Search

Rationale and Objectives: Parkinson's disease (PD) shows small structural changes in nigrostriatal pathways, which can be sensitively captured through diffusion kurtosis imaging (DKI). However, the value of DKI and its radiomic features in the classification performance of PD still need confirmation. This study aimed to compare the diagnostic efficiency of DKI-derived kurtosis metric and its radiomic features with different machine learning models for PD classification., Materials and Methods: 75 people with PD and 80 healthy individuals had their brains scanned using DKI. These images were pre-processed and the standard atlas were non-linearly registered to them. With the labels in atlas, different brain regions in nigrostriatal pathways, including the caudate nucleus, putamen, pallidum, thalamus, and substantia nigra, were chosen as the region of interests (ROIs) to warped to the native space to measure the mean kurtosis (MK). Additionally, new radiomic features were developed for comparison. To handle the large amount of data, a statistical method called Z-score normalization and another method called LASSO regression were used to simplify the information. From this, a few most important features were chosen, and a combined score called Radscore was calculated using LASSO regression. For the comprehensive analyses, three different conventional machine learning models were then created: logistic regression (LR), support vector machine (SVM), and random forest (RF). To ensure the models were accurate, a process called 10-fold cross-validation was used, where the data were split into 10 parts for training and testing., Results: Using MK alone, the models achieved good results in correctly identifying PD in the validation set, with LR at 0.90, RF at 0.93, and SVM at 0.90. When the radiomic features were added, the models performed even better, with LR at 0.92, RF at 0.95, and SVM at 0.91. Additionally, a nomogram combining all the information was created to predict the likelihood of someone having PD, which had an AUC of 0.91., Conclusion: These findings suggest that the combination of DKI measurements and radiomic features can effectively diagnose PD by providing more detailed information about the brain's condition and the processes involved in the disease., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 The Association of University Radiologists. Published by Elsevier Inc. All rights reserved.)

Background: Unfortunately, the morphologic magnetic resonance imaging (MRI) is unable to determine perineural invasion (PNI) status. This study applied histogram analysis of intravoxel incoherent motion diffusion-weighted imaging (IVIM-DWI) in the assessment of PNI status of rectal cancer (RC)., Methods: The retrospective analysis enrolled 175 patients with RC confirmed by postoperative pathology in The First Affiliated Hospital of Shandong First Medical University from January 2019 to December 2021. All patients underwent preoperative rectal MRI. Whole-tumor volume histogram features from IVIM-DWI were extracted using open-source software. Univariate analysis and multivariate logistic regression analysis were used to compare the differences in histogram parameters and clinical features between the PNI-positive group and PNI-negative group. Receiver operating characteristic curve analysis was used to evaluate the diagnostic performance, while the Delong test was used to compare the area under the curve of the models., Results: The interobserver agreement of the histogram features derived from DWI, including apparent diffusion coefficient (ADC), true diffusion coefficient (D), pseudo-diffusion coefficient (D*), perfusion fraction (f), water molecular diffusion heterogeneity index (α), and distributed diffusion coefficient (DDC) were good to excellent. A total of eight histogram features including DWI&#95;maximum, DWI&#95;skewness, D&#95;kurtosis, D&#95;minimum, D&#95;skewness, D*&#95;energy, D*&#95;skewness, and f&#95;minimum were significantly different between the PNI-positive and PNI-negative groups in the univariate analysis (P<0.05); among the clinicoradiologic factors, percentage of rectal wall circumference invasion (PCI) was significantly different between the two groups (P<0.05). Multivariate analysis demonstrated that the values of D*&#95;energy, D*&#95;skewness, and f&#95;minimum differed significantly between the PNI-positive patients and PNI-negative patients (P<0.05), with the independent risk factors being D*&#95;skewness [odds ratio (OR) =1.157; 95% confidence interval (CI): 1.050-1.276; P=0.003] and PCI (OR =11.108, 95% CI: 1.767-69.838; P=0.0002). The area under the curve of the model combining the three histogram features and PCI to assess PNI status in RC was 0.807 (95% CI: 0.741-0.863). The results of the Delong test showed that the combined model was significantly different from each single-parameter model (P<0.05)., Conclusions: The combined model constructed on the basis of IVIM-DWI histogram features may help to assess the status of RC PNI., Competing Interests: Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-23-1614/coif). W.D. was an employee of GE HealthCare throughout his involvement in the study. The other authors have no conflicts of interest to declare., (2024 Quantitative Imaging in Medicine and Surgery. All rights reserved.)

Background: The noninvasive prediction of sentinel lymph node (SLN) metastasis using quantitative magnetic resonance imaging (MRI), particularly with synthetic MRI (syMRI), is an emerging field. This study aimed to explore the potential added benefits of syMRI over conventional MRI and diffusion-weighted imaging (DWI) in predicting metastases in SLNs., Methods: This retrospective study consecutively enrolled 101 patients who were diagnosed with breast cancer (BC) and underwent SLN biopsy from December 2022 to October 2023 at the Affiliated Hospital of Jiangnan University. These patients underwent preoperative MRI including conventional MRI, DWI, and syMRI and were categorized into two groups according to the postoperative pathological results: those with and without metastatic SLNs. MRI morphological features, DWI, and syMRI-derived quantitative parameters of breast tumors were statistically compared between these two groups. Binary logistic regression was used to separately develop predictive models for determining the presence of SLN involvement, with variables that exhibited significant differences being incorporated. The performance of each model was evaluated through receiver operating characteristic (ROC) curve analysis, including the area under the curve (AUC), sensitivity, and specificity., Results: Compared to the group of 54 patients with BC but no metastatic SLNs, the group of 47 patients with BC and metastatic SLNs had a significantly larger maximum axis diameter [metastatic SLNs: median 2.40 cm, interquartile range (IQR) 1.50-3.00 cm; no metastatic SLNs: median 1.80 cm, IQR 1.37-2.50 cm; P=0.03], a higher proton density (PD) (78.44±11.92 vs. 69.20±10.63 pu; P<0.001), and a lower apparent diffusion coefficient (ADC) (metastatic SLNs: median 0.91×10 -3 mm 2 /s, IQR 0.79-1.01 mm 2 /s; no metastatic SLNs: median 1.02×10 -3 mm 2 /s, IQR 0.92-1.12 mm 2 /s; P=0.001). Moreover, the prediction model with maximum axis diameter and ADC yielded an AUC of 0.71 [95% confidence interval (CI): 0.618-0.802], with a sensitivity of 78.72% and a specificity of 51.85%; After addition of syMRI-derived PD to the prediction model, the AUC increased significantly to 0.86 (AUC: 0.86 vs. 0.71; 95% CI: 0.778-0.922; P=0.002), with a sensitivity of 80.85% and a specificity of 81.50%., Conclusions: Combined with conventional MRI and DWI, syMRI can offer additional value in enhancing the predictive performance of determining SLN status before surgery in patients with BC., Competing Interests: Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-24-1/coif). J.S. and W.D. are employees of GE HealthCare. The other authors have no conflicts of interest to declare., (2024 Quantitative Imaging in Medicine and Surgery. All rights reserved.)

Objectives: To systematically test the reproducibility of DKI technique in normal liver and report a complete set of DKI measurement data., Materials and Methods: Thirty-two healthy volunteers were examined with liver DKI twice on the GE 3.0 T MRI scanner and reviewed by three professional experts. DKI-derived parameters fractional anisotropy of kurtosis (FAk), mean diffusivity (Md), axial diffusivity (Da), radial diffusivity (Dr), mean kurtosis (Mk), axial kurtosis (Ka), and radial kurtosis (Kr) in eight segments divided by Couinaud octagonal method were collected. Inter-class correlation coefficient (ICC) was used to assess the agreement between three experts. For each expert, the reproducibility of twice scans was evaluated by Bland-Altman method. Multivariate analysis of variance was to explore the regional distribution characteristics of DKI-derived parameters, and showed with box-plot graph., Results: Using ICC analysis, except for FAk (ICC 0.312, 0.307), other DKI metric values showed high reproducibility (0.716 < ICC < 0.907) between three experts for each of two DKI measurements. With Bland-Altman method, liver segment 5 (S5) showed the best reproducibility between two DKI measurement, and the reproducibility of segment 4 (S4) was the worst. The reproducibility of the right lobe was significantly higher than the left lobe. The values of diffusion metrics (Md, Da, and Dr) and kurtosis metrics (Mk, Ka, and Kr) existed significantly difference between the right and left hepatic lobes., Conclusion: DKI has shown excellent reproducibility in liver imaging. The range of values for multiple DKI parameters, derived from the normal liver, was reported, and may provide data reference for further clinical DKI applications. Additionally, DKI technique is a non-invasive method to reflect the perfusion or structural differences between the left and right hepatic lobes from the molecular level.

Objectives: To compare diffusion parameters obtained from mono-exponential, bi-exponential, and stretched-exponential diffusion-weighted imaging (DWI) in stratifying non-alcoholic fatty liver disease (NAFLD)., Methods: Thirty-two New Zealand rabbits were fed a high-fat/cholesterol or standard diet to obtain different stages of NAFLD before 12 b-values (0-800 s/mm 2 ) DWI. The apparent diffusion coefficient (ADC) from the mono-exponential model; pure water diffusion (D), pseudo-diffusion (D*), and perfusion fraction (f) from bi-exponential DWI; and distributed diffusion coefficient (DDC) and water molecular diffusion heterogeneity index (α) from stretched-exponential DWI were calculated for hepatic parenchyma. The goodness of fit of the three models was compared. NAFLD severity was pathologically graded as normal, simple steatosis, borderline, and non-alcoholic steatohepatitis (NASH). Spearman rank correlation analysis and receiver operating characteristic curves were used to assess NAFLD severity., Results: Upon comparison, the goodness of fit chi-square from stretched-exponential fitting (0.077 ± 0.012) was significantly lower than that for the bi-exponential (0.110 ± 0.090) and mono-exponential (0.181 ± 0.131) models (p < 0.05). Seven normal, 8 simple steatosis, 6 borderline, and 11 NASH livers were pathologically confirmed from 32 rabbits. Both α and D increased with increasing NAFLD severity (r = 0.811 and 0.373, respectively; p < 0.05). ADC, f, and DDC decreased as NAFLD severity increased (r = - 0.529, - 0.717, and - 0.541, respectively; p < 0.05). Both α (area under the curve [AUC] = 0.952) and f (AUC = 0.931) had significantly greater AUCs than ADC (AUC = 0.727) in the differentiation of NASH from borderline or less severe groups (p < 0.05)., Conclusions: Stretched-exponential DWI with higher fitting efficiency performed, as well as bi-exponential DWI, better than mono-exponential DWI in the stratification of NAFLD severity., Key Points: • Stretched-exponential diffusion model fitting was more reliable than the bi-exponential and mono-exponential diffusion models (p = 0.039 and p < 0.001, respectively). • As NAFLD severity increased, the diffusion heterogeneity index (α) increased, while the perfusion fraction (f) decreased (r = 0.811, - 0.717, p < 0.05). • Both α and f showed superior NASH diagnostic performance (AUC = 0.952, 0.931) compared with ADC (AUC = 0.727, p < 0.05).

Purpose: The purpose of this study was to assess and compare the diagnostic performances of preoperative ultrasonography (US) and magnetic resonance imaging (MRI) in predicting extrathyroidal extension (ETE) in patients with papillary thyroid carcinoma (PTC)., Materials and Methods: This retrospective study was approved by our institutional review board. Preoperative US and MRI were performed on 225 patients who underwent surgery for PTC between May 2014 and December 2018. The US and MRI features of ETE of each case were retrospectively and independently investigated by two radiologists. The diagnostic performances of US and MRI, including their sensitivity, specificity, positive predictive values (PPV), and negative predictive values (NPV) for ETE, and their accuracy in predicting ETE were analyzed., Results: Higher sensitivity and NPV in predicting minimal ETE were observed in US (87.5% and 76.2%, respectively) compared with MRI (71.3% and 61.7%, respectively) (p = 0.006 and p = 0.046, respectively). Meanwhile, MRI (85.4%) showed higher sensitivity than US (66.7%) in assessing extensive ETE (p = 0.005). MRI also showed significantly higher specificity and PPV than US in assessing overall ETE (p = 0.025 and p = 0.025, respectively)., Conclusion: Preoperative US should be used as the first line in predicting minimal ETE, and MRI should be added in extensive ETE assessment. Compared with US, MRI had higher specificity and PPV in detecting the overall ETE of PTC.

Purpose: This study aimed to investigate the relationship between the fractional anisotropy (FA) values of compressed nerves derived in diffusion tensor imaging (DTI) and the corresponding clinical symptoms for quantitative and clinical evaluation in patients with lumbosacral radiculopathy., Methods: Thirty-six patients and ten volunteers participated in the study and measured with DTI. The resultant FA values for L5-S1 lumbar nerve roots were calculated at three sub-regions. Additionally, the DTI relevant tractography was also performed on L4-S1 nerve roots. Clinical symptoms were performed by Japanese Orthopedic Association (JOA) scoring for each patient and volunteer., Results: The FA values of the nerves at the symptomatic side were significantly lower than those at the asymptomatic side (p < 0.001). Diffusion tensor tractography distinctly showed abnormalities in the symptomatic nerve tracts. There was a significant correlation between JOA scores and the FA values of the compressed nerves at middle and distal sub-regions (p < 0.005)., Conclusion: The clinical symptoms associated robustly with the DTI derived FA values of the compressed nerves in patients with lumbosacral radiculopathy. Therefore, the FA values can be a potential clinical tool to evaluate the nerve roots in lumbosacral radiculopathy quantitatively.

Magnetic resonance imaging (MRI) is a non-invasive diagnostic imaging tool based on the detection of protons into the tissues. This imaging technique is remarkable because of high spatial resolution, strong soft tissue contrast and specificity, and good depth penetration. However, MR imaging of hard tissues, such as bone and teeth, remains challenging due to low proton content in such tissues as well as to very short transverse relaxation times (T 2 ). To overcome these issues, new MRI techniques, such as sweep imaging with Fourier transformation (SWIFT), ultrashort echo time (UTE) imaging, and zero echo time (ZTE) imaging, have been developed for hard tissues imaging with promising results reported. Within this article, MRI techniques developed for the detection of hard tissues, such as bone and dental tissues, have been reviewed. The main goal was thus to give a comprehensive overview on the corresponding (pre-) clinical applications and on the potential future directions with such techniques applied. In addition, a section dedicated to MR imaging of novel biomaterials developed for hard tissue applications was given as well.

Purpose: To examine the feasibility of MR diffusion kurtosis imaging (DKI) for characterizing nonalcoholic fatty liver disease (NAFLD) and diagnosing nonalcoholic steatohepatitis (NASH)., Methods: Thirty-two rabbits on high fat diet with different severities of NAFLD were imaged at 3 T MR including diffusion weighted imaging (DWI) and DKI using b values of 0, 400, 800 s/mm 2 with 15 diffusion directions at each b value. Apparent diffusion coefficient (ADC) was derived from the linear exponential DWI model. Mean diffusion (MD) and mean kurtosis (MK) were derived from the quadratic exponential model of DKI. Correlations between MR parameters and hepatic pathology determined by the NAFLD activity scoring system were analyzed by Spearman rank correlation analysis. Receiver operating characteristic analyses were applied to determine the cutoff values of MD, MK as well as ADC in distinguishing NASH from non-NASH. The diagnostic efficacies of MD and MK in detecting NASH were compared with that of ADC., Results: Values for ADC and MD significantly decreased as the severity of NAFLD increased (ρ = -0.529, -0.904, respectively; P < 0.05). MK values significantly increased as the severity of NAFLD increased (ρ = 0.761; P < 0.05). In addition, both MD and MK values were significantly different between borderline NASH and NASH groups (MD: 1.729 ± 0.144 vs. 1.458 ± 0.240[×10 -3  mm 2 /s]; MK: 1.096 ± 0.079 vs. 1.237 ± 0.180; P < 0.05). Moreover, there was a significantly higher area under the curve (AUC) for both MD (0.955) and MK (0.905), as compared to ADC (0.736)., Conclusion: Diffusion kurtosis imaging was feasible for stratifying NAFLD, and more accurately discriminated NASH from non-NASH when compared with DWI., (Copyright © 2019 Elsevier Inc. All rights reserved.)

Monitoring hydrogel degradation in real time using noninvasive imaging techniques is of great interest for designing a scaffold in tissue engineering. We report the preparation of gadolinium (Gd)-labeled and injectable hyaluronic acid (HA) hydrogels that can be visualized using T 1 - and T 2 -weighted magnetic resonance imaging (MRI). An HA derivative functionalized with thiol and hydrazide was labeled using a diethylenetriaminepentaacetate complex modified with "clickable" dithiopyridyl functionalities (degree of modification was 3.77% with respect to HA repeat units). The HA derivative modified with cross-linkable groups and Gd complex exhibited relaxivities r 1  = 3.78 mM -1 s -1 and r 2  = 56.3 mM -1 s -1 . A hydrazone hydrogel network was obtained by mixing Gd-labeled HA-hydrazide and HA-aldehyde derivatives. Enzymatic hydrogel degradation could be followed using MRI because the MR images showed great correlation with the hydrogel mass loss. Ex vivo MRI of injected Gd-labeled hydrogels demonstrated that they show a significant contrast difference (SNR coronal  = 456; SNR axial  = 459) from the surrounding tissues. These results indicate that our Gd-labeled HA hydrogel has great potential as an injectable biocompatible hydrogel that can be used for longitudinal tracking in vivo using MRI., (Copyright © 2018 Elsevier Ltd. All rights reserved.)

Purpose: To test if the proposed deep learning based denoising method denoising convolutional neural networks (DnCNN) with residual learning and multi-channel strategy can denoise three dimensional MR images with Rician noise robustly., Materials and Methods: Multi-channel DnCNN (MCDnCNN) method with two training strategies was developed to denoise MR images with and without a specific noise level, respectively. To evaluate our method, three datasets from two public data sources of IXI dataset and Brainweb, including T1 weighted MR images acquired at 1.5 and 3 T as well as MR images simulated with a widely used MR simulator, were randomly selected and artificially added with different noise levels ranging from 1 to 15%. For comparison, four other state-of-the-art denoising methods were also tested using these datasets., Results: In terms of the highest peak-signal-to-noise-ratio and global of structure similarity index, our proposed MCDnCNN model for a specific noise level showed the most robust denoising performance in all three datasets. Next to that, our general noise-applicable model also performed better than the rest four methods in two datasets. Furthermore, our training model showed good general applicability., Conclusion: Our proposed MCDnCNN model has been demonstrated to robustly denoise three dimensional MR images with Rician noise.

1 H magnetic resonance imaging (MRI) by a zero echo time (ZTE) sequence is an excellent method to image teeth. Calcium phosphate cement (CPC) materials are applied in the restoration of tooth lesions, but it has not yet been investigated whether they can be detected by computed tomography (CT) or MRI. The aim of this study was to optimize high-field ZTE imaging to enable the visualization of a new CPC formulation implanted in teeth and to apply this in the assessment of its decomposition in vivo. CPC was implanted in three human and three goat teeth ex vivo and in three goat teeth in vivo. An ultrashort echo time (UTE) sequence with multiple flip angles and echo times was applied at 11.7 T to measure T 1 and T 2 * values of CPC, enamel and dentin. Teeth with CPC were imaged with an optimized ZTE sequence. Goat teeth implanted with CPC in vivo were imaged after 7 weeks ex vivo. T 2 * relaxation of implanted CPC, dentin and enamel was better fitted by a model assuming a Gaussian rather than a Lorentzian distribution. For CPC and human enamel and dentin, the average T 2 * values were 273 ± 19, 562 ± 221 and 476 ± 147 μs, respectively, the average T 2 values were 1234 ± 27, 963 ± 151 and 577 ± 41 μs, respectively, and the average T 1 values were 1065 ± 45, 972 ± 40 and 903 ± 7 ms, respectively. In ZTE images, CPC had a higher signal-to-noise-ratio than dentin and enamel because of the higher water content. Seven weeks after in vivo implantation, the CPC-filled lesions showed less homogeneous structures, a lower T 1 value and T 2 * separated into two components. MRI by ZTE provides excellent contrast for CPC in teeth and allows its decomposition to be followed., (Copyright © 2017 John Wiley & Sons, Ltd.)

Calcium phosphate cement (CPC) is used in bone repair because of its biocompatibility. However, high similarity between CPC and the natural osseous phase results in poor image contrast in most of the available in vivo imaging techniques such as computed tomography (CT) and magnetic resonance imaging (MRI). For accurate identification and localization during and after implantation in vivo, a composition with enhanced image contrast is needed. In this study, we labeled CPC with perfluoro-15-crown-5-ether-loaded (PFCE) poly(latic-co-glycolic acid) nanoparticles (hydrodynamic radius 100 nm) and gold nanoparticles (diameter 40 nm), as 19 F MRI and CT contrast agents, respectively. The resulting CPC/PFCE/gold composite is implanted in a rat model for in vivo longitudinal imaging. Our findings show that the incorporation of the two types of different nanoparticles did result in adequate handling properties of the cement. Qualitative and quantitative long-term assessment of CPC/PFCE/gold degradation was achieved in vivo and correlated to the new bone formation. Finally, no adverse biological effects on the bone tissue are observed via histology. In conclusion, an easy and efficient strategy for following CPC implantation and degradation in vivo is developed. As all materials used are biocompatible, this CPC/PFCE/gold composite is clinically applicable.