Sensitivity of an AI method for [18F]FDG PET/CT outcome prediction of diffuse large B-cell lymphoma patients to image reconstruction protocols

EJNMMI Research

Original Research

Gerben J. C. Zwezerijnen¹  Sanne E. Wiegers¹  Maria C. Ferrández¹  Bart M. de Vries¹  Ronald Boellaard¹  Sandeep S. V. Golla¹  Jakoba J. Eertink²  Simone Pieplenbosch²  Josée M. Zijlstra²  Martijn W. Heymans³  Louise Schilder⁴ 

[1] Cancer Center Amsterdam, Department of Radiology and Nuclear Medicine, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands;Cancer Center Amsterdam, Imaging and Biomarkers, Amsterdam, The Netherlands;Cancer Center Amsterdam, Imaging and Biomarkers, Amsterdam, The Netherlands;Cancer Center Amsterdam, Department of Hematology, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands;Department of Epidemiology and Data Science, Amsterdam Public Health Research Institute, Amsterdam UMC, Vrije Universiteit Amsterdam, Amsterdam, The Netherlands;Amsterdam Public Health Research Institute, Methodology, Amsterdam, The Netherlands;Department of Internal Medicine, Amstelland Hospital, Amstelveen, The Netherlands;

关键词: Diffuse large B-cell lymphoma;    PET;    Convolutional neural networks;    Reconstruction;   

DOI  :  10.1186/s13550-023-01036-8

 received in 2023-07-05, accepted in 2023-09-19,  发布年份 2023

来源: Springer

【 摘 要 】

BackgroundConvolutional neural networks (CNNs), applied to baseline [18F]-FDG PET/CT maximum intensity projections (MIPs), show potential for treatment outcome prediction in diffuse large B-cell lymphoma (DLBCL). The aim of this study is to investigate the robustness of CNN predictions to different image reconstruction protocols. Baseline [18F]FDG PET/CT scans were collected from 20 DLBCL patients. EARL1, EARL2 and high-resolution (HR) protocols were applied per scan, generating three images with different image qualities. Image-based transformation was applied by blurring EARL2 and HR images to generate EARL1 compliant images using a Gaussian filter of 5 and 7 mm, respectively. MIPs were generated for each of the reconstructions, before and after image transformation. An in-house developed CNN predicted the probability of tumor progression within 2 years for each MIP. The difference in probabilities per patient was then calculated between both EARL2 and HR with respect to EARL1 (delta probabilities or ΔP). We compared these to the probabilities obtained after aligning the data with ComBat using the difference in median and interquartile range (IQR).ResultsCNN probabilities were found to be sensitive to different reconstruction protocols (EARL2 ΔP: median = 0.09, interquartile range (IQR) = [0.06, 0.10] and HR ΔP: median = 0.1, IQR = [0.08, 0.16]). Moreover, higher resolution images (EARL2 and HR) led to higher probability values. After image-based and ComBat transformation, an improved agreement of CNN probabilities among reconstructions was found for all patients. This agreement was slightly better after image-based transformation (transformed EARL2 ΔP: median = 0.022, IQR = [0.01, 0.02] and transformed HR ΔP: median = 0.029, IQR = [0.01, 0.03]).ConclusionOur CNN-based outcome predictions are affected by the applied reconstruction protocols, yet in a predictable manner. Image-based harmonization is a suitable approach to harmonize CNN predictions across image reconstruction protocols.

【 授权许可】

CC BY   
© Springer-Verlag GmbH Germany, part of Springer Nature 2023

【 图 表 】

Fig. 1

Fig. 5

13690_2023_1170_Article_IEq95.gif

Fig. 3

Fig. 4