Abstract
<h2>Abstract</h2><h3>Purpose</h3> Radiomic texture calculation requires discretizing image intensities within the region-of-interest. FBN (fixed-bin-number), FBS (fixed-bin-size) and FBN and FBS with intensity equalization (FBNequal, FBSequal) are four discretization approaches. A crucial choice is the voxel intensity (Hounsfield units, or HU) binning range. We assessed the effect of this choice on radiomic features. <h3>Methods</h3> The dataset comprised 95 patients with head-and-neck squamous-cell-carcinoma. Dual energy CT data was reconstructed at 21 electron energies (40, 45,… 140 keV). Each of 94 texture features were calculated with 64 extraction parameters. All features were calculated five times: original choice, left shift (-10/-20 HU), right shift (+10/+20 HU). For each feature, Spearman correlation between nominal and four variants were calculated to determine feature stability. This was done for six texture feature types (GLCM, GLRLM, GLSZM, GLDZM, NGTDM, and NGLDM) separately. This analysis was repeated for the four binning algorithms. Effect of feature instability on predictive ability was studied for lymphadenopathy as endpoint. <h3>Results</h3> FBN and FBNequal algorithms showed good stability (correlation values consistently <mml:math><mml:mrow><mml:mo>></mml:mo></mml:mrow></mml:math>0.9). For FBS and FBSequal algorithms, while median values exceeded 0.9, the 95% lower bound decreased as a function of energy, with poor performance over the entire spectrum. FBNequal was the most stable algorithm, and FBS the least. <h3>Conclusions</h3> We believe this is the first multi-energy systematic study of the impact of CT HU range used during intensity discretization for radiomic feature extraction. Future analyses should account for this source of uncertainty when evaluating the robustness of their radiomic signature.
Highlights
Radiomics converts medical images to mineable data which may be used to predict various clinical endpoints or outcomes [1]
The dataset consisted of 95 patients with biopsy-proven mucosal head-and-neck squamous cell carcinomas (HNSCCs) below the hard palate including tumors arising from the oral cavity, oropharynx, hy popharynx, and larynx
To gain a deeper under standing of the instability observed in the fixed bin size (FBS) and FBSequal algorithms, we decided to study how this depends on the four bin sizes used (6.25, 12.5, 25, and 50 Hounsfield Units (HU))
Summary
Radiomics converts medical images to mineable data which may be used to predict various clinical endpoints or outcomes [1]. Commonly used imaging modalities are CT, MR, and PET. Radiomics has been part of the research lexicon since 2012 [2,3], and the field has burgeoned ever since. Clinical translation has been slow to non-existent. A crucial barrier for the translation of radiomics to the clinical setting is the concern regarding replicability of published re sults. One approach towards overcoming these hurdles is to standardize the image acquisition as well as various parts of the typical radiomic workflow [1,4]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
