Predicting recurrence risks in lung cancer patients using multimodal radiomics and random survival forests

Jaryd R Christie; Omar Daher; Mohamed Abdelrazek; Perrin E Romine; Richard A Malthaner; Mehdi Qiabi; Rahul Nayak; Sandy Napel; Viswam S Nair; Sarah A Mattonen

doi:10.1117/1.JMI.9.6.066001

Predicting recurrence risks in lung cancer patients using multimodal radiomics and random survival forests

J Med Imaging (Bellingham). 2022 Nov;9(6):066001. doi: 10.1117/1.JMI.9.6.066001. Epub 2022 Nov 8.

Authors

Jaryd R Christie^{1

2}, Omar Daher³, Mohamed Abdelrazek³, Perrin E Romine^{4

5}, Richard A Malthaner⁶, Mehdi Qiabi⁶, Rahul Nayak⁶, Sandy Napel⁷, Viswam S Nair^{4

8}, Sarah A Mattonen^{1

2

9}

Affiliations

¹ Western University, Department of Medical Biophysics, London, Ontario, Canada.
² London Regional Cancer Program, Baines Imaging Research Laboratory, London, Ontario, Canada.
³ Western University, Department of Medical Imaging, London, Ontario, Canada.
⁴ Fred Hutchinson Cancer Research Center, Clinical Research Division, Seattle, Washington, United States.
⁵ University of Washington School of Medicine, Division of Medical Oncology, Seattle, Washington, United States.
⁶ Western University, Division of Thoracic Surgery, Department of Surgery, London, Ontario, Canada.
⁷ Stanford University, Department of Radiology, Stanford, California, United States.
⁸ University of Washington School of Medicine, Division of Pulmonary and Critical Care Medicine, Seattle, Washington, United States.
⁹ Western University, Department of Oncology, London, Ontario, Canada.

Abstract

Purpose: We developed a model integrating multimodal quantitative imaging features from tumor and nontumor regions, qualitative features, and clinical data to improve the risk stratification of patients with resectable non-small cell lung cancer (NSCLC).

Approach: We retrospectively analyzed 135 patients [mean age, 69 years (43 to 87, range); 100 male patients and 35 female patients] with NSCLC who underwent upfront surgical resection between 2008 and 2012. The tumor and peritumoral regions on both preoperative CT and FDG PET-CT and the vertebral bodies L3 to L5 on FDG PET were segmented to assess the tumor and bone marrow uptake, respectively. Radiomic features were extracted and combined with clinical and CT qualitative features. A random survival forest model was developed using the top-performing features to predict the time to recurrence/progression in the training cohort ( $n = 101$ ), validated in the testing cohort ( $n = 34$ ) using the concordance, and compared with a stage-only model. Patients were stratified into high- and low-risks of recurrence/progression using Kaplan-Meier analysis.

Results: The model, consisting of stage, three wavelet texture features, and three wavelet first-order features, achieved a concordance of 0.78 and 0.76 in the training and testing cohorts, respectively, significantly outperforming the baseline stage-only model results of 0.67 ( $p < 0.005$ ) and 0.60 ( $p = 0.008$ ), respectively. Patients at high- and low-risks of recurrence/progression were significantly stratified in both the training ( $p < 0.005$ ) and the testing ( $p = 0.03$ ) cohorts.

Conclusions: Our radiomic model, consisting of stage and tumor, peritumoral, and bone marrow features from CT and FDG PET-CT significantly stratified patients into low- and high-risk of recurrence/progression.

Keywords: computed tomography; lung cancer; machine learning; positron emission tomography; radiomics.

Grants and funding

T32 CA009515/CA/NCI NIH HHS/United States