A novel method for generating a fused ventilation–perfusion map using CT-derived ventilation and FDG-PET for lung functional avoidance in radiotherapy: A feasibility study

Abstract

Background: Pulmonary ventilation–perfusion function plays a crucial role in both radiotherapy planning and prognosis assessment in patients with lung cancer. However, there remains a lack of rapid and cost-effective imaging modalities capable of accurately capturing this functional parameter.

Purpose: This study aimed to develop a lung functional image that reflects ventilation–perfusion characteristics by integrating computed tomography (CT) and (positron emission tomography) PET imaging techniques, with the objective of enhancing lung dose evaluation in radiotherapy planning.

Approach: A retrospective analysis was performed on twenty lung cancer patients using CT images acquired at two respiratory phases and FDG-PET/CT images. The Elastic Distortion algorithm was applied for deformable image registration, with the end-expiration phase CT serving as the baseline. Values derived from the determinant Jacobian matrices of ventilation CT (V-CT) images and the gray-value matrices of PET images were normalized to a range of 0 to 1. These normalized values were multiplied to generate a ventilation–perfusion matrix. Three types of lung functional images were produced from these matrices: ventilation-imaging (V-imaging), perfusion-imaging (P-imaging), and ventilation–perfusion- imaging (VP-imaging). The Dice Similarity Coefficient (DSC) and Bland–Altman plots were used to assess the correlations and discrepancies among the imaging modalities.

Results: The DSC values for the entire lung, regions with low 30% functionality, and regions with high 40% functionality were 0.39 ± 0.05, 0.50 ± 0.03, and 0.20 ± 0.05 for V–P; 0.58 ± 0.03, 0.73 ± 0.03, and 0.32 ± 0.02 for V–VP; and 0.68 ± 0.04, 0.78 ± 0.04, and 0.34 ± 0.04 for P–VP, respectively. Notably, significant concordance was observed between V–VP and P–VP images within the delineated functional lung regions.#x02013;Altman analysis supported the DSC results, revealing high correlation coefficients in the low 30% functional lung region: 0.628 for V–P, 0.857 for V–VP, and 0.779 for P–VP. In contrast, similarity within the high 40% functional regions was markedly lower.

Conclusion: This study developed a novel method for generating a fused VP map by integrating CT-derived ventilation and FDG-PET data. The method demonstrated feasibility, and the resulting VP map provided a balanced representation of both ventilation and perfusion signals, particularly in regions with reduced lung function.

Keywords

CT-based functional imaging / deformable registration / lung functional imaging / PET-based functional imaging / ventilation–perfusion imaging

References

[1] Shin-Hyung Park, Jae-Kwang Lim, Min Kyu Kang, et al. Predictive factors for severe radiation-induced lung injury in patients with lung cancer and coexisting interstitial lung disease. Radiother Oncol. 2024, 192:110053.

[2] Li Tong, Chen Luqing, Wang Xiaohong, et al. The value of FEV1%, FEV1/FVC and DLCO% in predicting radiation pneumonia in elderly patients with non-small cell lung cancer before radiotherapy. Chin J Prac Med, 2019, 14(20): 8–10.

[3] Remy-Jardin M, Hutt A, Flohr T, et al. Ultra-High-Resolution Photon-Counting CT Imaging of the Chest: A New Era for Morphology and Function. Invest Radiol. 2023; 58(7): 482-487.

[4] Zhou PX, Zhang SX. Functional lung imaging in thoracic tumor radiotherapy: Application and progress. Front Oncol. 2022; 23(12):908345.

[5] Vinogradskiy Y. CT-based ventilation imaging in radiation oncology. BJR Open. 2019, 1(1):20180035.

[6] Latifi K, Forster KM, Hoffe SE, et al. Dependence of ventilation image derived from 4D CT on deformable image registration and ventilation algorithms. J Appl Clin Med Phys. 2013; 14(4): 4247.

[7] Hegi-Johnson F, Keall P, Barber J, et al. Evaluating the accuracy of 4D-CT ventilation imaging: first comparison with Technegas SPECT ventilation. Med Phys. 2017, 44(8): 4045-4055.

[8] Yamamoto T, Kabus S, Lorenz C, et al. Pulmonary ventilation imaging based on 4-dimensional computed tomography: comparison with pulmonary function tests and SPECT ventilation images. Int J Radiat Oncol Biol Phys. 2014; 90(2): 414-422.

[9] Mirsadraee S, van Beek EJ. Functional imaging: computed tomography and MRI. Clin Chest Med. 2015; 36(2): 349-363.

[10] Kajikawa T, Kadoya N, Maehara Y, et al. A deep learning method for translating 3DCT to SPECT ventilation imaging: First comparison with 81m Kr-gas SPECT ventilation imaging. Med Phys. 2022; 49(7): 4353-4364.

[11] MATUSZAK M M, MATROSIC C, JAREMA D, et al. Priority-driven plan optimization inlocally advanced lung patients based on perfusion SPECT imaging. Adv Radiat Oncol. 2016; 1(4): 281-289.

[12] Castillo E, Castillo R, Vinogradskiy Y, et al. Technical Note: On the spatial correlation between robust CT-ventilation methods and SPECT ventilation. Med Phys. 2020; 47(11): 5731-5738.

[13] Zhou PX, Zhang SX. Functional lung imaging in thoracic tumor radiotherapy: Application and progress. Front Oncol. 2022; 23(12):908345.

[14] Pinot F, Bourhis D, Bourbonne V, et al. New Automated Method for Lung Functional Volumes Delineation with Lung Perfusion PET/CT Imaging. Cancers (Basel). 2023; 15(7): 2166.

[15] Kinahan PE. The emergence of PET/CT: Engineering, innovation, and usage. Med Phys. 2023; 50 Suppl 1: 62-65.

[16] Nakamura H, Hirai T, Kurosawa H, et al. Current advances in pulmonary functional imaging. Respir Investig. 2024; 62(1): 49-65.

[17] West JB. State of the art: ventilation-perfusion relationships. Am Rev Respir Dis. 1977; 116(5): 919-943.

[18] Lenfant C, Okubo T. Distribution function of pulmonary blood flow and ventilation-perfusion ratio in man. J Appl Physiol. 1968; 24(5): 668-677.

[19] Powers KA, Dhamoon AS. Physiology, Pulmonary Ventilation and Perfusion. In: StatPearls [Internet]. Treasure Island (FL). PA: StatPearls 2023. Available from: https://www.ncbi.nlm.nih.gov/books/NBK539907

[20] Li P, Wang S, Li T, et al. A Large-Scale CT and PET/CT Dataset for Lung Cancer Diagnosis (Lung-PET-CT-Dx) [Data set]. Cancer Imaging Arch. https://doi.org/10.7937/TCIA.2020.NNC2-0461

[21] You Y, Niu Y, Sun F, et al. Three-dimensional printing and 3D slicer powerful tools in understanding and treating neurosurgical diseases. Front Surg. 2022; 9:1030081.

[22] Stȩpień P, Krauze W, Kujawińska M. Preprocessing methods for quantitative phase image stitching. Biomed Opt Express. 2021; 13(1): 1-13. https://doi.org/10.1364/BOE.439045

[23] Delaunay R, Hu Y, Vercauteren T. An unsupervised learning approach to ultrasound strain elastography with spatio-temporal consistency. Phys Med Biol. 2021; 66(17):175031.

[24] Holsgrove KM, Duchamp M, Moreno MS, et al. Elastic distortion determining conduction in BiFeO3 phase boundaries. RSC Adv. 2020; 10(47): 27954-27960.

[25] Lai K, Yu Y, Sui Q, et al. Jacobian methods for dynamic polarization control in optical applications. OptExpress. 2023; 31(8): 12175-12188.

[26] Iqbal GMD, Zhang H, D'Souza W, et, al. Four-dimensional computed tomography-based ventilation imaging in intensity-modulated radiation therapy treatment planning for pulmonary functional avoidance. J Appl Clin Med Phys. 2023; 24(6):e13920.

[27] Nakamura H, Hirai T, Kurosawa H, et al. Current advances in pulmonary functional imaging. Respir Investig. 2024; 62(1): 49-65.

[28] Du JQ, Zeng SC, Wu KN, et al. Methods and evaluation of pulmonary ventilation/perfusion function images. Mod Clin Biomed Eng J. 1997; 3(2): 4.

[29] Wang Jing, Zhao Shujun, Dong Jie. Methods and Image Processing for Quantitative Analysis of Radionuclide Pulmonary Ventilation/Perfusion Imaging [C]//National Conference on Nuclear Electronics and Nuclear Detection Technology. 2008. https://ConferenceArticle/5aa0c023c095d722207f108e

[30] Yeung M, Rundo L, Nan Y, et, al. Calibrating the Dice Loss to Handle Neural Network Overconfidence for Biomedical Image Segmentation. J Digit Imaging. 2023; 36(2): 739-752.

[31] Ruppel GL, Enright PL. Pulmonary function testing. Respir Care. 2012; 57(1) : 165–175.

[32] Bland JM and Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986; 1(8476): 307–310.

[33] Marquez BA, Singh J, Morison H, et al. Fully-integrated photonic tensor core for image convolutions. Nanotechnology. 2023; 34(39):10.1088/1361-6528/acde83

[34] Petersson J, Glenny RW. Gas exchange and ventilation-perfusion relationships in the lung. Eur Respir J. 2014; 44(4): 1023-1041.

[35] West JB. Understanding pulmonary gas exchange: ventilation-perfusion relationships. J Appl Physiol (1985). 2004; 97(5): 1603-1604.

[36] Lanks CW. Match Me If You Can: The Relationship between Ventilation and Perfusion with Position Changes in Nonhomogenous Lung Injury. Ann Am Thorac Soc. 2022; 19(2): 320–326.

[37] Honda N, Osada H, Watanabe W, Nakayama M, Nishimura K, Krauss B, Otani K. Imaging of ventilation with dual-energy CT during breath hold after single vital-capacity inspiration of stable xenon. Radiology. 2012; 262(1): 262-268.

[38] Ronneberger O, Fischer P, Brox T. U-net: convolutional networks for biomedical image segmentation. Lect Notes Comput Sci. 2015; 9351: 234-241.

[39] Fain SB. Machine Learning Reveals the Texture of Regional Lung Ventilation at CT. Radiology. 2019; 293(3): 685-686.

[40] Zhong Y, Vinogradskiy Y, Chen L, et al. Technical Note: Deriving ventilation imaging from 4DCT by deep convolutional neural network. Med Phys. 2019; 46(5): 2323-2329.

[41] Nawwar AA, Green C, Searle J, Hess S, Lyburn ID. Impact of pre-scan patient-related factors on diagnostic yield of 18F FDG PET/CT in fever of unknown origin. Q J Nucl Med Mol Imaging. 2025; 69(3): 200-207.

RIGHTS & PERMISSIONS

2026 The Author(s). Precision Radiation Oncology published by John Wiley & Sons Australia, Ltd on behalf of Shandong Cancer Hospital & Institute.