ESTRO 2020 Abstract Book

S849 ESTRO 2020

Purpose or Objective To describe the possibility to build a classifier for patients at risk of lymph-nodal relapse and a predictive model for disease specific survival in patients with early stage non- A cohort of 102, who received stereotactic body radiation treatment, was retrospectively investigated. A set of 45 textural features was computed for the tumor volumes on the treatment planning CT images. Patients were split into two independent cohorts ((70 patients (68.9%) for training and 32 patients (31.4%) for validation). Three different models were built in the study. A stepwise backward linear discriminant analysis was applied to identify patients at risk of lymph-nodal progression. The performance of the model was assessed by means of standard metrics derived from the confusion matrix. Furthermore, all textural features were correlated to survival data to build two separate predictive models for progression free survival (PFS) and disease specific survival (DS-OS). These models were built from the features/predictors found significant at univariate analysis and to elastic net regularization, by means of a multivarate cox regression with backward selection. Low and high risk groups were identified by maximizing the separation by means of the Youden method Results In the total cohort (77 (75.5%) males and 25 (24.5%) females, median age 76.6 years), 15 patients presented nodal progression at the time of analysis; 19 patients (18.6%) died because of disease specific causes, 25 (24.5%) died from other reasons, 28 (27.5%) were alive without disease and 30 (29.4%) with either local or distant progression. The specificity, sensitivity, accuracy of the classifier resulted 83.1±24.5, 87.4±1.2 and 85.4±12.5 in the validation group (coherently with the findings in the training). The area under the curve for the classifier resulted 0.84±0.04 and 0.73±0.05 for training and validation respectively. The mean time for DS-OS and PFS for the low- and high- risk sub-groups of patients (in the validation groups) resulted: 88.2±9.0 vs 84.1±7.8 (low risk) and 52.7±5.9 vs 44.6±9.2 (high risk) Conclusion Radiomics analysis, based on planning CT images, allowed to build a classifier and predictive models capable to identify patients at risk of nodal relapse and at high risk of bad prognosis. The radiomics signatures identified were mostly related to tumor heterogeneity PO-1568 The Impact of Varying Number of OSEM Subsets on PET Radiomic Features: A Preclinical Study E. Alsyed 1 , R. Smith 2 , S. Paisey 2 , C. Marshall 2 , C. Parkinson 1 , P. Whybra 1 , E. Spezi 1 1 Cardiff University, School of Engineering, Cardiff, United Kingdom ; 2 Cardiff University, Wales Research & Diagnostic PET Imaging Centre- School of Medicine, Cardiff, United Kingdom Purpose or Objective Positron emission tomography (PET) imaging plays a vital role in the assessment and management of cancer [1]. Accelerated reconstruction of PET imaging is widely accomplished through the use of the ordered subsets expectation maximization (OSEM) algorithm. Nevertheless, a trade-off exists between the number of subsets and image quality (e.g. noise) [2]. Consequently, radiomic analysis involving the extraction of quantitative features from medical images, may be affected as a result of degradation of the image quality [2], [3]. The aim of this study was to assess the impact of changing the number of OSEM subsets upon stability of PET radiomic features. Material and Methods Analysis was performed on PET scans of 8 mice with 4T1 tumours. They were injected with 10.0 ± 2.0 MBq of 18F- FDG, 50 minutes post injection they were imaged (by small cell lung cancer Material and Methods

Mediso Nanoscan PET/CT) for 20 minutes. Scans were reconstructed with four different OSEM subsets (1, 2, 4, 6) as showing in figure 1. One hundred thirty eight radiomic features were extracted using SPAARC (Spaarc Pipeline for Automated Analysis and Radiomic Computing, an in-house developed tool built on Matlab [4] ). Coefficient of variation (COV) for the features across the different subsets were calculated. Features were categorized based on their COV values into four groups including very small (COV ≤ 5%), small (5% < COV ≤ 10%), intermediate (10% < COV ≤ 20%) and large (COV > 20%) variation [5].

Results Thirty-three (24%) features showed COV ≤ 5% (very small variation). Twenty-seven (19%) features exhibited variation with a COV in the range of 5%-10% (small variation). Only 24 (17%) features showed variation with a COV in the range of 10%-20% (intermediate variation). More than a third (40%) of the features were found to be not-stable (COV > 20%). Figure 2 showing the percentage of features for each category.

Conclusion In this study, we investigated the impact of variations in the number of OSEM subsets on pre-clinical PET radiomic features. Our results indicate that varying the OSEM subsets changes the radiomic output derived from these images. Therefore, more research into the impact of OSEM reconstruction parameters on clinical PET needs to be undertaken in order to further understand its impact on radiomic features. References : [1] M. E. Juweid et al,“Positron-emission tomography and assessment of cancer therapy” N. Engl. J. Med. , vol. 354, no. 5,pp. 496–507, 2006. [2] A. M. Morey et al,“Effect of Varying Number of OSEM Subsets on PET Lesion Detectability” J. Nucl. Med. Technol. , vol.41, no.4, pp.268–273,2013. [3] G. J. R. Cook et al,“Radiomics in PET:principles and applications” Neuroimage , pp.269–276,2014. [4] P. Whybra et al, “Assessing radiomic feature robustness to interpolation in F-FDG PET imaging” Sci. Rep. , no. Dec, pp.0–10, 2019. [5] I. Shiri et al,“The impact of image reconstruction settings on 18F-FDG PET radiomic features: multi-scanner phantom and patient studies” Eur.Radiol. , vol. 27, no. 11, pp. 4498–4509, 2017.