The purpose of this study was to explore the effects of CT slice thickness, reconstruction algorithm, and radiation dose on quantification of CT features to characterize lung nodules using a chest phantom. Spherical lung nodule phantoms of known densities (−630 and + 100 HU) were inserted into an anthropomorphic thorax phantom. CT scan was performed ten times with relocations. CT data were reconstructed using 12 different imaging settings; three different slice thicknesses of 1.25, 2.5, and 5.0 mm, two reconstruction kernels of sharp and standard, and two radiation dose of 30 mAs and 12 mAs. Lesions were segmented using a semiautomated method. Twenty representative CT quantitative features representing CT density and texture were compared using multiple regression analysis. In 100 HU nodule phantoms, 18 and 19 among 20 computer features showed significant difference between different mAs and reconstruction algorithms, respectively (
Biomedical images may contain information that reflects underlying pathophysiology of many diseases. Nowadays, based on highthroughput computing, extracting many quantitative features from tomographic images is possible. Therefore, many studies have focused on how to convert information on images to quantitative computer features. The conversion of digital medical images into highdimensional computer data is known as radiomics. Radiomics is also a decision support tool, and it can involve combining radiomic data with other patient characteristics such as survival and disease phenotype [
Computer features based on computed tomography (CT) histogram and texture are most frequently used for the differential diagnosis of various cancers including lung cancer [
Recently, the number of researches on radiomics or deep learning is increasing [
Therefore, the purpose of this study was to analyze the effect of various scan parameters on the quantitative CT features of lung nodule phantoms. We evaluated the effect of different CT slice thicknesses, mAs, and reconstruction algorithms on 3dimensional computer features including CT histogram, graylevel cooccurrence matrix (GLCM), and graylevel run length matrix (GLRLM).
In this study, we used an anthropomorphic thorax phantom (KYOTO KAGAKU co., Kyoto, Japan) and nodule phantoms of two different attenuation values (100 Hounsfield Unit (HU) and −630 HU) (Figure
Anthropomorphic thorax phantom and nodule phantoms: (a) chest phantom, (b) nodule phantoms of 100 HU, (c) nodule phantoms of −630 HU, (d) an example of a 100 HU nodule phantom attached to the pulmonary vasculature, and (e) an example of a −630 HU nodule phantom attached to the pulmonary vasculature.
CT scan was performed by using a 64 channel multidetector row CT scanner (GE Discovery CT 750 HD; GE Healthcare, USA). The CT scan parameters were 120 kVp, 64 × 0.625 collimator configuration, and pitch of 0.984 : 1. The raw data were then reconstructed using 12 different combinations of scan parameters; slice thicknesses (1.25, 2.5, 5.0), mAs (30, 120), and reconstruction algorithms (lung, standard). In each combination of scan parameter, CT scan was repeated 10 times with relocation of nodule phantoms (Figures
CT images of a 12 mm sized 100 HU nodule phantom in 12 different scan parameters.
CT images of a 12 mm sized −630 HU nodule phantom in 12 different scan parameters.
In general, it is known that definition and noise decreases as slice thickness increases, noise decreases as mAs increases, and the standard algorithm has less definition and noise than the lung algorithm. Figure
Relationship between scan parameters and image quality: (a) relationship between slice thickness variation and noise, (b) relationship between slice thickness variation and definition, (c) relationship between mAs variation and noise, and (d) relationship between reconstruction algorithms and noise definition.
In this study, we used an inhouse software for the computerized analysis of CT images. This software was developed by using Microsoft Visual Studio (Ver. 2010, Microsoft, Redmond, WA, USA), ITK (Insight Segmentation and Registration Toolkit, Kitware Inc., NY, USA), and VTK (Visualization Toolkit, Kitware Inc., NY, USA). For the first step of the volume measurement of the nodules, the entire tumor mass was separated from surrounding anatomic structures by using a semiautomated segmentation algorithm developed in the Laboratory for Computational Image Analysis in the Department of Biomedical Engineering of Gachon University College of Medicine. This algorithm combined the image analysis techniques of seed regiongrowing algorithm [
In this study, we used the features mainly used in lung nodule analysis. Among the various features, 20 radiomic features were selected based on several related papers that performed lung nodule analysis [
In this study, we quantized the pixel values of 4,096 gray colors into 256 gray colors, and texture features are extracted based on the discretized pixel values [
Definition of the 20 computer features.
Features  Definition  Description  

Histogram  Mean 

The mean value of the histogram distribution 
Stddev 

The square root of the variance  
Variance 

The amount of variation of the histogram distribution  
Skewness 

The asymmetry of the histogram distribution  
Kurtosis 

The flatness of the histogram distribution  
Energy 

The uniformity of the histogram distribution  
Entropy 

The randomness of the histogram distribution  


LCM  Contrast 

The local variation of voxel pairs 
Dissimilarity 

The variation of voxel pairs  
Homogeneity 

The homogeneity of voxel pairs  
Angular second moment (ASM) 

The uniformity of voxel pairs  
Energy 

Square root of the ASM  
Probability max 

High max value of voxel pairs  
Entropy 

The randomness of voxel pairs  
Correlation 

The linear dependency of gray levels  


GLRLM  Long runs emphasis (LRE) 

The distribution of the long run length 
Graylevel nonuniformity (GLN) 

The nonuniformity of the gray level  
Run length nonuniformity (RLN) 

The nonuniformity of the run length  
Lowgraylevel run emphasis (LGRE) 

The distribution of the low gray level groups  
Highgraylevel run emphasis (HGRE) 

The distribution of the high gray level groups 
GLCM is a matrix that represents the frequency of occurrence in the relationship of gray level between neighboring voxels with a specific direction. GLRLM is a matrix characterized by the frequency of occurrence in the consecutive voxels with the same attenuation value along a specific direction [
The GLCM is represented by the maximum size of the gray level, both in rows and columns. The relationship between gray values of all the pixels and the gray values of neighboring pixels according to a given direction and distance is expressed in the number of occurrences in the matrix. In Figure
The calculation process of (a) GLCM and (b) GLRLM (distance: 1, direction: 0°).
In the GLRLM, the rows are represented by gray values and the columns are expressed by the same number of adjacent pixels. The number of occurrences for the case where the gray value of each of the pixels is the same as the gray value of the neighboring pixels according to a given direction, and distance is represented by a matrix. In Figure
On a threedimensional space, GLCM and GLRLM can generally conduct a matrix calculation in 13 directions (Figure
13 Directions for matrix calculation on a threedimensional space. On a threedimensional space, GLCM and GLRLM can generally conduct a matrix calculation in 13 directions.
Multiple regression analysis was performed to evaluate the effect of different scan parameters on the computer features of nodule phantoms (SPSS version 18.0, SPSS Inc., USA) [
The results of multiple regression analysis in 100 HU nodule phantoms are presented in Tables
Regression coefficients in the comparison between different slice thickness, mAs, and reconstruction algorithm in 100 HU nodule phantoms.
Image features  Constant  Slice thickness  mAs  Reconstruction algorithm  

5.00 (ref.)  2.50  1.25  30 (ref.)  120  Lung (ref.)  Standard  
Histogram  Mean  0.323  0  0.418^{†}  0.613^{†}  0  0.020^{†}  0  −0.271^{†}  
Stddev  0.442  0  −0.085^{†}  −0.237^{†}  0  −0.055^{†}  0  −0.075^{†}  
Variance  0.260  0  −0.058^{†}  −0.152^{†}  0  −0.036^{†}  0  −0.047^{†}  
Skewness  0.008  0  0.179^{†}  0.463^{†}  0  0.051^{†}  0  0.108^{†}  
Kurtosis  0.772  0  −0.135^{†}  −0.177^{†}  0  −0.032  0  −0.234^{†}  
Energy  0.112  0  0.388^{†}  0.642^{†}  0  0.103^{†}  0  −0.209^{†}  
Entropy  0.814  0  −0.430^{†}  −0.664^{†}  0  −0.079^{†}  0  0.188^{†}  


GLCM  Contrast  0.078  0  0.171^{†}  0.383^{†}  0  −0.022^{†}  0  −0.026^{†}  
Dissimilarity  0.231  0  0.011  0.135^{†}  0  −0.058^{†}  0  0.036^{†}  
Homogeneity  0.147  0  0.472^{†}  0.600^{†}  0  0.131^{†}  0  −0.127^{†}  
ASM  0.025  0  0.387^{†}  0.552^{†}  0  0.186^{†}  0  −0.159^{†}  
Energy  0.130  0  0.476^{†}  0.614^{†}  0  0.142^{†}  0  −0.145^{†}  
Probability max  0.150  0  0.466^{†}  0.583^{†}  0  0.141^{†}  0  −0.137^{†}  
Entropy  0.856  0  −0.454^{†}  −0.646^{†}  0  −0.116^{†}  0  0.153^{†}  
Correlation  0.927  0  −0.129^{†}  −0.299^{†}  0  0.026^{†}  0  0.015  


GLRLM  LRE  0.036  0  0.452^{†}  0.574^{†}  0  0.178^{†}  0  −0.097^{†}  
GLN  0.105  0  0.418^{†}  0.621^{†}  0  0.117^{†}  0  −0.181^{†}  
RLN  0.745  0  −0.541^{†}  −0.636^{†}  0  −0.105^{†}  0  0.116^{†}  
LGRE  0.212  0  −0.130^{†}  −0.180^{†}  0  −0.011  0  0.037^{†}  
HGRE  0.288  0  0.509^{†}  0.672^{†}  0  0.039^{†}  0  −0.161^{†} 
Absolute effect size in 100 HU phantom nodules.
Image features  Slice thickness  mAs  Reconstruction algorithm  

5.00 vs 1.25  5.00 vs 2.50  2.50 vs 1.25  30 vs 120  Lung vs standard  
Effect 

Effect 

Effect 

Effect 

Effect 


Histogram  Mean  <  4.62  <  2.81  <  1.18  <  0.07  >  1.02  
Stddev  >  3.06  >  1.05  >  1.40  >  0.42  >  0.59  
Variance  >  3.35  >  0.72  >  1.06  >  0.38  >  0.50  
Skewness  <  1.98  <  1.77  <  1.20  <  0.18  <  0.40  
Kurtosis  >  0.64  >  1.58  >  0.15  0  0.13  >  1.09  
Energy  <  5.63  <  2.70  <  1.44  <  0.35  >  0.74  
Entropy  >  7.06  >  3.25  >  1.67  >  0.26  <  0.65  


GLCM  Contrast  <  5.73  <  2.27  <  2.26  >  0.13  >  0.15  
Dissimilarity  <  2.04  0  0.11  0  1.05  >  0.52  <  0.31  
Homogeneity  <  5.81  <  3.36  <  0.87  <  0.46  >  0.45  
ASM  <  4.20  <  2.32  <  0.79  <  0.68  >  0.57  
Energy  <  5.53  <  3.28  <  0.88  <  0.49  >  0.50  
Probability max  <  5.25  <  3.28  <  0.77  <  0.51  >  0.49  
Entropy  >  6.56  >  3.38  >  1.34  >  0.39  <  0.53  
Correlation  >  5.61  >  1.69  >  1.96  <  0.18  0  0.10  


GLRLM  LRE  <  4.40  <  3.05  <  0.65  <  0.64  >  0.34  
GLN  <  5.36  <  2.96  <  1.17  <  0.40  >  0.64  
RLN  >  5.35  >  3.85  >  0.88  >  0.35  <  0.39  
LGRE  >  5.18  >  1.59  >  0.64  0  0.11  <  0.37  
HGRE  <  6.30  <  3.93  <  1.66  <  0.13  >  0.54 
The results of multiple regression analysis in −630 HU nodule phantoms are presented in Tables
Regression coefficients in the comparison between different slice thickness, mAs, and reconstruction algorithm in −630 HU nodule phantoms.
Image features  Constant  Slice thickness  mAs  Reconstruction algorithm  

5.00 (ref.)  2.50  1.25  30 (ref.)  120  Lung (ref.)  Standard  
Histogram  Mean  0.617  0  0.247^{†}  0.277^{†}  0  −0.011  0  −0.154^{†}  
Stddev  0.369  0  0.036  0.158^{†}  0  −0.149^{†}  0  −0.279^{†}  
Variance  0.230  0  0.034  0.154^{†}  0  −0.125^{†}  0  −0.200^{†}  
Skewness  −0.023  0  0.209^{†}  0.256^{†}  0  0.143^{†}  0  0.224^{†}  
Kurtosis  0.985  0  −0.146^{†}  −0.098^{†}  0  −0.149^{†}  0  −0.436^{†}  
Energy  0.061  0  0.063^{†}  −0.021  0  0.189^{†}  0  0.361^{†}  
Entropy  0.664  0  0.014  0.122^{†}  0  −0.196^{†}  0  −0.394^{†}  


GLCM  Contrast  0.237  0  0.111^{†}  0.238^{†}  0  −0.108^{†}  0  −0.206^{†}  
Dissimilarity  0.349  0  0.121^{†}  0.251^{†}  0  −0.141^{†}  0  −0.292^{†}  
Homogeneity  0.193  0  −0.039^{†}  −0.154^{†}  0  0.191^{†}  0  0.422^{†}  
ASM  0.027  0  −0.004  −0.060^{†}  0  0.109^{†}  0  0.166^{†}  
Energy  0.080  0  −0.015  −0.086^{†}  0  0.139^{†}  0  0.233^{†}  
Probability max  0.052  0  −0.008  −0.036^{†}  0  0.084^{†}  0  0.122^{†}  
Entropy  0.788  0  0.042  0.130^{†}  0  −0.153^{†}  0  −0.284^{†}  
Correlation  0.752  0  −0.100^{†}  −0.228^{†}  0  0.109^{†}  0  0.204^{†}  


GLRLM  LRE  0.068  0  0.014  −0.054^{†}  0  0.130^{†}  0  0.258^{†}  
GLN  0.033  0  0.033^{†}  −0.018  0  0.125^{†}  0  0.223^{†}  
RLN  0.891  0  −0.004  0.078^{†}  0  −0.163^{†}  0  −0.318^{†}  
LGRE  0.167  0  −0.106^{†}  −0.086^{†}  0  −0.018  0  0.011  
HGRE  0.639  0  0.227^{†}  0.269^{†}  0  −0.030^{†}  0  −0.144^{†} 
Effect of scan parameters on computer features in −630 HU phantom nodules.
Image features  Slice thickness  mAs  Reconstruction algorithm  

5.00 vs 1.25  5.00 vs 2.50  2.50 vs 1.25  30 vs 120  Lung vs standard  
Effect 

Effect 

Effect 

Effect 

Effect 


Histogram  Mean  <  2.85  <  2.71  <  0.28  0  0.07  >  1.11 
Stddev  <  0.71  0  0.25  <  0.50  >  0.72  >  1.65  
Variance  <  0.74  0  0.35  <  0.55  >  0.68  >  1.20  
Skewness  <  1.15  <  1.25  0  0.19  <  0.61  <  1.04  
Kurtosis  >  0.40  >  0.63  <  0.16  >  0.58  >  2.77  
Energy  0  0.10  <  0.29  >  0.31  <  0.89  <  2.46  
Entropy  <  0.52  0  0.07  <  0.38  >  0.85  >  2.56  


GLCM  Contrast  <  1.32  <  0.97  <  0.65  >  0.58  >  1.27 
Dissimilarity  <  1.21  <  0.74  <  0.56  >  0.65  >  1.67  
Homogeneity  >  0.69  >  0.16  >  0.40  <  0.79  <  2.81  
ASM  >  0.48  0  0.03  >  0.42  <  0.87  <  1.53  
Energy  >  0.55  0  0.08  >  0.41  <  0.87  <  1.80  
Probability max  >  0.34  0  0.06  0  0.26  <  0.76  <  1.21  
Entropy  <  0.63  0  0.20  <  0.41  >  0.75  >  1.72  
Correlation  >  1.20  >  0.80  >  0.65  <  0.58  <  1.21  


GLRLM  LRE  >  0.40  0  0.08  >  0.34  <  0.79  <  2.12 
GLN  0  0.13  <  0.23  >  0.31  <  0.91  <  2.23  
RLN  <  0.46  0  0.02  <  0.34  >  0.83  >  2.27  
LGRE  >  0.98  >  1.49  0  0.35  0  0.21  0  0.13  
HGRE  <  2.89  <  2.64  <  0.44  >  0.20  >  1.09 
In the regression analysis, the absolute value of regression coefficient (RC) can represent the scale of difference. RCs in all features in comparison between 5.0 mm and 1.25 mm were larger than RCs between 5.0 mm and 2.5 mm (Tables
In 100 HU nodule phantoms, the maximum (Max), median (Med), and minimum (Min) values of RC were 0.541, 0.388, and 0.011 between 5.0 mm and 2.5 mm slice thickness, and 0.672, 0.574, and 0.135 between 5.0 mm and 1.25 mm slice thickness, respectively. Max, Med, and Min values of RC were 0.186, 0.079, and 0.011 between 30 mAs and 120 mAs and 0.271, 0.127, and 0.015 between lung and standard reconstruction algorithms, respectively (Table
In −630 HU nodule phantoms, Max, Med, and Min values of RC were 0.247, 0.039, and 0.004 between 5.0 mm and 2.5 mm slice thickness, and 0.277, 0.122, and 0.018 between 5.0 mm and 1.5 mm slice thickness, respectively. Max, Med, and Min values of RC were 0.196, 0130, and 0.011 between 30 mAs and 120 mAs and 0.436, 0.224, and 0.011 between lung and standard reconstruction algorithms, respectively (Table
The values of computer features in different slice thicknesses, mAs, and reconstruction algorithms are presented in Supplementary Materials
Our study showed that (a) in both of 100 HU and −630 HU nodule phantoms, differences in the scan parameters had a significant effect on almost all computer features with few exceptions, (b) in the 100 HU nodule phantoms, considering the Max and Med values of RCs between different slice thicknesses were larger than the Max and Med values of RCs between different mAs or algorithms, we speculate slice thickness had a greater effect than mAs or algorithm, and (c) in the −630 HU nodule phantoms, considering the Max and Med values of RCs between different algorithms were larger than the Max and Med values of RCs between different slice thicknesses or mAs, we speculate algorithm had a greater effect than slice thickness or mAs.
In this study, differences in the scan parameters had a significant effect on almost all computer features. These results indicate that noise or artifacts affected the attenuation and texture in the nodule, which indicates that the scan parameters are related to noise and artifact. Also, our results are consistent with several previous studies [
In the regression analysis, the RC can represent the scale of difference in feature values in different scan parameters. We found in this study that RCs in all features in comparison between 5.0 mm and 1.25 mm were larger than RCs between 5.0 mm and 2.5 mm. As the slice thickness becomes thinner, the noise increases. As the slice thickness becomes thicker, the noise decreases and greater effect of the partial volume effect [
In the 100 HU nodule phantoms, we found slice thickness had a greater effect than mAs or algorithm. The CT image of 100 HU nodule phantoms has smaller noise than CT of −630 HU nodule phantom due to higher average attenuation and a smaller variation in the distribution of attenuation. Therefore, the 100 HU nodule phantom is more affected by the partial volume effect than the −630 HU nodule phantom. As the result, the slice thickness that is closely related to partial volume effect was the most influential parameter in the 100 HU nodule phantom.
We also found in this study that, in the 100 HU nodule phantoms, GLCMdissimilarity showed no difference between slice thickness of 5.0 and 2.5 mm and between 2.5 mm and 1.25 mm. The weight of the dissimilarity increases linearly unlike other texture features which increases exponentially [
In the −630 HU nodule phantoms, the reconstruction algorithm had a greater effect than slice thickness or mAs. The amount of noise in the −630 HU nodule phantom is greater than the amount in the 100 HU nodule phantom. The lung reconstruction algorithm usually makes higher noise level that the standard algorithm. The change of reconstruction algorithm from standard to lung algorithm made bigger increase of noise in the nodule with inherently higher level of noise, that is, −630 HU nodule.
In this study, in the −630 HU nodule phantoms, the effect of slice thickness was smaller than the effect of the reconstruction algorithm. We can notice that smaller difference of slice thickness might make smaller difference of noise, thus the statistical difference was not significant. On the other hand, due to higher average attenuation and a smaller variation in the distribution of attenuation, the CT image of 100 HU nodule phantoms was more affected by the difference of slice thickness that cause difference of partial volume effect.
CT images of the lung reconstruction algorithm contain higher noise level that the image of standard algorithm. Therefore, the change of the reconstruction algorithm can affect the features associated with CT histogram or CT texture. In this study, we found that change of the algorithms had significant effect on 19 computer features in the 100 and −630 HU nodule phantoms. Only a feature, that is, correlation, in 100 HU phantoms and LGRE in −630 HU nodule phantom showed no difference between lung and standard algorithms.
In this study, we found that most computer features showed significant difference between 30 mAs and 120 mAs. We speculate that this significant difference was originated by the change of noise level. We also found that LGRE showed no difference between 30 mAs and 120 mAs in both of 100 HU and −630 HU nodule phantoms. LGRE, that is, low gray level run emphasis, can be defined by distribution of run length in the low gray values. The value of LGRE is high when there is the large number of pixels with low gray level [
This study demonstrated that the change of CT scan parameters can affect the quantitative CT features. In clinical studies involving deep learning or radiomics, it should be noted that differences in values can occur when using computer features obtained from different CT scan parameters in combination. Therefore, when interpreting the statistical analysis results, it is necessary to reflect the difference in the computer features depending on the scan parameters. In further studies, we need to develop methods for the standardization of computer features obtained from different scan parameters.
The data used to support the findings of this study are available from the corresponding author upon request.
The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.
HyunJu Lee and Kwnag Gi Kim contributed equally to this study.
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2017R1D1A1B04032467).
Supplementary Material 1: the computer features according to the scanning parameters in the 100 HU nodule phantoms. Supplementary Material 2: the computer features according to the scanning parameters in the −630 HU nodule phantoms. Supplementary Material 3: the computer features in 12 different scan parameters in 100 HU nodule phantoms. Supplementary Material 4: the computer features in 12 different scan parameters in −630 HU nodule phantoms.