Quantitative Imaging in Medicine and Surgery最新文献

Background: Anatomical variations of the recurrent motor branch (RMB) are at risk of injury during carpal tunnel release procedures. Previous studies have visualized the RMB using ultrasound (US) and magnetic resonance imaging (MRI) but have not compared the imaging capabilities of the two. Previous investigations have overlooked two specific types of carpal tunnel syndrome (CTS): simultaneous compression of the median nerve and the RMB and isolated compression of the latter. This study aims to identify the best imaging method to prevent iatrogenic injury to the RMB by comparing US and MRI capabilities. It also aims to devise a new method for the comprehensive diagnosis of CTS by evaluating the initial diameter of the RMB (RMB-ID). Additionally, this study aims to gain insights into the distribution patterns of the different anatomic variations of the RMB in healthy individuals and patients through an analysis of these variations. A cross-sectional study was conducted.

Methods: Forty healthy adults subjected to bilateral US and MRI of the RMB were included in this study. The US and magnetic resonance images of each patient were subsequently compared. US imaging of the RMB was performed on 102 hands of healthy adults and 112 hands of patients with CTS. The cross-sectional area of the median nerve (MN-CSA) and RMB-ID were measured.

Results: US provided better visibility of the RMB than did MRI (P<0.05). No statistically significant difference was observed in the variation type composition of the RMB between the healthy and patient groups (P>0.05). The RMB-ID and the MN-CSA significantly differed between groups (P<0.001). The RMB-ID increased with the increase of the MN-CSA (R=0.842; P<0.001). The optimal cutoff point for diagnosing CTS of the RMB-ID was 0.85 mm, yielding a sensitivity of 83.0%, a specificity of 92.2%, and the area under the curve of 0.945. The MN-CSA was 0.115 cm², with a sensitivity of 73.2%, a specificity of 96.1%, and an area under the curve of 0.923 [95% confidence interval (CI): 0.887-0.958]. No statistically significant difference was observed in the area under the receiver operator characteristic curve between the two diagnostic methods (P>0.05). The interexaminer reliability for the RMB-ID and the MN-CSA measurements was 0.983 (95% CI: 0.978-0.987) and 0.966 (95% CI: 0.955-0.974), respectively.

Conclusions: US outperformed MRI in visualizing the anatomical variations of the RMB. The RMB-ID was an accurate and valid indicator for comprehensive diagnosis of CTS.

Background: Pineal region tumors encompass a wide range of pathologies, presenting challenges in pre-surgical characterization and exhibiting variable prognostic outcomes across different tumor types. This study aims to identify key imaging features from pre-treatment magnetic resonance imaging (MRI) of pineal region tumors to aid in prognostic analysis.

Methods: We retrospectively enrolled 33 patients with pineal region tumors who underwent pre-treatment imaging at the Seventh Medical Center of the Chinese PLA General Hospital between January 2010 and June 2023. Two radiologists assessed imaging features including lesion morphology, border delineation, intensity, enhancement pattern, maximum tumor diameter, and secondary changes such as intratumoral hemorrhage and cystic changes. Tumor prognoses were categorized based on reported overall survival rates from recent literature as either good (overall survival rate ≥90%) or poor (overall survival rate <90%). Significant imaging features were selected using one-way analysis of variance (ANOVA) and binary logistic regression.

Results: The study cohort comprised 33 patients (27 males), aged 1 to 72 years [mean ± standard deviation (SD), 26.4±17.7 years]. We identified 13 distinct pathology types, with 15 cases classified as having a good prognosis and 18 cases as having a poor prognosis. A significant feature identified through one-way ANOVA was intratumoral hemorrhage (P=0.017). Binary logistic regression also highlighted intratumoral hemorrhage as a significant independent predictor of prognosis, irrespective of age and other factors. Tumors with intratumoral hemorrhage had a nearly 19-fold (P=0.015, 95% CI: 1.780-202.798) higher likelihood of an unfavorable prognosis compared to those without.

Conclusions: Intratumoral hemorrhage emerges as a significant indicator of poor prognosis in pineal region tumors. This finding underscores the importance of incorporating specific imaging features, particularly intratumoral hemorrhage, into the prognostic evaluation of pineal region tumors.

Background: Pediatric growth hormone deficiency (GHD) is a disease resulting from the impaired growth hormone-insulin-like growth factor-1 (GH-IGF-1) axis, but the effects of GHD on children's behavior and brain microstructural structure alterations have not yet been fully clarified. We aimed to investigate the quantitative profiles of gray matter and white matter in pediatric GHD using synthetic magnetic resonance imaging (MRI).

Methods: The data of 50 children with GHD and 50 typically developing (TD) children were prospectively collected. Group differences in brain volumetric parameters, individual-level T1 and T2 relaxometry values, and myelin volume fraction (MVF) were assessed. Subsequently, magnetic resonance-based indices with significant differences between groups were correlated with clinical variables via partial correlation.

Results: Compared with TD children, children with GHD showed significantly decreased whole-brain gray-matter volume [P false discovery rate (P_FDR) <0.001] and increased non-gray-matter/white-matter/cerebrospinal fluid (NoN) volume (P_FDR<0.001). For gray-matter microstructural profiles, altered T1 and T2 relaxometry values in children with GHD were mainly distributed in the default mode (P_FDR<0.001) and central executive networks (P_FDR<0.001). For white-matter microstructural profiles, widespread increased regional MVF was mainly distributed in the corpus callosum, corticospinal tract, internal capsule, external capsule, and cingulum (all P_FDR values <0.001). Meanwhile, the T2 relaxation values in the left cuneus (r=0.400; P=0.005) and MVF in the right corticospinal tract (r=0.313; P=0.032) had a positive relationship with IGF-1.

Conclusions: Altered T1 and T2 relaxometry values and MVF in gray and white matter indicate the relevance of the default mode, central executive, somatosensory, visual, and cerebellar networks underlying pediatric GHD, which may imply the involvement of the GH-IGF-1 axis and myelin in the pathophysiological mechanism of GHD. Moreover, the brain microstructure alteration in cortico-striatal-limbic loop might be influenced by the GH-IGF-1 axis and play an important role in the behavioral impairments in children with GHD.

Background: Echocardiography can conveniently, rapidly, and economically evaluate the structure and function of the heart, and has important value in the diagnosis and evaluation of cardiovascular diseases (CVDs). However, echocardiography still exhibits significant variability in image acquisition and diagnosis, with a heavy dependency on the operator's experience. Image quality affects disease diagnosis in the later stage, and even image quality assessment still has variability in human evaluation. This study aimed to develop an automated and real-time quality assessment system using deep learning (DL) techniques while decreasing the measurement error of left ventricular ejection fraction (LVEF).

Methods: This study involved over 5,000 echocardiography datasets from 2,461 participants across 10 medical centers in China to build the model. A 5-point quality scoring system was used to assess the integrity, clarity, and alignment of anatomical structures in each echocardiogram view. Additionally, an innovative DL model was developed to autonomously detect these essential cardiac anatomical structures in real-time, subsequently providing quality score estimations and LVEF. A total of 175 participants from two distinct external medical centers were enrolled for model validation. This dataset was employed to assess the consistency and repeatability of quality score and ejection fraction (EF) measurements, and the assessments made by human experts were compared with those of our model.

Results: The developed model demonstrated exceptional performance, achieving Intersection over Union (IoU) scores exceeding 0.8 for left ventricular (LV) segmentation, a mean average precision when IoU >0.5 (mAP50) of 0.91 for cardiac anatomical structures detection, and a 0.96±0.05 accuracy in view classification. The quality scores assessed by the model closely matched those of human experts, indicating strong agreement. The weighted average precision and weighted average recall scores fell within the range of 0.5 to 0.6. Notably, there was no statistically significant difference in LVEF assessments between human experts and our model (P=0.09), as demonstrated by an intraclass correlation coefficient (ICC) analysis of 0.821, reflecting high-level consistency. When assessing echocardiograms with high-quality scores, the model demonstrated a significantly closer alignment and a higher correlation coefficient with human experts (R=0.90±0.04).

Conclusions: This study demonstrates that artificial intelligence-assisted echocardiography scoring system aligns well with manual quality scoring. Through the supervision of real-time echocardiogram quality, the artificial intelligence model can assist doctors in providing more reproducible and consistent assessments of cardiac function.

Background: The presence of hepatic steatosis (HS) is a crucial histological parameter for evaluating the suitability of liver transplantation. However, to date, no studies have used contrast-enhanced ultrasonography (CEUS) to diagnose and grade HS in brain-dead donors. This study aimed to detect and quantify hepatic microcirculatory perfusion in brain-dead donors using CEUS and to assess the utility of CEUS in the diagnosis and grading of HS.

Methods: This prospective study enrolled 88 livers from brain-dead donors (44 with HS and 44 without HS) aged ≥18 years between June 2020 and January 2024. The donors had a mean age of 45.42±9.59 years, and 70 were male (79.5%). CEUS was conducted on the livers of the brain-dead donors 24 h before organ procurement, and time-intensity curves were generated. The main measures included the arrival time, time-to-peak, peak intensity of the hepatic artery (PIHA), peak intensity of the portal vein (PIPV), and peak intensity of the liver parenchyma (PILP), and hepatorenal index (HRI). Logistic regression analyses were used to identify the significant factors associated with HS, and the areas under the curve (AUC) of the receiver operating characteristic curves were used to evaluate diagnostic performance.

Results: The PIHA (P<0.001), PIPV (P<0.001), and PILP (P=0.001) were significantly shorter in the steatosis group than the non-steatosis group. The one-way analysis of variance revealed significant decreases in the PIHA (P<0.001), PIPV (P<0.001), and PILP (P<0.001) as HS grades increased. The multivariate regression analysis indicated that the PIHA was an independent factor for both HS [odds ratio (OR) =0.97, 95% confidence interval (CI): 0.94-0.99, P=0.01] and moderate-to-severe HS (OR =0.96, 95% CI: 0.93-0.99, P=0.009). The AUC values of the PIHA and HRI for diagnosing moderate-to-severe HS were 0.88 and 0.69, respectively. The optimal cut-off value of the PIHA for diagnosing moderate-to-severe HS was 173.04, with a sensitivity of 92.9% (13 of 14 livers), specificity of 68.9% (51 of 74 livers), and positive and negative predictive values of 36.1% and 98.1%, respectively.

Conclusions: CEUS showed promising results in the diagnosis and grading of HS in brain-dead donors. The PIHA, a CEUS-derived parameter, could serve as a diagnostic tool for moderate-to-severe HS.

Background: There are currently no deep learning models applying resting-state functional magnetic resonance imaging (rs-fMRI) data to distinguish patients with Parkinson's disease (PD) and healthy controls (HCs). Moreover, no study has correlated objective gait parameters with brain network alterations in patients with PD. We propose BrainNetCNN + CL, applying a convolutional neural network (CNN) and joint contrastive learning (CL) method to brain network analysis to classify patients with PD and HCs, and compare their performance with classical classification methods. This study aimed to explore more accurate abnormal connecting regions that may serve as potential therapeutic targets, and to explore the correlation between abnormal connecting regions and gait parameters.

Methods: We enrolled 29 patients with PD and 38 HCs. Rs-fMRI data and high-resolution three-dimensional structural T1-weighted images were acquired for each participant. BrainNetCNN + CL were utilized to classify the PD and HC groups.

Results: The top 20 connections with the highest contribution to the classification results obtained using BrainNetCNN + CL included the default mode network (DMN), ventral attention network (VAN), and limbic network (LN). The strength of the functional connectivity (FC) between the right inferior occipital gyrus and left postcentral gyrus in the PD group was negatively correlated with the step length at the self-selected pace (SSP) speed in the "ON" state (P=0.001, r=-0.589). The strength of the FC between the right fusiform gyrus and the right calcarine fissure and surrounding cortex was negatively correlated with the Beck Anxiety Inventory (BAI) score (P=0.032, r=-0.406) and positively correlated with the Berg Balance Scale (BBS) score measured in the "ON" state (P=0.037, r=0.395).

Conclusions: BrainNetCNN + CL accurately identified abnormally connected regions associated with gait impairments, which may serve as potential therapeutic targets for PD.

Background: Approximately half of human immunodeficiency virus (HIV) patients experience HIV-associated neurocognitive disorders (HAND); however, the neurophysiological mechanisms underlying HAND remain unclear. This study aimed to evaluate changes in functional brain activity patterns during the early stages of HIV infection by comparing local and global indicators using resting-state functional magnetic resonance imaging (rs-fMRI).

Methods: A total of 165 people living with HIV (PLWH) but without neurocognitive disorders (PWND), 173 patients with asymptomatic neurocognitive impairment (ANI), and 100 matched healthy controls (HCs) were included in the study. A cross-sectional study of the participants was conducted. The metrics of functional segregation and integration were computed, using graph theory to explore differences across methodologies. Brain functional changes in the PWND and ANI groups were assessed, and correlations between the rs-fMRI metrics, clinical data, and cognitive function were examined.

Results: As cognitive function declined, changes reflected by regional homogeneity (ReHo) were primarily observed in the default mode network (DMN). In the DMN and visual network (VIS), amplitude of low-frequency fluctuation (ALFF) decreases were mainly observed in the parieto-occipital lobes, while increases were mainly observed in the limbic network (LIM). Reductions in fractional ALFF (fALFF) were mainly observed in the somatomotor network (SMN) and LIM, while increases were observed in the DMN and LIM. Unlike local indicators, global functional connectivity (FC) significantly decreased in both the PWND and ANI groups compared to the HC group. The ANI group showed partial increases in FC compared to the PWND group, with major changes observed in the DMN, VIS, and LIM. Notably, FC between the right insula and right supramarginal gyrus decreased significantly following HIV infection, while FC between the right caudate nucleus and the left middle frontal gyrus declined further in the ANI group. Graph theory further confirmed the significance of the DMN, and revealed changes in the eigenvector centrality mapping (ECM) values of the frontoparietal network (FPN) and dorsal attention network (DAN).

Conclusions: HIV patients exhibit complex changes in both local and global brain activity, regardless of cognitive impairment. Widespread abnormalities primarily involve the DMN, VIS, and LIM. Changes in FC along the fronto-striatal pathway may play a crucial role in the decline of cognitive function in individuals with HAND. Our findings provide new insights that may assist in the early detection of brain damage in the early stages of HIV infection. The use of multiple methodologies may offer a more comprehensive and effective approach, enabling the early detection of brain damage in HIV patients.

Background: Missed or delayed diagnosis of occult fractures of tibial plateau may cause adverse effects on patients. The objective of this study was to evaluate the diagnostic performance of texture analysis (TA) of knee joint radiographs combined with machine learning (ML) in identifying patients at risk of tibial plateau occult fractures.

Methods: A total of 169 patients with negative fracture on knee X-ray films from 2018 to 2022 who were diagnosed with occult tibial plateau fractures or no fractures by subsequent magnetic resonance imaging (MRI) examination were retrospectively enrolled. The X-ray images of the patient's knee joint were used for texture feature extraction. A total of 9 ML feature selection methods (including 6 mainstream methods and 3 methods provided by MaZda software) combined with 3 classification methods were used to build the best diagnostic model. The performance of each model was evaluated by accuracy, F1-value, and area under the curve (AUC).

Results: The least absolute shrinkage and selection operator (LASSO) method had the best performance of the 6 mainstream methods, with an accuracy of 0.81, an F1 value of 0.80, and an AUC of 0.920, all of which were higher than those of the other five methods (accuracy range: 0.65-0.80, F1 score range: 0.61-0.79, AUC range: 0.722-0.895). Among the three feature selection models in MaZda software, the most ideal method for accuracy measurement was the MI method, reaching 0.77. In the measurement of the F1 value and AUC, MaZda's best method was Fisher, reaching 0.78 and 0.888, respectively. All indicators were lower than those of the LASSO method. The combination of LASSO and support vector machine (SVM) yielded the best classification performance, while the performance of the combination of LASSO and logistic regression was slightly inferior, but the difference was not statistically significant.

Conclusions: TA of knee joint radiography combined with ML has achieved high performance in identifying patients at risk of occult fractures of the tibial plateau. Considering both the model performance and computational complexity, the LASSO feature selection method combined with the logistic regression classifier yielded the best classification performance in this process.