Kenta Ninomiya1, Hidetaka Arimura2. 1. Division of Medical Quantum Science, Department of Health Sciences, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-ku, Fukuoka 812-8582, Japan. 2. Division of Medical Quantum Science, Department of Health Sciences, Faculty of Medical Sciences, Kyushu University, 3-1-1, Maidashi, Higashi-ku, Fukuoka 812-8582, Japan. Electronic address: arimurah@med.kyushu-u.ac.jp.
Abstract
PURPOSE: This study explored a novel homological analysis method for prognostic prediction in lung cancer patients. MATERIALS AND METHODS: The potential of homology-based radiomic features (HFs) was investigated by comparing HFs to conventional wavelet-based radiomic features (WFs) and combined radiomic features consisting of HFs and WFs (HWFs), using training (n = 135) and validation (n = 70) datasets, and Kaplan-Meier analysis. A total of 13,824 HFs were derived through homology-based texture analysis using Betti numbers, which represent the topologically invariant morphological characteristics of lung cancer. The prognostic potential of HFs was evaluated using statistically significant differences (p-values, log-rank test) to compare the survival curves of high- and low-risk patients. Those patients were stratified into high- and low-risk groups using the medians of the radiomic scores of signatures constructed with an elastic-net-regularized Cox proportional hazard model. Furthermore, deep learning (DL) based on AlexNet was utilized to compare HFs by stratifying patients into the two groups using a network that was pre-trained with over one million natural images from an ImageNet database. RESULTS: For the training dataset, the p-values between the two survival curves were 6.7 × 10-6 (HF), 5.9 × 10-3 (WF), 7.4 × 10-6 (HWF), and 1.1 × 10-3 (DL). The p-values for the validation dataset were 3.4 × 10-5 (HF), 6.7 × 10-1 (WF), 1.7 × 10-7 (HWF), and 1.2 × 10-1 (DL). CONCLUSION: This study demonstrates the excellent potential of HFs for prognostic prediction in lung cancer patients.
PURPOSE: This study explored a novel homological analysis method for prognostic prediction in lung cancerpatients. MATERIALS AND METHODS: The potential of homology-based radiomic features (HFs) was investigated by comparing HFs to conventional wavelet-based radiomic features (WFs) and combined radiomic features consisting of HFs and WFs (HWFs), using training (n = 135) and validation (n = 70) datasets, and Kaplan-Meier analysis. A total of 13,824 HFs were derived through homology-based texture analysis using Betti numbers, which represent the topologically invariant morphological characteristics of lung cancer. The prognostic potential of HFs was evaluated using statistically significant differences (p-values, log-rank test) to compare the survival curves of high- and low-risk patients. Those patients were stratified into high- and low-risk groups using the medians of the radiomic scores of signatures constructed with an elastic-net-regularized Cox proportional hazard model. Furthermore, deep learning (DL) based on AlexNet was utilized to compare HFs by stratifying patients into the two groups using a network that was pre-trained with over one million natural images from an ImageNet database. RESULTS: For the training dataset, the p-values between the two survival curves were 6.7 × 10-6 (HF), 5.9 × 10-3 (WF), 7.4 × 10-6 (HWF), and 1.1 × 10-3 (DL). The p-values for the validation dataset were 3.4 × 10-5 (HF), 6.7 × 10-1 (WF), 1.7 × 10-7 (HWF), and 1.2 × 10-1 (DL). CONCLUSION: This study demonstrates the excellent potential of HFs for prognostic prediction in lung cancerpatients.