Do Radiologists Report the TNM Staging in Radiology Reports for Head and Neck Cancers? A National Survey Study

Communication in radiology: evaluation of terminology and TNM descriptor use at a cancer center

Objective

The purpose of our study was to evaluate the transmission of information from radiologists to physicians, focusing on the level of certainty and the use of imaging descriptors from the tumor-node-metastasis (TNM) staging system.

Materials and Methods

Radiologists (n = 56) and referring physicians (n = 50) participated in this questionnaire-based, singlecenter study, conducted between March 20, 2020, and January 21, 2021. Participants were presented with terms commonly used by the radiologists at the institution and were asked to order them hierarchically in terms of the level of certainty they communicate regarding a diagnosis, using a scale ranging from 1 (most contrary to) to 10 (most favoring). They then assessed TNM system descriptors and their interpretation. Student's t-tests and the kappa statistic were used in order to compare the rankings of the terms of certainty. Items related to T and N staging were analyzed by Fisher's exact test. The confidence level was set to 97% (p < 0.03).

Results

Although overall agreement among the radiologists and referring physicians on term ranking was poor (kappa = 0.10- 0.35), the mean and median values for the two groups were similar. Most of the radiologists and referring physicians (67% and 86%, respectively) approved of the proposal to establish a standard lexicon. Such a lexicon, based on the participant responses, was developed and graphically represented. Regarding the TNM system descriptors, there were significant differences between the two groups in the reporting of lymph node numbers, of features indicating capsular rupture, and of vessel wall irregularities, as well as in the preference for clear descriptions of vascular involvement.

Conclusion

Our findings indicate that ineffective communication and differences in report interpretation between radiologists and referring physicians are still prevalent in the fields of radiology and oncology. Efforts to gain a better understanding of those impediments might improve the objectivity of reporting and the quality of care.

Common Data Elements in Head and Neck Radiology Reporting

Radiologists must convert the complex information in head and neck imaging into text reports that can be understood and used by clinicians, patients, and fellow radiologists for patient care, research, and quality initiatives. Common data elements in reporting, through use of defined questions with constrained answers and terminology, allow radiologists to incorporate best practice standards and improve communication of information regardless of individual reporting style. Use of common data elements for head and neck reporting has the potential to improve outcomes, reduce errors, and transition data consumption not only for humans but future machine learning systems.

One-slice CT image based kernelized radiomics model for the prediction of low/mid-grade and high-grade HNSCC

An accurate grade prediction can help to appropriate treatment strategy and effective diagnosis to Head and neck squamous cell carcinoma (HNSCC). Radiomics has been studied for the prediction of carcinoma characteristics in medical images. The success of previous researches in radiomics is attributed to the availability of annotated all-slice medical images. However, it is very challenging to annotate all slices, as annotating biomedical images is not only tedious, laborious, and time consuming, but also demanding of costly, specialty-oriented skills, which are not easily accessible. To address this problem, this paper presents a model to integrate radiomics and kernelized dimension reduction into a single framework, which maps handcrafted radiomics features to a kernelized space where they are linearly separable and then reduces the dimension of features through principal component analysis. Three methods including baseline radiomics models, proposed kernelized model and convolutional neural network (CNN) model were compared in experiments. Results suggested proposed kernelized model best fit in one-slice data. We reached AUC of 95.91 % on self-made one-slice dataset, 67.33 % in predicting localregional recurrence on H&N dataset and 64.33 % on H&N1 dataset. While all other models were <76 %, <65 %, and <62 %. Though CNN model reached an incredible performance when predicting distant metastasis on H&N (AUC 0.88), model faced serious problem of overfitting in small datasets. When changing all-slice data to one-slice on both H&N and H&N1, proposed model suffered less loss on AUC (<1.3 %) than any other models (>3 %). These proved our proposed model is efficient to deal with the one-slice problem and makes using one-slice data to reduce annotation cost practical. This is attributed to the several advantages derived from the proposed kernelized radiomics model, including (1) the prior radiomics features reduced the demanding of huge amount of data and avoided overfitting; (2) the kernelized method mined the potential information contributed to predict; (3) generating principal components in kernelized features reduced redundant features.

Magnetic resonance imaging based radiomics signature for the preoperative discrimination of stage I-II and III-IV head and neck squamous cell carcinoma

PURPOSE

This study aimed to investigate the predictive ability of magnetic resonance imaging (MRI) based radiomics signature for the preoperative staging in HNSCC.

METHODS

This study involved127 consecutive patients (training cohort: n = 85; testing cohort, n = 42) with stage I-IV HNSCC. A total of 970 radiomics features were extracted from T2-weighted (T2W) (n = 485) and contrast-enhanced T1-weighted (ceT1W) (n = 485) MRI for each case. Radiomics signatures were constructed with least absolute shrinkage and selection operator (LASSO) logistic regression. Associations between radiomics signatures and HNSCC staging were explored. Areas under the receiver operating characteristic curve (AUC) and classification performance of radiomics signatures were determined and compared with those of the visual assessment.

RESULTS

Ten features from T2W images, six from ceT1W images, and six from combined T2W and ceT1W images were selected by LASSO logistic regression. The three radiomics signatures of stage III-IV HNSCC were significantly higher than that for stage I-II in both cohorts (all P < 0.05). The radiomics signatures from ceT1W and combined images performed well in the discrimination of stage I-II and III-IV HNSCC, with AUCs of 0.828 and 0.850 in the training cohort, and AUCs of 0.853 and 0.849 in the testing cohort. Based on the cut-off value of the training cohort, the radiomics signature from combined images achieved best classification performance in both cohorts, with accuracies of 0.788 and 0.857, sensitivities of 0.836 and 0.885, and specificities of 0.700 and 0.813. Significant differences in accuracy and sensitivity were found between the radiomics signature from combined images and the visual assessment of the radiologists in the training cohort.

CONCLUSION

Radiomics signature based on MRI could discriminate stage I-II from stage III-IV HNSCC, which may serve as a complementary tool for preoperative staging.

Pathological tumor volume predicts survival outcomes in oral squamous cell carcinoma

The present study examined whether the pathological tumor volume (PTV) was correlated with the survival outcomes in patients with oral squamous cell carcinoma (SCC) and clinical lymph node metastasis. Forty-seven patients who underwent radical surgery without preoperative treatment were enrolled. The PTV of the primary tumor, which was surgically resected without preoperative treatment, was calculated based on the diameters in three dimensions. A survival analysis was performed using a Cox proportional hazards model. A PTV of ≥18 cm³ was significantly correlated with shorter overall survival (P<0.01) and local recurrence-free survival (P<0.01) in a univariate analysis. A multivariate analysis with adjustment for the pathological stage (stage I–II/III–IV), primary site (tongue/others) and positive surgical margin and/or extracapsular extension (absent/present) showed that a PTV of ≥18 cm³ was significantly correlated with shorter overall survival (P<0.01) and local recurrence-free survival (P<0.01). The present findings suggested that PTV in oral SCC provides a prognostic parameter that may predict shorter or longer overall and local recurrence-free survival.

[Imaging for surgical planning : Tumor surgery including reconstructive procedures]

Alongside clinical investigation, imaging is an important diagnostic modality for guiding treatment decisions and particularly for surgical planning in head and neck cancer. The significance and type of imaging depends on localization of the primary tumor. Beside the primary tumor, each imaging procedure must also include the lymph nodes, in order to develop an overall concept of surgical treatment. In addition to the superficial growth of a tumor, it is of utmost importance that its infiltration also be detected by imaging, in order to define the extent of tumor resection and identify potentially necessary reconstructive procedures. The type of imaging needed to understand tumor localization and size, e.g., CT or MRI, depends on the region. In some cases the methods are complementary.