Patients
Ninety patients, including 60 PD cases and 30 MSA cases from December 2017 to June 2019, were retrospectively enrolled in our study. Inclusion criteria were as follows: (1) the diagnosis of PD was based on the Movement Disorder Society PD Criteria [30], while MSA was diagnosed based on the second consensus statement on the diagnosis of multiple system atrophy [31] (which were international standard diagnostic criteria for PD and MSA for unavailable pathological diagnoses), and all fulfilling diagnoses were confirmed by neurological physicians with more than 10 years of experience; (2) disease duration was no more than 5 years and the average follow-up period was more than 1.5 years; and (3) all patients had undergone 18F-FDG PET/MRI examination with MRI sequences including T1-weighted imaging (T1WI), T2-weighted imaging (T2WI), T2-weighted fluid-attenuated inversion recovery (T2/FLAIR) imaging, diffusion-weighted imaging (DWI), and susceptibility-weighted imaging (SWI). Exclusion criteria were: (1) evidence of vascular disease confirmed on computed tomography (CT) or MRI; (2) organic lesions (such as trauma, tumours and infections) and other degenerative diseases; (3) severe motion artefacts on images and significant head movement during the scan; and (4) a blood glucose level was ≥ 11.1 mmol/L.
The patient screening process is shown in the study flowchart in Fig. 1a. Patients cohorts were randomly divided into a training set (n=63) and validation set (n=27) with a ratio of 7:3. This study is a retrospective study based on data from one of our clinical studies, in which all patients signed informed consent forms and ethical approval was obtained by the Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology.
Image acquisition and reconstruction of 18F-FDG PET/MRI
PET/MRI images were acquired on a 3.0T time-of-flight (TOF) Signa PET/MRI (GE Healthcare, Milwaukee WI, USA). All participants fasted for at least 6 h and stopped any drugs that could affect brain metabolism for at least 12 h before the 18F-FDG PET/MRI acquisition. An 18F-FDG dose of 0.1 mCi/kg (3.7 MBq/kg) was intravenously injected after ensuring the blood glucose level was < 11.1 mmol/L. The participants rested in a quiet and dimly lit room before and after the 18F-FDG injection until the start of imaging. At 1h after intravenous injection of 18F-FDG, an MRI and a 15-min PET scan in the three-dimensional mode were acquired. MR imaging protocols included T1WI (time of Repetition [TR]/time of echo [TE], 7.9 ms/3.0 ms; flip angles 12.0°; bandwidth 41.67 MHz; matrix 288 × 224; section thickness 1 mm), T2/FLAIR (TR/TE, 9000 ms/100 ms; inversion time 2475; refocus flip angles 160.0°; bandwidth 41.67 MHz; matrix 256 × 192; section thickness 5 mm; intersection gap 1.5 mm), T2WI (TR/TE, 5523 ms/105 ms; refocus flip angles 142.0°; bandwidth 50.00 MHz; matrix 384 × 240; section thickness 5mm; intersection gap 1.5mm), SWI (TR/TE, 45.5 ms/4.0 ms; flip angles 15.0°; bandwidth 41.67 MHz; matrix 384 × 320; section thickness 3 mm), and DWI (TR/TE, 6379.0 ms/70.0 ms; bandwidth 41.67 MHz; matrix 128 × 128; section thickness 5 mm; intersection gap 1.5mm). MRI images were acquired simultaneously with the PET acquisition without re-positioning. The apparent diffusion coefficient (ADC) map (b = 1000) was generated from DWI images. An atlas-based method was used for PET attenuation correction. For PET imaging, the ordered subsets expectation maximization (OSEM) iterative reconstruction algorithm with 28 subsets, 2 iterations, and 2.14 mm (full width at half maximum) post-filtering was used.
Region-of-interest segmentation, image pre-processing and feature extraction
The radiomics workflow is shown in Fig. 1b. The bilateral putamina and caudate nuclei were selected as the regions of interest (ROIs), which were segmented on T1WI images though the open-source software ITK-SNAP (version 3.6.0, www.itksnap.org). To minimize partial volume effects, these ROIs excluded the most inferior and most superior slices including these structures [32]. The ROIs were delineated manually by a nuclear medicine physician with 2-3 years’ experience (Hu X) who was blinded to subject information, and repeated measurements were performed at an interval of 6 weeks. All ROIs were confirmed by two neuroradiologists who had over 10 years of experience (Sun X and Liu F). The intra-observer differences were calculated by the intraclass correlation coefficient (ICC).
Based on the T1WI images, spatial registration of PET and other MR images was carried out using the SPM software package (Version 12.0, http://www.fil.ion.ucl.ac.uk/spm/) implemented in MATLAB 2016a (MathWorks, Natick, MA, USA) to provide the same spatial information (thickness, slice and interlamellar space). 18F-FDG PET images were transformed into SUV maps by normalisation by injected dose and patients’ weight using LIFEx software (version 6.20; www.lifexsoft.org), to provide the SUV values (SUVmax, SUVmean, SUVmin) in the corresponding ROIs.
The radiomics features were extracted using Anaconda Prompt (version 4.2.0) importing the feature package of pyradiomics (github.com/Radiomics/pyradiomics), according to the feature guidelines of the Image Biomarker Standardization Initiative (IBSI) [33, 34].
Optimal multimodal radiomics signature construction
To improve diagnostic performance and control scanning time to reduce the waste of medical resources, an optimal multimodal radiomics signature was constructed in the following three steps: (1) single-sequence radiomics signatures: 18F-FDG, T1WI, T2WI, T2/FLAIR, SWI, and DWI, respectively; (2) double-sequence radiomics signatures: 18F-FDG plus structural MRI (sMRI) sequences (18F-FDG + T1WI, 18F-FDG + T2WI, 18F-FDG + T2/FLAIR), or 18F-FDG plus functional MRI (fMRI) sequences (18F-FDG + SWI, 18F-FDG + DWI); (3) optimal multimodal radiomics signatures: 18F-FDG + the best sMRI + the best fMRI, according to the results from the first and second steps.
We employed the same strategy of feature selection and model construction above all three steps. The datasets were randomly divided into a training set and a validation set with a case number ratio of 7:3. The minimal redundancy maximal relevance (mRMR) algorithm, which can considerably improve the accuracy of feature selection and classification [35], was performed for initial feature selection in the training set. The least absolute shrinkage and selection operator (LASSO) method, which is suitable for the regression of high-dimensional data, was used to select significant distinguishable features to construct the radiomics signature with 10-fold cross-validation.
Clinical-radiomics diagnostic model construction
The clinical variables, including clinical characteristics (age, sex, weight, pre-injection glucose levels, disease duration [DD], age at onset, hypermyotonia, asymmetric symptoms at onset, bradykinesia, limbs tremor, dysarthria, and autonomic failure [AF]) and SUV values, were collected and compared between PD and MSA in training and validation sets. The clinical-radiomics model was constructed by combining the clinical variables with the optimal multimodal radiomics signature obtained above. The clinical-radiomics model was built by a multivariate logistic regression model with 10-fold cross-validation to distinguish PD and MSA through a likelihood ratio test with back-ward step-down. A nomogram was then constructed on the basis of the clinical-radiomics model.
Model effectiveness evaluation
The area under the receiver-operating characteristic (ROC) curve (AUC) was used to evaluate the diagnostic performance of models constructed by the training set and validated by the validation set, whereby the radiomics score (Radscore) was calculated via the formula built in the training set. The accuracy of the radiomics signature was evaluated in both the training and validation sets. The models’ calibration was assessed using calibration curves and the Hosmer–Lemeshow test; decision curve analysis (DCA) was performed to estimate the clinical utility of models.
Statistical analysis
Statistical analysis was performed by R 3.6.1 (www.Rproject.org). The packages in R used in this study were tidyverse, caret, pROC, glmnet, DMWR, rmda, ggpubr, ModelGood, rms, mRMRe, DescTOOLs and irr. The Delong test was applied to compare the differences in ROC curves between two arbitrary models by Medcalc (www.medcalc.org).
The differences in demographic and clinical variables were compared between patients with PD and MSA in both the training set and validation set by Graphpad prism 8 (www.graphpad-prism.cn). The Mann-Whitney U-test was used for non-normally distributed quantitative data; for normally distributed data, the independent sample t-test was used. Chi-squared tests were used for categorical data.