Dataset Information

Revealing Tumor Habitats from Texture Heterogeneity Analysis for Classification of Lung Cancer Malignancy and Aggressiveness.

ABSTRACT: We propose an approach for characterizing structural heterogeneity of lung cancer nodules using Computed Tomography Texture Analysis (CTTA). Measures of heterogeneity were used to test the hypothesis that heterogeneity can be used as predictor of nodule malignancy and patient survival. To do this, we use the National Lung Screening Trial (NLST) dataset to determine if heterogeneity can represent differences between nodules in lung cancer and nodules in non-lung cancer patients. 253 participants are in the training set and 207 participants in the test set. To discriminate cancerous from non-cancerous nodules at the time of diagnosis, a combination of heterogeneity and radiomic features were evaluated to produce the best area under receiver operating characteristic curve (AUROC) of 0.85 and accuracy 81.64%. Second, we tested the hypothesis that heterogeneity can predict patient survival. We analyzed 40 patients diagnosed with lung adenocarcinoma (20 short-term and 20 long-term survival patients) using a leave-one-out cross validation approach for performance evaluation. A combination of heterogeneity features and radiomic features produce an AUROC of 0.9 and an accuracy of 85% to discriminate long- and short-term survivors.

SUBMITTER: Cherezov D

PROVIDER: S-EPMC6418269 | biostudies-literature | 2019 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Revealing Tumor Habitats from Texture Heterogeneity Analysis for Classification of Lung Cancer Malignancy and Aggressiveness.

Cherezov Dmitry D Goldgof Dmitry D Hall Lawrence L Gillies Robert R Schabath Matthew M Müller Henning H Depeursinge Adrien A

Scientific reports 20190314 1

We propose an approach for characterizing structural heterogeneity of lung cancer nodules using Computed Tomography Texture Analysis (CTTA). Measures of heterogeneity were used to test the hypothesis that heterogeneity can be used as predictor of nodule malignancy and patient survival. To do this, we use the National Lung Screening Trial (NLST) dataset to determine if heterogeneity can represent differences between nodules in lung cancer and nodules in non-lung cancer patients. 253 participants ...[more]

PMID: 30872600

Similar Datasets

Project description:There is no study that investigates the potential correlation between the heterogeneity obtained from texture analysis of medical images and the heterogeneity observed from histopathological findings. We investigated whether texture analysis of magnetic resonance images correlates with histopathological findings.Seventy-five patients with estrogen receptor positive invasive ductal carcinoma who underwent preoperative breast magnetic resonance imaging (MRI) were included. Tumor entropy and uniformity were determined on T2- and contrast-enhanced T1-weighted subtraction images under different filter levels. Two pathologists evaluated the detailed histopathological findings of the tumors including tumor cellularity, dominant stroma type, central scar, histologic grade, extensive intraductal component (EIC), and lymphovascular invasion. Entropy and uniformity values on both T2- and contrast-enhanced T1-weighted subtraction images were compared with detailed pathological findings.In a multivariate analysis, entropy significantly increased only on unfiltered T2-weighted images (P?=?0.013). Tumor cellularity and predominant stroma did not affect the uniformity or entropy on both T2- and contrast-enhanced T1-weighted subtraction images. High histologic grades showed increased uniformity and decreased entropy on contrast-enhanced T1-weighted subtraction images, whereas the opposite tendency was observed on T2-weighted images. Invasive ductal carcinoma with an EIC or lymphovascular invasion only affected the contrast-enhanced T1-weighted subtraction images, through increased uniformity and decreased entropy. The best uniformity results were recorded on T2- and contrast-enhanced T1-weighted subtraction images at a filter level of 0.5. Entropy showed the best results at a filter level of 0.5 on contrast-enhanced T1-weighted subtraction images. However, on T2-weighted images, an ideal model was achieved on unfiltered images.MRI texture analysis correlated with pathological tumor heterogeneity.

Dataset Information

Revealing Tumor Habitats from Texture Heterogeneity Analysis for Classification of Lung Cancer Malignancy and Aggressiveness.

Publications

Revealing Tumor Habitats from Texture Heterogeneity Analysis for Classification of Lung Cancer Malignancy and Aggressiveness.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets