3.1. Patient characteristics
A total of 21 PIPT patients were enrolled in this study, including 15 males and 6 females (maximum age 74 years, minimum age 44 years, median age 58 years). In the second step, 21 patients with peripheral lung cancer were randomly collected, including 13 males and 8 females (maximum age 78 years, minimum age 44 years, median age 67 years). 14 were diagnosed with adenocarcinoma, 6 with squamous cell carcinoma and 1 with atypical carcinoid
3.2. Feature results
In this study, a total of 435 radiomics features were extracted (composed of 368 texture features, 18 shape features and 49 tumor intensity features), According to the different calculation methods, these features were divided into 5 categories. Among them, the gray-level co-occurrence matrix (GLCM,22 parent features, 330 child features), gray-level run length matrix (GLRLM, 11 parent features, 33 child features), intensity histogram (IH, 9 parent features,49 child features), gray-level neighbor intensity difference matrix(GLNIDM, 5 parent features, none), shape(null,18 patient features, none).
3.3. Statistical results
All statistical differences of PIPT and peripheral lung cancer were tested by Mann-Whitney U test. A total of 32 feature differences were found to be statistically significant, of which the GLCM has 24 child features, which belong to 5 parent features, respectively, 1 parent feature in IH, 2 parent features in NIDM, 5 parent features in shape, and their child features are all 0. The 32 radiomics features with statistically significant differences are shown in Table 1.
Binary logistic regression model analysis showed that 25 of the 32 child features were significantly different and could be used to distinguish PIPT from peripheral lung cancer. These 25 child features respectively belong to GLCM(parent feature correlation (n=9) , parent feature information measure corr1 (n=5) and parent feature information measure corr2 (n= 4) )、IH(parent feature range(n=0)) 、GLNIDM(parent feature texture strength(n=0)) and shape(parent feature compactness2(n=0), Roundness(n=0), parent feature spherical disproportion(n=0), parent feature sphericity (n=0) and parent feature surface area density(n=0)),respectively. (Table 1)
Table 1 Feature parameters differentiating between pulmonary inflammatory pseudotumor and peripheral lung cancer
Category
|
Parent Feature
|
Child Feature
|
P1 value
|
P2 value
|
GLCM
|
Cluster Prominence
|
135-1Cluster Prominence
|
0.0325
|
0.2731
|
Correlation
|
333-1 Correlation
|
0.0111
|
0.0002
|
333-4 Correlation
|
0.0497
|
0.0035
|
45-1 Correlation
|
0.0128
|
0.0007
|
45-4 Correlation
|
0.0020
|
0.0021
|
45-7 Correlation
|
0.0025
|
0.0079
|
90-1 Correlation
|
0.0018
|
0.0001
|
90-4 Correlation
|
0.0028
|
0.0014
|
90-7 Correlation
|
0.0086
|
0.0010
|
135-7 Correlation
|
0.0265
|
0.0368
|
Information Measure Corr1
|
333-1 Information Measure Corr1
|
0.0268
|
0.0042
|
333-4 Information Measure Corr1
|
0.0157
|
0.0604
|
0-1 Information Measure Corr1
|
0.0128
|
0.0384
|
45-1 Information Measure Corr1
|
2.97E-5
|
0.0009
|
90-1 Information Measure Corr1
|
5.7E-5
|
0.0001
|
90-4 Information Measure Corr1
|
0.0489
|
0.1070
|
135-1 Information Measure Corr1
|
0.0137
|
0.0079
|
Information Measure Corr2
|
333-1 Information Measure Corr2
|
0.0442
|
0.0070
|
333-4 Information Measure Corr2
|
0.0442
|
0.0955
|
0-1 Information Measure Corr2
|
0.0497
|
0.0978
|
45-1 Information Measure Corr2
|
0.0020
|
0.0184
|
90-1 Information Measure Corr2
|
0.0083
|
0.0065
|
135-1 Information Measure Corr2
|
0.0168
|
0.0273
|
Inverse Diff Moment Norm
|
135-7 Inverse Diff moment Norm
|
0.0391
|
0.2151
|
IH
|
Range
|
None
|
0.0169
|
0.0116
|
GLNIDM
|
Complexity
|
None
|
0.0268
|
0.0595
|
Texture Strength
|
None
|
0.0089
|
0.0289
|
SHAPE
|
Compactness2
|
None
|
0.0017
|
0.0027
|
Roundness
|
None
|
0.0083
|
0.0051
|
Spherical Disproportion
|
None
|
0.0017
|
0.0010
|
Sphericity
|
None
|
0.0017
|
0.0014
|
|
Surface Area Density
|
None
|
0.0207
|
0.0443
|
Note: gray-level co-occurrence matrix, GLCM; Intensity Histogram, IH; neighbor intensity difference matrix, GLNIDM; number1-number2-feature (angle,number1; distance,number2); the significant difference index of Mann-Whitney U test; P1 value; the significant difference index of Binary logistic regression, P2 value; indicates a significant difference (p<0.05).
ROC curves of 25 features were performed to evaluate the ability of features differentiating peripheral lung cancer from PIPT. The curves (AUC < 0.7) was abandoned in this study, because of its limited discriminant ability. Finally, A total of 23 ROC curves of features were obtained in this study. In addition, we calculated the average value of the features at the same angle and different distance and drew ROC curves, which were curve Mean1 and Mean2. we also calculated the average value of the features at different angles and different distances and the drew of ROC curves were curve mean3, mean4, mean5, respectively. The P-values of statistical differences among ROC were 0.0499 (F11, F23), 0.0472 (F12, F13) and 0.0250 (F13, Mean4), and the others were 0.5908 0.2803. All ROC curves are shown in Figure1.
At the same time, we calculate the AUC, sensitivity, truncated value, specificity and Youden index. For features with discriminating ability and features after averaging the feature values, we calculated their interquartile range (IQR) and median values, respectively. The specific values of all statistical parameters differentiating between peripheral lung cancer and PIPT are shown in Table. 2.
Table. 2 Statistical parameters differentiating between peripheral lung cancer and pulmonary inflammatory pseudotumor
Feature
|
Peripheral Lung Cancer
|
PIPT
|
AUC
|
Sensitivity(%)
|
Specificity(%)
|
Truncated-Value
|
Younden-Index
|
IQR
|
median
|
IQR
|
median
|
F1
|
0.0429
|
0.7334
|
0.0874
|
0.7913
|
0.730
|
80.95
|
76.19
|
0.7657
|
0.5714
|
F2
|
0.1056
|
0.6783
|
0.1218
|
0.7846
|
0.726
|
80.95
|
66.67
|
0.7401
|
0.4762
|
F3
|
0.1399
|
0.0738
|
0.1480
|
0.2719
|
0.800
|
90.00
|
70.59
|
0.2047
|
0.6059
|
F4
|
0.1889
|
-0.0284
|
0.2054
|
0.1434
|
0.806
|
88.89
|
68.75
|
0.0246
|
0.5764
|
F5
|
0.0682
|
0.7960
|
0.0773
|
0.8801
|
0.782
|
90.48
|
71.43
|
0.8629
|
0.6190
|
F6
|
0.1965
|
0.2102
|
0.1405
|
0.3946
|
0.774
|
90.48
|
65.00
|
0.3589
|
0.5548
|
F7
|
0.1370
|
0.0223
|
0.1438
|
0.1954
|
0.754
|
84.21
|
72.22
|
0.1234
|
0.5643
|
F8
|
0.1467
|
-0.0576
|
0.2119
|
0.410
|
0.718
|
73.68
|
76.47
|
-0.0316
|
0.5015
|
Mean1
|
0.1500
|
0.2280
|
0.1403
|
0.4112
|
0.800
|
95.24
|
71.43
|
0.3512
|
0.6667
|
Mean2
|
0.1377
|
0.3465
|
0.2000
|
0.5030
|
0.728
|
85.71
|
66.67
|
0.4526
|
0.5238
|
Mean3
|
0.1264
|
0.2513
|
0.1745
|
0.3979
|
0.744
|
85.71
|
71.43
|
0.3100
|
0.5714
|
F9
|
0.0436
|
-0.2981
|
0.0790
|
-0.3183
|
0.701
|
93.48
|
47.62
|
-0.3395
|
0.4762
|
F10
|
0.0436
|
-0.3821
|
0.1112
|
-0.4560
|
0.726
|
95.24
|
57.14
|
-0.4461
|
0.5238
|
F11
|
0.0409
|
-0.2944
|
0.1154
|
0.3496
|
0.878
|
80.95
|
90.48
|
-0.3106
|
0.7143
|
F12
|
0.0506
|
-0.3605
|
0.0892
|
-0.4266
|
0.864
|
71.43
|
100.00
|
-0.3813
|
0.7143
|
F13
|
0.0389
|
-0.2849
|
0.1490
|
-0.3602
|
0.723
|
95.24
|
52,38
|
-0.3602
|
0.4762
|
Mean4
|
0.0263
|
-0.3250
|
0.0788
|
-0.3887
|
0.837
|
90.48
|
66.67
|
-0.3597
|
0.5714
|
F14
|
0.0428
|
0.8997
|
0.0692
|
0.9302
|
0.780
|
85.71
|
66.67
|
0.9175
|
0.5238
|
F15
|
0.0190
|
0.9330
|
0.0450
|
0.9623
|
0.739
|
85.71
|
66.71
|
0.9567
|
0.5238
|
F16
|
0.0378
|
0.8930
|
0.0788
|
0.9202
|
0.717
|
71.43
|
71.43
|
0.8975
|
0.4286
|
Mean5
|
0.0244
|
0.9065
|
0.0625
|
0.9419
|
0.751
|
85.71
|
61.91
|
0.9252
|
0.4762
|
F17
|
115
|
346
|
189
|
494
|
0.717
|
76.19
|
71.43
|
400.0000
|
0.4762
|
F18
|
24.4979
|
30.8545
|
123.2077
|
101.5105
|
0.737
|
76.19
|
80.95
|
42.1144
|
0.5714
|
F19
|
0.1460
|
0.5928
|
0.4127
|
0.2817
|
0.785
|
95.24
|
71.43
|
0.3259
|
0.6667
|
F20
|
0.1063
|
0.3569
|
0.2001
|
0.2136
|
0.739
|
85.71
|
57.14
|
0.2454
|
0.4286
|
F21
|
0.0953
|
1.1904
|
0.5439
|
1.5253
|
0.785
|
95.24
|
71.43
|
1.3783
|
0.6667
|
F22
|
0.0680
|
0.8400
|
0.2671
|
0.6556
|
0.785
|
95.24
|
71.43
|
0.6882
|
0.6667
|
F23
|
1.7698
|
2.9145
|
5.3141
|
4.5316
|
0.710
|
66.67
|
76.19
|
2.9718
|
0.4286
|
Note: F1-333-1 Correlation; F2-45-1 Correlation; F3-45-4 Correlation; F4-45-7 Correlation; F5-90-1 Correlation;F6-90-4 Correlation; F7-90-7 Correlation; F8-135-7 Correlation; F9-333-1 Information Measure Corr1; F10-0-1 Information- Measure Corr1; F11-45-1 Information Measure Corr1; F12-90-1 Information Measure Corr1; F13-135-1 Information Measure Corr1; F14-45-1 Information Measure Corr2; F15-90-1 Information Measure Corr2; F16-135-1 Information Measure Corr2; F17-range; F18-Texture Strength; F19-Compactness2; F20-Roundness; F21-Spherical Disproportion; F22-Sphericity; F23-Surface Area Density; Mean1- Mean(F2+…+F4); Mean2- Mean(F5+…+F7); Mean3- Mean(F2+…+F9); Mean4- Mean(F9+…+F13); Mean5- Mean(F14+…+F16); IQR- indicates interquartile range