A more objective PD diagnostic model: integrating texture feature markers of cerebellar gray matter and white matter through machine learning.

Objective: The purpose of this study is to explore whether machine learning can be used to establish an effective model for the diagnosis of Parkinson's disease (PD) by using texture features extracted from cerebellar gray matter and white matter, so as to identify subtle changes that cannot be observed by the naked eye. Method: This study involved a data collection period from June 2010 to March 2023, including 374 subjects from two cohorts. The Parkinson's Progression Markers Initiative (PPMI) served as the training set, with control group and PD patients (HC: 102 and PD: 102) from 24 global sites. Our institution's data was utilized as the test set (HC: 91 and PD: 79). Machine learning was employed to establish multiple models for PD diagnosis based on texture features of the cerebellum's gray and white matter. Results underwent evaluation through 5-fold cross-validation analysis, calculating the area under the receiver operating characteristic curve (AUC) for each model. The performance of each model was compared using the Delong test, and the interpretability of the optimized model was further augmented by employing Shapley additive explanations (SHAP). Results: The AUCs for all pipelines in the validation dataset were compared using FeAture Explorer (FAE) software. Among the models established by Kruskal-Wallis (KW) and logistic regression via Lasso (LRLasso), the AUC was highest using the "one-standard error" rule. 'WM_original_glrlm_GrayLevelNonUniformity' was considered the most stable and predictive feature. Conclusion: The texture features of cerebellar gray matter and white matter combined with machine learning may have potential value in the diagnosis of Parkinson's disease, in which the heterogeneity of white matter may be a more valuable imaging marker.

Implementation of Individualized Low-Dose Computed Tomography-Guided Hook Wire Localization of Pulmonary Nodules: Feasibility and Safety in the Clinical Setting.

Background: CT-guided hook-wire localization is an essential step in the management of small pulmonary nodules. Few studies, however, have focused on reducing radiation exposure during the procedure. Purpose: This study aims to explore the feasibility of implementing a low-dose computed tomography (CT)-guided hook wire localization using tailored kVp based on patients' body size. Materials and Methods: A total of 151 patients with small pulmonary nodules were prospectively enrolled for CT-guided hook wire localization using individualized low-dose CT (LDCT) vs. standard-dose CT (SDCT) protocols. Radiation dose, image quality, characteristics of target nodules and procedure-related variables were compared. All variables were analyzed using Chi-Square and Student's t-test. Results: The mean CTDIvol was significantly reduced for LDCT (for BMI ≤ 21 kg/m2, 0.56 ± 0.00 mGy and for BMI > 21 kg/m2, 1.48 ± 0.00 mGy) when compared with SDCT (for BMI ≤ 21 kg/m2, 5.24 ± 0.95 mGy and for BMI > 21 kg/m2, 6.69 ± 1.47 mGy). Accordingly, the DLP of LDCT was significantly reduced as compared with that of SDCT (for BMI ≤ 21 kg/m2, 56.86 ± 4.73 vs. 533.58 ± 122.06 mGy.cm, and for BMI > 21 kg/m2, 167.02 ± 38.76 vs. 746.01 ± 230.91 mGy.cm). In comparison with SDCT, the effective dose (ED) of LDCT decreased by an average of 89.42% (for BMI ≤ 21 kg/m2) and 77.68% (for BMI > 21 kg/m2), respectively. Although the images acquired with the LDCT protocol yielded inferior quality to those acquired with the SDCT protocol, they were clinically acceptable for hook wire localization. Conclusions: LDCT-guided localization can provide safety and nodule detection performance comparable to SDCT-guided localization, benefiting radiation dose reduction dramatically, especially for patients with small body mass indexes.

Quantification of the renal sinus fat and exploration of its relationship with ectopic fat deposition in normal subjects using MRI fat fraction mapping.

Purpose: To determine the renal sinus fat (RSF) volume and fat fraction (FF) in normal Chinese subjects using MRI fat fraction mapping and to explore their associations with age, gender, body mass index (BMI) and ectopic fat deposition. Methods: A total of 126 subjects were included in the analysis. RSF volume and FF, visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) area, and hepatic and pancreatic FFs were measured for each subject. The comparisons in gender were determined using two-tailed t-tests or the nonparametric Mann-Whitney U-test for normally or non-normally distributed data for continuous variables and the chi-square test for categorical variables. Comparisons of RFS volume and FF between right and left kidneys were determined using paired sample t-tests. Multivariable logistic models were performed to confirm whether RSF differences between men and women are independent of VAT or SAT area. When parameters were normally distributed, the Pearson correlation coefficient was used; otherwise, the Spearman correlation coefficient was applied. Results: The RSF volumes (cm3) of both kidneys in men (26.86 ± 8.81 for right and 31.62 ± 10.32 for left kidneys) were significantly bigger than those of women (21.47 ± 6.90 for right and 26.03 ± 8.55 for left kidneys) (P < 0.05). The RSF FFs (%) of both kidneys in men (28.33 ± 6.73 for right and 31.21 ± 6.29 for left kidneys) were significantly higher than those of the women (23.82 ± 7.74 for right and 27.92 ± 8.15 for left kidneys) (P < 0.05). The RSF differences between men and women are independent of SAT area and dependent of VAT area (except for right RSF volume). In addition, the RSF volumes and FFs in both kidneys in the overall subjects show significant correlations with age, BMI, VAT area, hepatic fat fraction and pancreatic fat fraction (P < 0.05). However, the patterns of these correlations varied by gender. The RSF volume and FF of left kidney were significantly larger than those of the right kidney (P < 0.05). Conclusion: The association between renal sinus fat and ectopic fat deposition explored in this study may help establish a consensus on the normal values of RSF volume and FF for the Chinese population. This will facilitate the identification of clinicopathological changes and aid in the investigation of whether RSF volume and FF can serve as early biomarkers for metabolic diseases and renal dysfunction in future studies.

Assessing the Blood Supply Status of the Focal Ground-Glass Opacity in Lungs Using Spectral Computed Tomography.

Objective: To exploit material decomposition analysis in dual-energy spectral computed tomography (CT) to assess the blood supply status of the ground-glass opacity (GGO) in lungs. Materials and Methods: This retrospective study included 48 patients with lung adenocarcinoma, who underwent a contrast-enhanced dual-energy spectral CT scan before treatment (53 GGOs in total). The iodine concentration (IC) and water content (WC) of the GGO, the contralateral and ipsilateral normal lung tissues were measured in the arterial phase (AP) and their differences were analyzed. IC, normalized IC (NIC), and WC values were compared between the pure ground-glass opacity (pGGO) and the mixed ground-glass opacity (mGGO), and between the group of preinvasive lesions and the minimally invasive adenocarcinoma (MIA) and invasive adenocarcinoma (IA) groups. Results: The values of pGGO (IC = 20.9 ± 6.2 mg/mL and WC = 345.1 ± 87.1 mg/mL) and mGGO (IC = 23.8 ± 8.3 mg/mL and WC = 606.8 ± 124.5 mg/mL) in the AP were significantly higher than those of the contralateral normal lung tissues (IC = 15.0 ± 4.9 mg/mL and WC = 156.4 ± 36.8 mg/mL; IC = 16.2 ± 5.7 mg/mL and WC = 169.4 ± 41.0 mg/mL) and ipsilateral normal lung tissues (IC = 15.1 ± 6.2 mg/mL and WC = 156.3 ± 38.8 mg/mL; IC = 15.9 ± 6.0 mg/mL and WC = 174.7 ± 39.2 mg/mL; all p < 0.001). After normalizing the data according to the values of the artery, pGGO (NIC = 0.1 and WC = 345.1 ± 87.1 mg/mL) and mGGO (NIC = 0.2 and WC = 606.8 ± 124.5 mg/mL) were statistically different (p = 0.049 and p < 0.001, respectively), but not for the IC value (p = 0.161). The WC values of the group with preinvasive lesions and MIA (345.4 ± 96.1 mg/mL) and IA (550.1 ± 158.2 mg/mL) were statistically different (p < 0.001). Conclusion: Using dual-energy spectral CT and material decomposition analysis, the IC in GGO can be quantitatively measured which can be an indicator of the blood supply status in the GGO.

Feasibility of spectral imaging with low-concentration contrast medium in abdominal CT angiography of obese patients.

OBJECTIVE: To explore the application value of dual-energy spectral imaging scanning with low radiation dose and low-concentration contrast medium (270 mg I/mL) in abdominal CT angiography (CTA) of obese patients. METHODS: A total of 127 obese patients (BMI≥25 kg/m(2) and waist circumference ≥900 mm) referred for abdominal CTA were prospectively enrolled in the study. The patients were divided into two groups; in group A (n=69), a spectral imaging scan mode and 270 mg I/mL iodine concentration contrast medium was used, and in group B (n=58), a conventional imaging scan mode using 120 kVp and 350 mg I/mL iodine concentration contrast medium was used. The image quality of the right renal artery in the two groups was evaluated by two observers using a 5-point scale, and the scores were compared using the Mann-Whitney U-test; the inter-observer agreement for the scores was analysed using the Kappa test. The CT values of the abdominal aorta, the superior mesenteric artery, the common hepatic artery and the splenic artery, and the CT value and standard deviation (SD) of the erector spinae at the level of the right renal hilum in groups A and B were measured by two observers; the inter-observer agreement of the measurement data was analysed using the inter-class correlation coefficient test. The following parameters were compared between the two groups using an independent sample t-test: the CT values of the abdominal aorta and its main branches; the image contrast-to-noise ratio (CNR) and figure of merit (FOM) of the abdominal aorta; the CT dose index (CTDIvol ); the dose length product (DLP); and the total iodine intake of the patients. P<.05 suggested a statistically significant difference. RESULTS: The image scores of the right renal artery in groups A (4.59±0.60) and B (4.53±0.63) were the same (P=.57), with good inter-observer agreement. The CT values of the abdominal aorta, the superior mesenteric artery, the common hepatic artery and the splenic artery were >300 HU in both the groups; there was no statistically significant difference between the two groups (all P>.05), and inter-observer agreement was also good. Group A had significantly higher CNR and FOM values in the abdominal aorta than group B (all P<.001). Compared with group B, the CTDIvol and DLP values in group A were decreased by 46% and 35%, respectively. The total iodine intake for patients in group A was 27 g, 23% lower than the 35 g intake for patients in group B. CONCLUSION: Compared with conventional CT, spectral CT imaging significantly reduces both radiation dose and contrast dose while maintaining image quality in abdominal CTA for patients with central obesity and high BMI.

[Preliminary study of improved quality of computed tomographic angiography on tumor blood supply arteries of gastrointestinal tract malignant tumors with low tube voltage and low concentration contrast medium].

OBJECTIVE: To assess the image quality of gastrointestinal tract malignant tumors arteries using low tube voltage, low contrast medium concentration and 50% adaptive statistical iterative reconstruction (ASiR) algorithm. METHODS: A total of 58 patients with pathologically confirmed gastrointestinal tract malignant tumors and body mass index (BMI) under 22 kg/m² undergoing computed tomography (CT) scanning during March-December 2013. They were randomly divided into group A (270 mgI/ml of contrast medium, 80 kVp tube voltage scanning and 50% ASiR; 21 males and 11 females with an age range of 40-90 years) and group B (350 mgI/ml of contrast medium, 120 kVp routine tube voltage; 22 males and 4 females with an age range of 40-76 years). The inter-group differences of age and gender ratio were compared with Mann-Whitney U and Fisher exact tests respectively. Two independent radiologists reviewed reconstructions and separated the reconstructed image into 5 points according to image quality. The CT value of tumor blood supplying artery (CT1 value) and fat in anterior abdominal wall (SD value) were measured and image noise, contrast-to-noise ratio (CNR) and figure of merit (FOM) calculated. The interobserver variation was estimated by weighted kappa statistics and intra-class correlation coefficients (ICC) test. The point of image quality of two groups was compared by Mann-Whitney U test. The paired Student t test was used to compare the inter-group differences in CT1, SD, CNR, FOM and CT dose index (CTDIvol). RESULTS: No inter-group difference existed in patient age or gender ratio. The consistency of two radiologists was excellent (kappa value > 0.80; ICC value > 0.75). The subjective image quality scores of tumor blood supplying arteries showed no inter-group difference (P = 0.09). The tumor blood supply arteries CT value, CNR and FOM of group A (459 ± 69 HU, 20.2 ± 3.3 and 85 ± 37) were significantly higher than those of group B (250 ± 42 HU, 9.3 ± 1.9 and 9 ± 4) (both P < 0.01). The SD value of group A (20.0 ± 1.7) was obviously lower than that of group B (22.4 ± 3.2) (P < 0.01). And the CTDIvol of group A (5.2 ± 1.2 mGy) was obviously lower than that of group B (13.5 ± 4.7 mGy) (P < 0.01). CONCLUSION: For patients of BMI ≤ 22 kg/m², low tube voltage and low contrast medium concentration scan condition with 50% ASiR algorithm is feasible for acquiring better quality image and ensuring significant reduction in effective dose.

The Value of Nonenhanced Single-Source Dual-Energy CT for Differentiating Metastases From Adenoma in Adrenal Glands.

RATIONALE AND OBJECTIVES: To evaluate the value of the nonenhanced single-source dual-energy computed tomography (ssDECT) in differentiating metastases from adenomas in adrenal glands. MATERIALS AND METHODS: This retrospective study was approved by our Institutional Review Board, and written informed consent was waived. One hundred twelve patients (66 men:46 women; mean age, 58 years) with 63 adrenal metastases (AMs) and 64 adrenal adenomas (AAs) underwent a plain dual-energy spectral CT imaging from August 2011 to December 2013 were included. The fat (water) density (DFa [Wa]) from the material decomposition (MD) images and CT number and effective atomic number (eff-Z) from the virtual monochromatic spectral (VMS) image sets were measured for the AMs and AAs. The spectral Hounsfield unit (HU) curve (CT number as a function of photon energy from 40 to 140 keV) was generated, and its slope (K) was calculated. The difference of these parameters between AMs and AAs was statistically compared by the Wilcoxon rank sum test. Receiver operating characteristic curve (ROC) curves were used to compare the diagnostic efficacies of these measures in the identification of AAs and AMs. The distribution of spectral HU curve was analyzed using the chi-square test in terms of its slope K: ascending (K > 0.1), straight (-0.1 ≤ K ≤ 0.1), and descending (K < -0.1). RESULTS: 1) The CT number (medium, range) of metastases (50.47, 29.93 HU at 40 keV and 29.00, 9.36 HU at 140 keV) was significantly higher than that of adenomas (-0.76, 33.04 to 13.73, 18.96 HU) at each energy level from 40 to 140 keV (P < .05). 2) The fat concentration of metastases (-177.37, 296.38 mg/mL) was statistically lower than that of adenomas (126.73, 328.07 mg/mL; P < .05). 3) The eff-Z of metastases (7.76, 0.23) was significantly higher than that of adenomas (7.42, 0.32; P < .05). 4) With CT number of VMS image at 40 keV of 21.78 HU as a threshold, the sensitivity and specificity for differentiating metastases from adenomas was 92.1% and 76.6%, respectively, and the area under the ROC curve was 0.90. 5) The spectral curve types included 3.2% (2 of 63) ascending, 20.6% (13 of 63) straight, and 76.2% (48 of 63) descending for the metastases, whereas the corresponding numbers were 60.9% (39 of 64), 21.9% (14 of 64), and 17.2% (11 of 64) for the adenomas. The difference was statistically significant (X(2) = 56.63; P < .05). CONCLUSIONS: The nonenhanced ssDECT enables a multiparametric approach to provide an excellent sensitivity for identifying AMs from AAs.

[Optimal monochromatic images of abdominal arteries with low concentrations of contrast agents in spectral computed tomography imaging].

OBJECTIVE: To explore the application of optimal monochromatic images according to computed tomography (CT) values of conventional CT scans with high concentration contrast agent and the feasibility in guiding abdominal arteries imaging with low concentration contrast. METHODS: From March 2013 to February 2014, 160 patients of suspected celiac disease with a body mass index (BMI) of >25 kg/m(2) underwent spectral CT scan. And they were randomly divided into two groups. In group A (n = 61), 270 mgI/ml low concentrations of contrast agents and spectral CT imaging; in group B (n = 99), 350 mgI/ml high concentrations of contrast agents and 120 kVp scan. For CT values of the first measurement of abdominal aorta B group of patients, the average CT value was calculated . The optimal monochromatic images of group A according to CT values of group B were reconstructed and 70 keV images (group C) also reconstructed. The CT values of hepatic, splenic, right renal and superior mesenteric arteries were measured and contrast to noise ratios (CNRs) calculated. CT values and CNR were compared through t test on independent samples. P > 0.05 indicated no significant difference. RESULTS: Mean CT value of abdominal aorta was (422 ± 65) HU in group B. And optimal keV was (67 ± 5) keV in group A. CT values of abdominal aorta and its branches showed no statistical significance; the CNRs of abdominal aorta and its branches showed statistical significance. The scoring consistency between two observers was excellent and the score (4.1 ± 0.9 vs 4.2 ± 0.7) showed no significant difference (P > 0.05). CONCLUSION: GSI abdominal scans with low concentration contrast may obtain better image qualities compared to conventional CTA of high concentration. And (67 ± 5) keV is recommended.