Quantitative Imaging in Medicine and Surgery最新文献

Background: Ossification of the posterior longitudinal ligament (OPLL) is a degenerative disease of the cervical spine. Due to the ectopic calcification of the posterior longitudinal ligament, the spinal cord or nerve roots may be compressed, producing neurological symptoms. Computed tomography (CT) is considered the "gold standard" for the diagnosis of OPLL; however, its ability to visualize soft-tissue lesions is limited. This study aimed to evaluate the utility of the magnetic resonance imaging (MRI) using the fast-field-echo resembling a computed tomography using restricted echo-spacing (FRACTURE) sequence in comparison to CT in the detection of OPLL of the cervical spine.

Methods: OPLL was evaluated in 40 patients (mean age: 56.8±9.5 years, 18 women) using CT and MRI with the FRACTURE sequence. Two radiologists evaluated the following morphological parameters on both examinations: type of OPLL, distribution of OPLL, ossification thickness, canal-occupying ratio, K-line, dural calcification, and intervertebral vacuum phenomenon. Inter-observer and inter-modality agreement were analyzed. A Bland-Altman analysis was used for the quantitative data, and the Kappa coefficient was used for the categorical data.

Results: The inter-observer agreement of FRACTURE-MRI images, and the inter-modality agreement between the FRACTURE-MRI images and CT images were calculated. Measurements of ossification thickness and canal-occupying ratio showed strong inter-observer agreement for the FRACTURE sequence, and strong inter-modality agreement between FRACTURE-MRI and CT. The inter-observer and inter-modality agreement was strong to perfect for the type of OPLL [0.864 (0.742-0.986) and 0.841 (0.697-0.984)], distribution of OPLL [0.785 (0.641-0.930) and 0.774 (0.645-0.903)], K-line [0.838 (0.662-1.000) and 0.893 (0.750-1.000)], dural calcification [0.806 (0.596-1.000) and 0.826 (0.638-1.000)], respectively. However, the consistency in detecting the intervertebral vacuum phenomenon was relatively poor, with values of 0.541 (0.186-0.896) and 0.483 (0.130-0.836), respectively.

Conclusions: MRI with the FRACTURE sequence may be sufficient for assessing OPLL of the cervical spine. Under specific conditions, the FRACTURE sequence could serve as an alternative to conventional CT examination, reducing both the associated costs and radiation exposure.

Background: Accurate recognition and quantitative assessment of cervical nerves are essential for ultrasound-guided nerve block, preoperative evaluation, and diagnosis of peripheral neuropathies. However, manual interpretation and measurement of nerve structures remain time-consuming and operator dependent. This study aimed to develop an intelligent, fully automated system for precise segmentation and quantitative measurement of cervical nerves in ultrasound images using a deep learning-based approach.

Methods: Ultrasound images from 200 healthy volunteers were collected and meticulously annotated to construct a large-scale standardized dataset comprising 117,729 images. A fully automated analysis framework was designed, incorporating a YOLOv11 network for precise localization of the effective imaging region [mean intersection over union (mIoU) =0.99] and an optical character recognition (OCR) module for automatic extraction of depth scale information. The proposed segmentation network, termed SZJ-SEG, combines a Deconv Block and an efficient upsampling convolution block (EUCB) based on a ResNet50 backbone to achieve high segmentation accuracy with low computational cost. Quantitative analysis of the C5-C7 nerve roots was performed through pixel-to-physical scale calibration for cross-sectional area (CSA) and perimeter calculation.

Results: On a clinical dataset comprising 117,729 images (training/validation/test =94,183/11,773/11,773), SZJ-SEG achieved mIoU values of 0.9124, 0.9109, and 0.9041 for the C5, C6, and C7 nerves, respectively, and 0.9227 for the continuous brachial plexus section. The mean absolute error (MAE) of CSA measurement ranged from 0.278 to 0.442 mm², with mean absolute percentage error (MAPE) between 2.43% and 6.16%, and Pearson correlation coefficients exceeding 0.96. For perimeter measurements, MAE ranged from 0.374 to 0.471 mm, MAPE from 3.05% to 4.49%, and Pearson R from 0.84 to 0.91, indicating excellent consistency with manual annotations.

Conclusions: The proposed SZJ-SEG-based system enables accurate, efficient, and reproducible segmentation and quantitative analysis of cervical ultrasound images. With strong performance and high clinical relevance, it provides a reliable tool for ultrasound-guided nerve block localization and preoperative assessment. Its modular architecture offers scalability for extension to other anatomical regions, highlighting broad potential in intelligent ultrasound diagnosis.

Background: Time-dependent diffusion magnetic resonance imaging (MRI) has the potential to help characterize tumor cell properties; however, its application in imaging soft tissue tumors is still unclear. The objective of this study was to assess whether quantitative parameters obtained from time-dependent diffusion MRI (td-dMRI) correlate with the Ki-67 proliferation status in soft tissue tumors.

Methods: Participants with soft tissue tumors who underwent pretreatment td-dMRI between December 2023 and August 2024 were prospectively enrolled. The Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion (IMPULSED) model was utilized to fit td-dMRI signals. The microstructural parameters extracted included the intracellular volume fraction (v _in), cell diameter (d), cellularity, and extracellular diffusivity (D _ex). Apparent diffusion coefficient (ADC) values were calculated from diffusion-weighted imaging (DWI). Ki-67 was categorized into high (>20%) or low (≤20%) groups, and the area under the receiver operating characteristic (ROC) curve (AUC) was used to assess its effectiveness in differentiating between the high and low proliferation groups. These parameters were analyzed for their correlations with the Ki-67 index.

Results: The analysis included 31 participants with soft tissue tumors (mean age 52.1±19.9 years). Among all the parameters, the relative change in ADC (rcADC) of td-dMRI exhibited the highest AUC of 0.91, with 81.3% sensitivity and 86.7% specificity for differentiating low and high Ki-67 states. Except for parameter d, the AUC values of other parameters were greater than 0.7 (P<0.05). The rcADC and v _in values from td-dMRI exhibited strong positive correlations with the Ki-67 index (r_rcADC=0.68, r _{v in}=0.61; all P<0.001). Cellularity and change in ADC (cADC) showed moderate positive correlations with the Ki-67 index (r_celluarity=0.54, r_cADC=0.54; all P=0.002), whereas ADC_{0 Hz} and ADC_{25 Hz} exhibited moderate negative correlations (r_{ADC0 Hz}=-0.52, r_{ADC25 Hz}=-0.42; P=0.003, P=0.020). Weak negative correlations with the Ki-67 index were observed for D_ex and ADC_{40 Hz} (r _{D ex}=-0.37, r_{ADC40 Hz} =-0.37; P=0.038, P=0.040).

Conclusions: Several quantitative parameters derived from td-dMRI were correlated with Ki-67 proliferation status in soft tissue tumors. The rcADC was identified as the most robust predictor of Ki-67 status.

Background: The choroid plexus (CP) regulates cerebrospinal fluid (CSF) production, neuroimmune monitoring, and glymphatic transport, but its role in glymphatic system and cerebral small vessel disease (CSVD) remains unclear. This study aimed to investigate whether CP enlargement correlates with poor functional status in CSVD and whether glymphatic activity, measured by diffusion tensor imaging along perivascular space (DTI-ALPS), mediates this relationship.

Methods: A total of 70 CSVD participants and 44 age- and sex-matched healthy controls (HCs) underwent 5T ultra-high-field magnetic resonance imaging (MRI). The volumes of CP, lateral ventricles volume, white matter hyperintensities (WMHs), enlarged perivascular spaces (EPVS), and cerebral microbleeds (CMBs) were segmented automatically. Functional status was classified as fair or poor based on modified Rankin Scale (mRS) of <2 or ≥2. Group differences were assessed, and partial correlations, mediation analyses, and logistic regression were performed to quantify associations.

Results: Severe CSVD showed a lower global DTI-ALPS index than HCs (1.261 vs. 1.465, P<0.001) and mild to moderate CSVD (1.261 vs. 1.382, P=0.011). Those with poor functional status had larger CP (1.495×10^-3 vs. 1.154×10^-3 cm³, P=0.006), lateral ventricles volume (41.710 vs. 28.619 cm³, P<0.001), greater CMB burden (54.071×10^-3 vs. 9.075×10^-3 cm³, P=0.006), higher Fazekas score {4 [3-4] vs. 4 [2-4], P<0.001}, and lower DTI-ALPS index (1.181 vs. 1.392, P<0.001) compared to those with fair status. The DTI-ALPS index correlated with CP volume (r=-0.271, P=0.036), lateral ventricles volume (r=-0.400, P=0.001), basal ganglia (BG) perivascular space (PVS) volume (r=-0.262, P=0.042), total CSVD score (r=-0.287, P=0.025), and mRS (r=-0.316, P=0.013). Lower DTI-ALPS index partially mediated the relationship between larger CP volume and both larger lateral ventricle volume and higher mRS, with mediation effects accounting for 15.8% and 28.72%, respectively. Logistic regression identified older age [odds ratio (OR) =1.094, P=0.021], increased CP volume (OR =3.165, P=0.045), and lower DTI-ALPS index (OR =0.992, P=0.017) as independent predictors of poor functional status.

Conclusions: CP enlargement is associated with poor functional status in CSVD, and this relationship is partly explained by reduced DTI-ALPS index, suggesting a glymphatic pathway involvement. Quantitative MRI metrics of CP volume and DTI-ALPS index may serve as potential imaging biomarkers for functional status in CSVD.

Background: Hyperactivity of the hypothalamic-pituitary-adrenal (HPA) axis is an independent risk factor for cardiovascular disease (CVD). Adrenal gland volume reflects disturbances in glucose-lipid metabolism and HPA axis activity, but its relationship with CVD risk remains unclear. This study aimed to investigate the association between adrenal gland volume and estimated 10-year CVD risk in patients with diabetes mellitus (DM).

Methods: A total of 223 patients with DM (30-74 years) without prior CVD were retrospectively included. Adrenal gland volume was measured from non-contrast computed tomography (CT) images using a three-dimensional no new U-net (nnU)-Net deep learning algorithm. The 10-year CVD risk was estimated using the Framingham Risk Score (FRS) and categorized into low-to-intermediate risk (<20%) and high risk (≥20%) groups. Spearman correlation and multivariable linear regression were used to assess the associations of volume with metabolic parameters and the FRS, whereas multivariable binary logistic regression evaluated its association with high CVD risk. Receiver operating characteristic (ROC) curve analysis determined the predictive performance of volume for high CVD risk.

Results: In multivariable linear regression, total volume was positively associated with male sex (P=0.036), waist circumference (P=0.044), and hypertension (P=0.012), and negatively associated with high-density lipoprotein cholesterol (HDL-C; P=0.020). After adjustment for confounders, total volume remained significantly correlated with FRS [β=0.085, 95% confidence interval (CI): 0.003-0.168, P=0.042]. Volume was significantly larger in the high risk group compared to the low-to-intermediate risk group (P<0.001). Moreover, volume was independently associated with high CVD risk [left: odds ratio (OR) =1.777, 95% CI: 1.179-2.959; right: OR =2.053, 95% CI: 1.239-3.899; total: OR =1.467, 95% CI: 1.139-2.021]. The area under the curve (AUC) for total volume in predicting high CVD risk was 0.713 (95% CI: 0.643-0.783). At the optimal cut-off value of 9.04 cm3, the sensitivity and specificity were 70.3% and 61.3%, respectively.

Conclusions: Adrenal gland volume is independently associated with high estimated 10-year CVD risk in patients with diabetes and may potentially provide imaging-based support for the early identification of high-risk individuals.

Background: Arterial tortuosity is closely related to the formation, progression, and rupture of unruptured intracranial aneurysms (UIAs). This study aimed to assess the relationship between 3-dimensional (3D) centerline curvatures and the UIA formation risk, at the level of the entire cerebral arterial trunks and the parent arteries of UIAs in different hemodynamic locations.

Methods: This retrospective study evaluated the association between 3D centerline curvatures and the risk of UIAs in a cohort of 101 patients (114 aneurysms) and 145 healthy controls (HC) enrolled between March 2020 and February 2023. Tortuosity index (TI) and 3D centerline curvatures of 9 major cerebral arteries were automatically extracted from magnetic resonance angiography (MRA). Curvature parameters of cerebral arterial trunks were compared between UIA patients and HC and analyzed for their association with UIA formation. Additionally, curvature parameters of the parent arteries were compared with their contralateral counterparts in both sidewall and bifurcation aneurysms.

Results: Compared to HC, UIA patients exhibited significantly higher curvature parameters (P<0.001), whereas TI showed no significant difference. Increased curvature parameters are strongly correlated with a higher UIA risk [odds ratio (OR) ranging from 2.69 to 10.62]. In parent artery analysis, sidewall aneurysms demonstrated elevated curvature parameters versus contralateral arteries, including total absolute curvature (P=0.049), average absolute curvature (P=0.023), root-mean-square curvature (P =0.014), standard deviation of curvature (P=0.014), and 90% maximum absolute curvature (P=0.008), whereas bifurcation aneurysms showed no differences.

Conclusions: Increased 3D centerline curvatures of cerebral arteries significantly correlate with higher UIA risk, particularly in sidewall aneurysms. Assessing 3D centerline curvatures provides valuable insights into aneurysm hemodynamics and supports a comprehensive assessment of aneurysm risk.

Background: Prospective clinical research on the use of ultrasound contrast agents for tubal patency assessment in hysterosalpingo-contrast sonography (HyCoSy) is limited. This study sought to evaluate the efficacy and safety of perfluoropropane human albumin microspheres as an ultrasound contrast agent for HyCoSy in assessing fallopian tube patency, using laparoscopy (LSC) as the gold standard.

Methods: This prospective, double-blind, phase III study was conducted at 11 centers in China from April 2022 to April 2023 (ClinicalTrials.gov: CTR20220578). Infertile women aged between 18 and 45 years, who were scheduled for tubal patency testing, were eligible to participate in the study. The HyCoSy examination was performed by an experienced sonographer, and the image analysis was conducted by two experienced sonographers. Subsequently, within one menstrual cycle, LSC was performed by gynecologists who were blinded to the HyCoSy results. The accuracy of HyCoSy was calculated using the LSC results as the gold standard. In addition, a binary logistic regression model was applied to analyze the clinical and ultrasound factors associated with false-positive and false-negative results. Safety assessments were conducted before HyCoSy, within one week after HyCoSy, and before discharge from hospital following LSC.

Results: A total of 120 participants were enrolled in the study, of whom 112 completed both HyCoSy and LSC. Two participants who had previously undergone unilateral salpingectomy, and one participant whose fallopian tube evaluation result was recorded as "uncertain" on one side by HyCoSy, resulting in a total of 221 fallopian tubes included in the efficacy analysis. Compared to LSC, the overall accuracy of HyCoSy was 88.24% (195/221). The total false-positive rate and false-negative rate were 3.62% (8/221) and 8.14% (18/221), respectively. The false-positive group had a higher proportion of previous ectopic pregnancy and a higher proportion of ovaries adjacent to the uterus (P=0.03, P=0.01), while the false-negative group had a higher incidence of contrast intravasation (P=0.046). In the safety assessment, pain and contrast intravasation were common treatment-emergent adverse events.

Conclusions: Perfluoropropane human albumin microspheres can be used to effectively evaluate tubal patency through HyCoSy, and their diagnostic accuracy is good compared to the gold standard, LSC. Additionally, this contrast agent was associated with fewer and milder adverse reactions, making it suitable for clinical application.