultrasound scans from the patient’s neck node during radiation treatment (after 24 h, 1 and 4 weeks). Artificial intelligence was used to interpret the ultrasound imaging and predict the response to radiation at the end of 3 months. The scans obtained after the first week were able to predict the treatment response with reasonable accuracy (86%).

Combined CT texture analysis and nodal axial ratio for detection of nodal metastasis in esophageal cancer

OBJECTIVE

To assess the accuracy of a combination of CT texture analysis (CTTA) and nodal axial ratio to detect metastatic lymph nodes (LNs) in esophageal squamous cell carcinoma (ESCC).

METHODS

The contrast-enhanced chest CT images of 78 LNs (40 metastasis, 38 benign) from 38 patients with ESCC were retrospectively analyzed. Nodal axial ratios (short-axis/long-axis diameter) were calculated. CCTA parameters (kurtosis, entropy, skewness) were extracted using commercial software (TexRAD) with fine, medium, and coarse spatial filters. Combinations of significant texture features and nodal axial ratios were entered as predictors in logistic regression models to differentiate metastatic from benign LNs, and the performance of the logistic regression models was analyzed using the area under the receiver operating characteristic curve (AUROC).

RESULTS

The mean axial ratio of metastatic LNs was significantly higher than that of benign LNs (0.81 ± 0.2 vs 0.71 ± 0.1, p = 0.005; sensitivity 82.5%, specificity 47.4%); namely, significantly more round than benign. The mean values of the entropy (all filters) and kurtosis (fine and medium) of metastatic LNs were significantly higher than those of benign LNs (all, p < 0.05). Medium entropy showed the best performance in the AUROC analysis with 0.802 (p < 0.001; sensitivity 85.0%, specificity 63.2%). A binary logistic regression analysis combining the nodal axial ratio, fine entropy, and fine kurtosis identified metastatic LNs with 87.5% sensitivity and 65.8% specificity (AUROC = 0.855, p < 0.001).

CONCLUSION

The combination of CTTA features and the axial ratio of LNs has the potential to differentiate metastatic from benign LNs and improves the sensitivity for detection of LN metastases in ESCC.

ADVANCES IN KNOWLEDGE

The combination of CTTA and nodal axial ratio has improved CT sensitivity (up to 87.5%) for the diagnosis of metastatic LNs in esophageal cancer.

Dual-Energy CT Texture Analysis With Machine Learning for the Evaluation and Characterization of Cervical Lymphadenopathy

Purpose

To determine whether machine learning assisted-texture analysis of multi-energy virtual monochromatic image (VMI) datasets from dual-energy CT (DECT) can be used to differentiate metastatic head and neck squamous cell carcinoma (HNSCC) lymph nodes from lymphoma, inflammatory, or normal lymph nodes.

Materials and methods

A retrospective evaluation of 412 cervical nodes from 5 different patient groups (50 patients in total) having undergone DECT of the neck between 2013 and 2015 was performed: (1) HNSCC with pathology proven metastatic adenopathy, (2) HNSCC with pathology proven benign nodes (controls for (1)), (3) lymphoma, (4) inflammatory, and (5) normal nodes (controls for (3) and (4)). Texture analysis was performed with TexRAD® software using two independent sets of contours to assess the impact of inter-rater variation. Two machine learning algorithms (Random Forests (RF) and Gradient Boosting Machine (GBM)) were used with independent training and testing sets and determination of accuracy, sensitivity, specificity, PPV, NPV, and AUC.

Results

In the independent testing (prediction) sets, the accuracy for distinguishing different groups of pathologic nodes or normal nodes ranged between 80 and 95%. The models generated using texture data extracted from the independent contour sets had substantial to almost perfect agreement. The accuracy, sensitivity, specificity, PPV, and NPV for correctly classifying a lymph node as malignant (i.e. metastatic HNSCC or lymphoma) versus benign were 92%, 91%, 93%, 95%, 87%, respectively.

Conclusion

Machine learning assisted-DECT texture analysis can help distinguish different nodal pathology and normal nodes with a high accuracy.