Dataset Information

Squamous Cell Carcinoma and Lymphoma of the Oropharynx: Differentiation Using a Radiomics Approach.

ABSTRACT: The purpose of this study was to evaluate the diagnostic performance of magnetic resonance (MR) radiomics-based machine learning algorithms in differentiating squamous cell carcinoma (SCC) from lymphoma in the oropharynx. MR images from 87 patients with oropharyngeal SCC (n=68) and lymphoma (n=19) were reviewed retrospectively. Tumors were semi-automatically segmented on contrast-enhanced T1-weighted images registered to T2-weighted images, and radiomic features (n=202) were extracted from contrast-enhanced T1- and T2-weighted images. The radiomics classifier was built using elastic-net regularized generalized linear model analyses with nested five-fold cross-validation. The diagnostic abilities of the radiomics classifier and visual assessment by two head and neck radiologists were evaluated using receiver operating characteristic (ROC) analyses for distinguishing SCC from lymphoma. Nineteen radiomics features were selected at least twice during the five-fold cross-validation. The mean area under the ROC curve (AUC) of the radiomics classifier was 0.750 [95% confidence interval (CI), 0.613-0.887], with a sensitivity of 84.2%, specificity of 60.3%, and an accuracy of 65.5%. Two human readers yielded AUCs of 0.613 (95% CI, 0.467-0.759) and 0.663 (95% CI, 0.531-0.795), respectively. The radiomics-based machine learning model can be useful for differentiating SCC from lymphoma of the oropharynx.

SUBMITTER: Bae S

PROVIDER: S-EPMC7515782 | biostudies-literature | 2020 Oct

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Squamous Cell Carcinoma and Lymphoma of the Oropharynx: Differentiation Using a Radiomics Approach.

Bae Sohi S Choi Yoon Seong YS Sohn Beomseok B Ahn Sung Soo SS Lee Seung Koo SK Yang Jaemoon J Kim Jinna J

Yonsei medical journal 20201001 10

The purpose of this study was to evaluate the diagnostic performance of magnetic resonance (MR) radiomics-based machine learning algorithms in differentiating squamous cell carcinoma (SCC) from lymphoma in the oropharynx. MR images from 87 patients with oropharyngeal SCC (n=68) and lymphoma (n=19) were reviewed retrospectively. Tumors were semi-automatically segmented on contrast-enhanced T1-weighted images registered to T2-weighted images, and radiomic features (n=202) were extracted from contr ...[more]

PMID: 32975065

Similar Datasets

Project description:PurposePatients with cancers of oropharynx have a favorable prognosis and are an ideal candidate for adaptive therapy. A replan to improve coverage or escalate/de-escalate dose based on morphological information alone may not be adequate as the grossly involved lymph nodes (LNs) of a subset of these patients tend to become cystic and often do not regress. Functional adaptation may be a better approach when considering replanning for these patients. The purpose of this study was to evaluate the weekly trends in treatment related morphological and physiological changes for these LNs using diffusion-weighted MRI (DW-MRI) and evaluate its implications for adaptive replanning.MethodsTen patients with histologically proven oropharynx HNSCC undergoing concurrent chemoradiation were analyzed in this study. MR imaging protocol included axial T1w, T2w, and DW-MRI using a 3 T Philips MR scanner. The patients were scanned weekly in radiation treatment planning position using a 16 element phased-array anterior coil and a 44 element posterior coil. A total of 65 DWI and T2w scans were analyzed. DWI was performed using an optimized single-shot echo planar imaging sequence (TR/TE = 5000/65 ms, slice thickness = 5 mm; slices = 28; b values = 0 and 800 s/mm(2)). Quantification of the DW-MRI images was performed by calculating the apparent diffusion coefficient (ADC). T2w and DWI scans were imported into the Eclipse treatment planning system and gross tumor volumes (GTVs) corresponding to grossly involved LNs were contoured on each axial slice by physician experts. An attempt was made to remove any cystic or necrotic components so that the ADC analysis was of viable tumor only. A pixel-by-pixel fit of signal intensities within the GTVs was performed assuming monoexponential behavior. From each GTV histogram mean, median, standard deviation, skewness, and kurtosis were calculated. Absolute and percent change in weekly ADC histogram parameters and percent change in T2w GTV were also calculated.ResultsFor all nodes, an immediate change in ADC was observed during first 2-3 weeks after which ADC values either continued to increase or plateaued. A few nodal volumes had a slightly decreased ADC value during later weeks. Percent increase in median ADC from weeks 1 to 6 with respect to baseline was 14%, 25%, 41%, 42%, 45%, and 58%. The corresponding change in median T2 volumes was 8%, 10%, 16%, 22%, 40%, and 42%, respectively. The ADC distribution of the viable tumors was initially highly kurtotic; however, the kurtosis decreased as treatment progressed. The ADC distribution also showed a higher degree of skewness in the first 2 weeks, progressively becoming less skewed as treatment progressed so as to slowly approach a more symmetric distribution.ConclusionsPhysiological changes in LNs represented by changes in ADC evaluated using DW-MRI are evident sooner than the morphological changes calculated from T2w MRI. The decisions for adaptive replanning may need to be individualized and should be based primarily on tumor functional information. The authors' data also suggest that for many patients, week 3 maybe the optimal time to intervene and replan. Larger studies are needed to confirm their findings.

Dataset Information

Squamous Cell Carcinoma and Lymphoma of the Oropharynx: Differentiation Using a Radiomics Approach.

Publications

Squamous Cell Carcinoma and Lymphoma of the Oropharynx: Differentiation Using a Radiomics Approach.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets