Performance of quantitative CT texture analysis in differentiation of gastric tumors

563 Background: In metastatic nonseminomatous testicular germ cell tumor (NSGCT), post-chemotherapy retroperitoneal lymph node dissection (PC-RPLND) is indicated for residual masses > 1 cm because of these 45% will be fibrosis/necrosis, 45% will be teratoma and 15% will be viable malignancy. There is no imaging test that reliably distinguishes lymph nodes (LNs) with tumor (teratoma or malignancy) from LNs with fibrosis/necrosis. We evaluated whether quantitative CT texture analysis (TA) could make this differentiation. Methods: Pre- and post-chemotherapy CTs (all same phase and slice thickness) were reviewed in 22 NSGCT patients with RP LNs > 1 cm post chemotherapy. After manual segmentation of RP LNs on a 3D workstation, 187 TA metrics were derived, using 2D/3D gray-level co-occurrence matrix (GLCM), 2D/3D gray-level difference matrix (GLDM), and spectral analysis. Metrics were derived 2 ways: from post-chemotherapy CTs alone, and also as a difference between pre- and post-chemotherapy CTs, resulting in 374 metrics. PC-RPLND pathology was correlated with CT data at 88 LN stations in these 22 patients. Results: 15 imaging metrics showed a significant difference (p ≤ 0.05) between LN stations with only fibrosis/necrosis and those with teratoma or viable tumor. Seven were derived from the difference between pre- and post-chemotherapy CTs: 4 using a 2D GLCM (coronal standard deviation, coronal square root of variance, coronal mean, and coronal sum of average), and 3 using a 2D GLDM (axial variance, axial square root of variance, and coronal variance). The other 8 were derived from post-chemotherapy CTs alone: 7 using a 2D GLCM (sagittal square root of variance, sagittal standard deviation, coronal square root of variance, coronal mean, coronal standard deviation, coronal sum of average, and coronal entropy) and 1 using a 2D GLDM (sagittal sum entropy). Conclusions: CT TA shows promise in differentiating necrosis from teratoma or viable tumor in RP LNs in post-chemotherapy NSGCT. A larger study is needed to further test this method, towards a long-term goal of potentially allowing some patients to avoid PC-RPLND.

Download Full-text

Development of CT texture analysis in COPD and association with visual scoring and DLCO

10.32920/ryerson.14635971.v1 ◽

2021 ◽

Author(s):

V. Lui ◽

W. C. Tan ◽

J. C. Hogg ◽

H. O. Coxson ◽

M. Kirby

Keyword(s):

Lung Function ◽

Healthy Volunteers ◽

Texture Analysis ◽

Texture Features ◽

Visual Assessment ◽

Quantitative Ct ◽

Visual Scoring ◽

Ct Texture Analysis

Objectives: • To determine if CT texture features, such as GLCM and FD, can differentiate patients with COPD from healthy volunteers, and are related to lung function • To determine if CT texture features are association with qualitative visual scoring • To determine if CT texture features are significantly associated with COPD outcomes, independent of qualitative scoring and standard quantitative CT emphysema measurements Hypothesis: • CT texture features can be developed to objectively aid in quantifying the severity of emphysema, and may provide information complementary to qualitative visual assessment

Download Full-text

Quantitative CT texture analysis for evaluating histologic grade of urothelial carcinoma

Abdominal Radiology ◽

10.1007/s00261-016-0897-2 ◽

2016 ◽

Vol 42 (2) ◽

pp. 561-568 ◽

Cited By ~ 32

Author(s):

Gu-Mu-Yang Zhang ◽

Hao Sun ◽

Bing Shi ◽

Zheng-Yu Jin ◽

Hua-Dan Xue

Keyword(s):

Urothelial Carcinoma ◽

Texture Analysis ◽

Histologic Grade ◽

Quantitative Ct ◽

Ct Texture Analysis

Download Full-text

Quantitative CT texture analysis in predicting PD-L1 expression in locally advanced or metastatic NSCLC patients

La radiologia medica ◽

10.1007/s11547-021-01399-9 ◽

2021 ◽

Author(s):

Stefano Bracci ◽

Miriam Dolciami ◽

Claudio Trobiani ◽

Antonella Izzo ◽

Angelina Pernazza ◽

...

Keyword(s):

Texture Analysis ◽

Advanced Nsclc ◽

Validation Cohort ◽

Locally Advanced ◽

Area Under The Curve ◽

Gray Level ◽

Quantitative Ct ◽

Training Cohort ◽

Contrast Enhanced Ct ◽

Ct Texture Analysis

Abstract Purpose The assessment of Programmed death-ligand 1 (PD-L1) expression has become a game changer in the treatment of patients with advanced non-small cell lung cancer (NSCLC). We aimed to investigate the ability of Radiomics applied to computed tomography (CT) in predicting PD-L1 expression in patients with advanced NSCLC. Methods By applying texture analysis, we retrospectively analyzed 72 patients with advanced NSCLC. The datasets were randomly split into a training cohort (2/3) and a validation cohort (1/3). Forty radiomic features were extracted by manually drawing tumor volumes of interest (VOIs) on baseline contrast-enhanced CT. After selecting features on the training cohort, two predictive models were created using binary logistic regression, one for PD-L1 values ≥ 50% and the other for values between 1 and 49%. The two models were analyzed with ROC curves and tested in the validation cohort. Results The Radiomic Score (Rad-Score) for PD-L1 values ≥ 50%, which consisted of Skewness and Low Gray-Level Zone Emphasis (GLZLM_LGZE), presented a cut-off value of − 0.745 with an area under the curve (AUC) of 0.811 and 0.789 in the training and validation cohort, respectively. The Rad-Score for PD-L1 values between 1 and 49% consisted of Sphericity, Skewness, Conv_Q3 and Gray Level Non-Uniformity (GLZLM_GLNU), showing a cut-off value of 0.111 with AUC of 0.763 and 0.806 in the two population, respectively. Conclusion Rad-Scores obtained from CT texture analysis could be useful for predicting PD-L1 expression and guiding the therapeutic choice in patients with advanced NSCLC.

Download Full-text