Article Figures & Data
Figures
- Fig 1.
ROI placement. A 51-year-old man with an IP involving the right maxillary sinus. Axial T2-weighted fat-suppressed MR imaging pulse sequence demonstrates the manual placement of the largest rectangular ROI that would fit within the tumor margins on the axial image with the greatest tumor cross-sectional area. The inset image in the lower right corner is representative of the final 16 × 16 matrix that was derived from the ROI isocenter and served as the input for texture analysis.
- Fig 2.
Heat map showing MR imaging texture feature significance in distinguishing tumor type. Univariate analysis compared the pathology status (SCC versus IP) with MR imaging–texture features. Color maps show the false discovery rate–adjusted P values of a 2-sample t test. MR imaging contrasts (pulse sequences) are listed above the columns, and MR imaging–based texture features are listed in rows. DOST features 0–9 correspond with low-to-high frequency patterns. LBP 0–11 are the normalized bin counts in the LBP histogram. The reader is referred to the “Materials and Methods” section for additional details about the features.
- Fig 3.
Relative contributions to model accuracy. Of the 90.9% overall model accuracy for the training dataset, the bar graph demonstrates the accuracy attributable to PCs derived from T1C-GFB, T1-DOST, and T1-GLCM (upper panel). Across all texture algorithms, the contribution to total model accuracy was derived predominantly from T1C, with minor contributions from T1 and no input from T2 (lower panel).
- Fig 4.
PC loading. The model with the greatest accuracy for discriminating SCC from IP was derived from T1C-GFB, T1-GLCM, and T1-DOST texture features (right). For the individually specified texture features (left), PC loadings are graphically represented, and larger values in the PC loading indicate greater significance in the final model.
Tables
Study Group Sample Size Sex (Female/Male) Age (yr) Tumor Volume (cm3) Tumor Stageb T1 T2 T3 T4 IP training 16 4:12 58.0 ± 12.1 21.2 ± 17.7 1 3 10 2 IP validation 6 1:5 58.2 ± 15.3 22.0 ± 6.9 1 1 3 1 IP combined 22 5:17c 58.1 ± 13.1d 21.4 ± 15.5e 2 4 13 3 SCC training 17 4:13 54.0 ± 13.5 55.8 ± 40.5 0 1 4 12 SCC validation 7 1:6 54.6 ± 9.4 43.5 ± 27.9 0 1 2 4 SCC combined 24 5:19c 54.2 ± 12.5d 52.2 ± 37.7e 0 2 6 16 ↵a Data are presented separately for the training and validation sets and also as a single combined cohort for each tumor type. Age and tumor volume are presented as means.
↵b Tumor stage represents the Krouse staging system39 for IP and the American Joint Committee on Cancer staging40 for SCC.
↵c Fisher exact test, P = .578.
↵d Two-sample t test, P = .317.
↵e Two-sample t test, P = .001.
- Table 2:
Diagnostic performance of machine-learning classification in training and validation datasets
Tumor Type (Pathologic Diagnosis) Diagnostic Performance SCC IP Model prediction for training dataset SCC 16 2 Accuracy 90.9%a Sensitivity 94.1% IP 1 14 Specificity 87.5% PPV 88.9% Total 17 16 NPV 93.3% Model prediction for validation dataset SCC 6 1 Accuracy 84.6%a Sensitivity 85.7% IP 1 5 Specificity 83.3% PPV 85.7% Total 7 6 NPV 83.3% Model prediction for entire cohort SCC 22 3 Accuracy 89.1% Sensitivity 91.7% IP 2 19 Specificity 86.4% PPV 88.0% Total 24 22 NPV 90.5% Note:—NPV indicates negative predictive value; PPV, positive predictive value.
↵a With a 2-tailed test of population proportion, the accuracies for the training and validation datasets were not significantly different (P = .537).
- Table 3:
Diagnostic performance of texture analysis with machine learning compared with neuroradiologists' review for the differentiation of SCC from IPa
Analysis Method Accuracyb Sensitivity Specificity PPV NPV Texture analysis with machine learning 89.1% 91.7% 86.4% 88.0% 90.5% Neuroradiologists' review, ROI 56.5% (P = .0004) 54.2% 59.1% 59.1% 54.2% Neuroradiologists' review, tumor 73.9% (P = .060) 75.0% 72.7% 75.0% 72.7% Neuroradiologists' review, image 87.0% (P = .748) 91.7% 81.8% 84.6% 90.0% Note:—NPV indicates negative predictive value; PPV, positive predictive value.
↵a Results are shown for the entire cohort (22 IPs, 24 SCCs) and reflect the best classification model. The labels for the neuroradiologists' assessment indicate whether they reviewed the 16 × 16 ROI (ROI), tumor alone (tumor), or entire images (image).
↵b P values represent comparison of texture analysis with machine learning against each neuroradiologist's review using a 2-tailed test of population proportion.
Jump to section
Related Articles
Cited By...
- No citing articles found.