RESUMEN
Magneto-acousto-electrical tomography (MAET) is a hybrid imaging method that combines the high spatial resolution of ultrasonography with the high contrast of electrical impedance tomography (EIT). While most previous studies on MAET have focused on two-dimensional imaging, our recent research proposed a novel three-dimensional (3D) MAET method using B-mode and translational scanning. This method has been the first to reconstruct a 3D volume image of conductivity interfaces. However, this method has its limitations in mapping irregular shapes of conductivity. To address this challenge, we propose a 3D magneto-acousto-electrical computed tomography (3D MAE-CT) method utilizing an ultrasound linear array transducer in this work. Both phantom and in vitro experiments were conducted to validate our proposed method. The results from the phantom experiments demonstrate that our method can map the 3D volume conductivity with high spatial resolution. The oblique angles extracted from the 3D image closely match practical value, with the relative error ranging between -2.80% and 4.07%. Furthermore, the in vitro experiment successfully obtained a 3D image of a chicken heart, marking the first MAET 3D conductivity image of a tissue sample to date.
RESUMEN
OBJECTIVES: Freehand three-dimensional (3D) ultrasound (US) is of great significance for clinical diagnosis and treatment, it is often achieved with the aid of external devices (optical and/or electromagnetic, etc.) that monitor the location and orientation of the US probe. However, this external monitoring is often impacted by imaging environment such as optical occlusions and/or electromagnetic (EM) interference. METHODS: To address the above issues, we integrated a binocular camera and an inertial measurement unit (IMU) on a US probe. Subsequently, we built a tight coupling model utilizing the unscented Kalman algorithm based on Lie groups (UKF-LG), combining vision and inertial information to infer the probe's movement, through which the position and orientation of the US image frame are calculated. Finally, the volume data was reconstructed with the voxel-based hole-filling method. RESULTS: The experiments including calibration experiments, tracking performance evaluation, phantom scans, and real scenarios scans have been conducted. The results show that the proposed system achieved the accumulated frame position error of 3.78 mm and the orientation error of 0.36° and reconstructed 3D US images with high quality in both phantom and real scenarios. CONCLUSIONS: The proposed method has been demonstrated to enhance the robustness and effectiveness of freehand 3D US. Follow-up research will focus on improving the accuracy and stability of multi-sensor fusion to make the system more practical in clinical environments.
Asunto(s)
Algoritmos , Imagenología Tridimensional , Fantasmas de Imagen , Ultrasonografía , Imagenología Tridimensional/métodos , Ultrasonografía/métodos , Ultrasonografía/instrumentación , Diseño de Equipo , HumanosRESUMEN
In pursuit of cultivating automated models for magnetic resonance imaging (MRI) to aid in diagnostics, an escalating demand for extensive, multisite, and heterogeneous brain imaging datasets has emerged. This potentially introduces biased outcomes when directly applied for subsequent analysis. Researchers have endeavored to address this issue by pursuing the harmonization of MRIs. However, most existing image-based harmonization methods for MRI are tailored for 2D slices, which may introduce inter-slice variations when they are combined into a 3D volume. In this study, we aim to resolve inconsistencies between slices by introducing a pseudo-warping field. This field is created randomly and utilized to transform a slice into an artificially warped subsequent slice. The objective of this pseudo-warping field is to ensure that generators can consistently harmonize adjacent slices to another domain, without being affected by the varying content present in different slices. Furthermore, we construct unsupervised spatial and recycle loss to enhance the spatial accuracy and slice-wise consistency across the 3D images. The results demonstrate that our model effectively mitigates inter-slice variations and successfully preserves the anatomical details of the images during the harmonization process. Compared to generative harmonization models that employ 3D operators, our model exhibits greater computational efficiency and flexibility.
Asunto(s)
Encéfalo , Imagenología Tridimensional , Imagen por Resonancia Magnética , Imagen por Resonancia Magnética/métodos , Imagen por Resonancia Magnética/normas , Humanos , Imagenología Tridimensional/métodos , Encéfalo/diagnóstico por imagen , Algoritmos , Neuroimagen/métodos , Neuroimagen/normasRESUMEN
BACKGROUND: Deep learning, as an artificial intelligence method has been proved to be powerful in analyzing images. The purpose of this study is to construct a deep learning-based model (ToothNet) for the simultaneous detection of dental caries and fissure sealants in intraoral photos. METHODS: A total of 1020 intraoral photos were collected from 762 volunteers. Teeth, caries and sealants were annotated by two endodontists using the LabelMe tool. ToothNet was developed by modifying the YOLOX framework for simultaneous detection of caries and fissure sealants. The area under curve (AUC) in the receiver operating characteristic curve (ROC) and free-response ROC (FROC) curves were used to evaluate model performance in the following aspects: (i) classification accuracy of detecting dental caries and fissure sealants from a photograph (image-level); and (ii) localization accuracy of the locations of predicted dental caries and fissure sealants (tooth-level). The performance of ToothNet and dentist with 1year of experience (1-year dentist) were compared at tooth-level and image-level using Wilcoxon test and DeLong test. RESULTS: At the image level, ToothNet achieved an AUC of 0.925 (95% CI, 0.880-0.958) for caries detection and 0.902 (95% CI, 0.853-0.940) for sealant detection. At the tooth level, with a confidence threshold of 0.5, the sensitivity, precision, and F1-score for caries detection were 0.807, 0.814, and 0.810, respectively. For fissure sealant detection, the values were 0.714, 0.750, and 0.731. Compared with ToothNet, the 1-year dentist had a lower F1 value (0.599, p < 0.0001) and AUC (0.749, p < 0.0001) in caries detection, and a lower F1 value (0.727, p = 0.023) and similar AUC (0.829, p = 0.154) in sealant detection. CONCLUSIONS: The proposed deep learning model achieved multi-task simultaneous detection in intraoral photos and showed good performance in the detection of dental caries and fissure sealants. Compared with 1-year dentist, the model has advantages in caries detection and is equivalent in fissure sealants detection.
Asunto(s)
Aprendizaje Profundo , Caries Dental , Selladores de Fosas y Fisuras , Humanos , Caries Dental/diagnóstico , Selladores de Fosas y Fisuras/uso terapéutico , Proyectos Piloto , Fotografía Dental/métodos , Adulto , Masculino , FemeninoRESUMEN
DNA damage response (DDR) plays a vital role in the development of cancer. Nevertheless, in osteosarcoma, the potential of DDR-related genes (DDRGs) remains unclear. Thus, the current research is intended to investigate the mechanisms of DDRGs in the development of osteosarcoma and to explore potential DDR-related biomarkers in forecasting the prognosis of osteosarcoma patients. The osteosarcoma genomic data from TCGA, GEO and cBioPortal databases were utilized for screening and identification of differentially expressed DDRGs (DEDDRGs). Consensus clustering analysis was performed to identify different subtypes of osteosarcoma based on the expressions of DDRGs. Key DEDRRGs were identified by overlapping DEDRRGs between different subtypes and DEDRRGs between tumor and control samples. Univariate, as well as LASSO regressions, were further applied to obtain robust prognostic signatures. GSVA and ssGSEA analysis were implemented to explore the underlying mechanisms of prognostic DDRG signature in regulating osteosarcoma. In addition, the drug sensitivity of patients in low- and high-risk groups was evaluated using pRRophetic algorithm. A total of 43 key DEDRRGs were identified. Followed by univariate Cox along with LASSO regression analyses, CDK6, CSF1R, EGFR, ERBB4, GATA3 and SOCS1 were identified as prognostic signatures in osteosarcoma. Cox regressions revealed that the risk score was an independent prognostic factor in osteosarcoma. DDR may affect osteosarcoma via regulating immune microenvironment along with influencing cell proliferation, migration, adhesion and apoptosis. The chemotherapeutic response between patients in low- and high-risk groups was much different. The role of DDRGs in osteosarcoma and identified six DDR-linked biomarkers for forecasting the prognosis of osteosarcoma patients. Our outcomes enhanced the understanding of DDR-related molecular mechanisms involved in osteosarcoma and provided potential therapeutic targets for osteosarcoma patients.
Asunto(s)
Neoplasias Óseas , Daño del ADN , Regulación Neoplásica de la Expresión Génica , Osteosarcoma , Osteosarcoma/genética , Osteosarcoma/patología , Humanos , Pronóstico , Daño del ADN/genética , Neoplasias Óseas/genética , Neoplasias Óseas/patología , Neoplasias Óseas/mortalidad , Biomarcadores de Tumor/genética , Biomarcadores de Tumor/metabolismo , Perfilación de la Expresión Génica , Femenino , Reparación del ADN/genéticaRESUMEN
This paper presents a cost-effective, quantitative, point-of-care solution for urinalysis screening, specifically targeting nitrite, protein, creatinine, and pH in urine samples. Detecting nitrite is crucial for the early identification of urinary tract infections (UTIs), while regularly measuring urinary protein-to-creatinine (UPC) ratios aids in managing kidney health. To address these needs, we developed a portable, transmission-based colorimeter using readily available components, controllable via a smartphone application through Bluetooth. Multiple colorimetric detection strategies for each analyte were identified and tested for sensitivity, specificity, and stability in a salt buffer, artificial urine, and human urine. The colorimeter successfully detected all analytes within their clinically relevant ranges: nitrite (6.25-200 µM), protein (2-1024 mg/dL), creatinine (2-1024 mg/dL), and pH (5.0-8.0). The introduction of quantitative protein and creatinine detection, and a calculated urinary protein-to-creatinine (UPC) ratio at the point-of-care, represents a significant advancement, allowing patients with proteinuria to monitor their condition without frequent lab visits. Furthermore, the colorimeter provides versatile data storage options, facilitating local storage on mobile devices or in the cloud. The paper further details the setup of the colorimeter's secure connection to a cloud-based environment, and the visualization of time-series analyte measurements in a web-based dashboard.
Asunto(s)
Nitritos , Urinálisis , Humanos , Creatinina/orina , Proteinuria/diagnóstico , Proteinuria/orina , Concentración de Iones de HidrógenoRESUMEN
Background: Osteoarthritis (OA) is a diverse disorder that most frequently affects elderly people and makes them disabled. Many investigations have shown that the etiology of OA depends on cartilage wear, but immunology also plays a significant role. Thus, the goal of this study was to define the immune-related etiology of OA. Methods: Data from the "Gene Expression Omnibus (GEO)" database were used to find differentially expressed genes (DEGs), and the "Cell-type Identification By Estimating Relative Subsets Of RNA Transcripts (CIBERSORT) algorithm" was employed to calculate the quantity of distinct immune cells. We analyzed the results to identify patient subgroups and compare major active pathways. Results: The macrophage cell population accounts for the greatest percentage of infiltrating immune cells in OA. One hundred and twenty-two common intersection genes were identified, with the network analysis of protein-protein interactions revealing ten hub genes related to OA, including CXCL8, JUN, ATF3, DUSP1, PTGS2, IL6, MMP9, FOS, NFKBIA, and MYC. The random forest model showed that memory-activated CD4 T cells are strongly correlated with other immune cell types, while neutrophils have the weakest correlation with other immune cell types. Violin plots showed that OA patients had a significantly larger quantity of plasma cells and resting mast cells, with a significantly smaller quantity of resting memory CD4 T cells and activated mast cells than healthy controls. Conclusions: Two immune-related subgroups of OA were identified by semi-supervised clustering analysis of microarray data, and core genes were also determined by network analysis. A group of the immune infiltrating cells was selected by random forest analysis suggesting they are related to the pathogenesis of OA.
RESUMEN
OBJECTIVES: Dramatic brain morphological changes occur throughout the third trimester of gestation. In this study, we investigated whether the predicted brain age (PBA) derived from graph convolutional network (GCN) that accounts for cortical morphometrics in third trimester is associated with postnatal abnormalities and neurodevelopmental outcome. METHODS: In total, 577 T1 MRI scans of preterm neonates from two different datasets were analyzed; the NEOCIVET pipeline generated cortical surfaces and morphological features, which were then fed to the GCN to predict brain age. The brain age index (BAI; PBA minus chronological age) was used to determine the relationships among preterm birth (i.e., birthweight and birth age), perinatal brain injuries, postnatal events/clinical conditions, BAI at postnatal scan, and neurodevelopmental scores at 30 months. RESULTS: Brain morphology and GCN-based age prediction of preterm neonates without brain lesions (mean absolute error [MAE]: 0.96 weeks) outperformed conventional machine learning methods using no topological information. Structural equation models (SEM) showed that BAI mediated the influence of preterm birth and postnatal clinical factors, but not perinatal brain injuries, on neurodevelopmental outcome at 30 months of age. CONCLUSIONS: Brain morphology may be clinically meaningful in measuring brain age, as it relates to postnatal factors, and predicting neurodevelopmental outcome. CLINICAL RELEVANCE STATEMENT: Understanding the neurodevelopmental trajectory of preterm neonates through the prediction of brain age using a graph convolutional neural network may allow for earlier detection of potential developmental abnormalities and improved interventions, consequently enhancing the prognosis and quality of life in this vulnerable population. KEY POINTS: â¢Brain age in preterm neonates predicted using a graph convolutional network with brain morphological changes mediates the pre-scan risk factors and post-scan neurodevelopmental outcomes. â¢Predicted brain age oriented from conventional deep learning approaches, which indicates the neurodevelopmental status in neonates, shows a lack of sensitivity to perinatal risk factors and predicting neurodevelopmental outcomes. â¢The new brain age index based on brain morphology and graph convolutional network enhances the accuracy and clinical interpretation of predicted brain age for neonates.
RESUMEN
PURPOSE: The purpose of this paper is to evaluate the clinical and radiographic outcomes of oblique lumbar interbody fusion (OLIF) to perform in L4/5 degenerative lumbar spondylolisthesis (DLS) patients who diagnosed with osteopenia. METHODS: From December 2018 to 2021 March, 94 patients were diagnosed with degenerative spondylolisthesis underwent OLIF and divided into two groups with different bone mineral density. Anterolateral screw and rod instrumentation was applied in two groups. The primary outcomes were VAS, JOA and ODI. The secondary outcomes included disc height (DH), cross-sectional height of the intervertebral foramina (CSH), cross-sectional area of the dural sac (CSA), lumbar lordorsis (LL), pelvic titlt (PT), pelvic incidence (PI) and sacrum slop (SS). RESULTS: All patients finished at least 1 years follow-up with 21.05 ± 4.42 months in the group A and 21.09 ± 4.28 months in the group B. The clinical symptoms were evaluated by VAS, JOA and ODI and 94 patients showed good outcomes at final follow-up (P < 0.05), with significant increases in DH, CSH and CSA. In group A, DH increased from 8.54 ± 2.48 to 11.11 ± 2.63 mm, while increased from 8.60 ± 2.29 to 11.23 ± 1.88 were recorded in group B. No statistical difference was found in DH between the two groups (P > 0.05). The cage subsidence was 1.14 ± 0.83 mm in group A and 0.87 ± 1.05 mm in group B (P > 0.05). There was no significant difference in the adjusted parameters of spino-pelvic between two groups (P > 0.05). CONCLUSION: Oblique lumbar interbody fusion with anterolateral screw and rod instrumentation is feasible to be performed in osteopenia patients who diagnosed with degenerative spondylolisthesis.
Asunto(s)
Enfermedades Óseas Metabólicas , Espondilolistesis , Humanos , Estudios Retrospectivos , Espondilolistesis/complicaciones , Espondilolistesis/diagnóstico por imagen , Espondilolistesis/cirugía , Enfermedades Óseas Metabólicas/diagnóstico por imagen , Enfermedades Óseas Metabólicas/cirugía , Tornillos Óseos , Sacro/diagnóstico por imagen , Sacro/cirugíaRESUMEN
BACKGROUND: To explore the role of Trans-rectal Color Doppler Flow Imaging (TR-CDFI) and risk-stratification nomogram in a MRI-directed biopsy pathway and examine its clinical performance, via comparisons between existing four biopsy pathways. METHODS: A Bi-centered retrospective cohort study on biopsy-naïve male population who received ultrasound-guided prostate biopsy from Jan. 2015 to Feb. 2022 was proposed. All enrolled patients should have undergone serum-PSA test, TR-CDFI and multiparametric MRI before biopsy, and subsequently opted for surgical intervention, enabling more accurate pathological grading. We then utilized univariate and multivariate logistic regression analysis to construct a predictive nomogram for risk-stratification. Outcome measurements were overall prostate cancer (PCA) detection rate, clinically significant PCA (csPCA) detection rate, clinically insignificant PCA (cisPCA) detection rate, biopsy avoidance rate and missed csPCA detection rate. Decision curve analysis was used to compare the performances between diagnostic pathways. RESULTS: Under the criteria mentioned above, 752 patients from two centers were included. Reference pathway (biopsy for all) showed that overall PCA detection rate was 46.1%, csPCA and cisPCA detection rates were 32.3% and 13.8% respectively. Risk-based MRI-directed TR-CDFI pathway, which incorporated both TR-CDFI and risk stratification nomogram, exhibited PCA detection rate of 38.7%, csPCA detection rate of 28.7%, cisPCA detection rate of 7.0%, Biopsy avoidance rate of 42.4%, and missed csPCA detection rate of 3.6%. Decision curve analysis revealed that the risk-based pathway held the most net benefit, under the threshold probability level between 0.1 and 0.5. CONCLUSIONS: The risk-based MRI-directed TR-CDFI pathway out-performed other strategies, balancing csPCA detection and biopsy avoidance. This suggested that incorporation of TR-CDFI and risk-stratification nomogram in the early PCA diagnostic procedures could reduce unnecessary biopsies.
Asunto(s)
Próstata , Neoplasias de la Próstata , Humanos , Masculino , Próstata/diagnóstico por imagen , Próstata/patología , Nomogramas , Estudios Retrospectivos , Biopsia , Neoplasias de la Próstata/diagnóstico por imagen , Neoplasias de la Próstata/patología , Imagen por Resonancia Magnética/métodos , Biopsia Guiada por Imagen/métodosRESUMEN
The resting-state functional magnetic resonance imaging (rs-fMRI) faithfully reflects the brain activities and thus provides a promising tool for autism spectrum disorder (ASD) classification. Up to now, graph convolutional networks (GCNs) have been successfully applied in rs-fMRI based ASD classification. However, most of these methods were developed based on functional connectivities (FCs) that only reflect low-level correlation between brain regions, without integrating both high-level discriminative knowledge and phenotypic information into classification. Besides, they suffered from the overfitting problem caused by insufficient training samples. To this end, we propose a novel contrastive multi-view composite GCN (CMV-CGCN) for ASD classification using both FCs and HOFCs. Specifically, a pair of graphs are constructed based on the FC and HOFC features of the subjects, respectively, and they share the phenotypic information in the graph edges. A novel contrastive multi-view learning method is proposed based on the consistent representation of both views. A contribution learning mechanism is further incorporated, encouraging the FC and HOFC features of different subjects to have various contribution in the contrastive multi-view learning. The proposed CMV-CGCN is evaluated on 613 subjects (including 286 ASD patients and 327 NCs) from the Autism Brain Imaging Data Exchange (ABIDE). We demonstrate the performance of the method for ASD classification, which yields an accuracy of 75.20% and an area under the curve (AUC) of 0.7338. Experimental results show that our proposed method outperforms state-of-the-art methods on the ABIDE database.
Asunto(s)
Trastorno del Espectro Autista , Infecciones por Citomegalovirus , Humanos , Trastorno del Espectro Autista/diagnóstico por imagen , Imagen por Resonancia Magnética/métodos , Encéfalo/diagnóstico por imagen , Mapeo Encefálico/métodosRESUMEN
Ultrafast ultrasound imaging modalities have been studied extensively in the ultrasound community. It breaks the compromise between the frame rate and the region of interest by imaging the whole medium with wide unfocused waves. Continuously available data allow monitoring fast transient dynamics at hundreds to thousands of frames per second. This feature enables a more accurate and robust velocity estimation in vector flow imaging (VFI). On the other hand, the huge amount of data and real-time processing demands are still challenging in VFI. A solution is to provide a more efficient beamforming approach with smaller computation complexity than the conventional time-domain beamformer like delay-and-sum (DAS). Fourier-domain beamformers are shown to be more computationally efficient and can provide equally good image quality as DAS. However, previous studies generally focus on B-mode imaging. In this study, we propose a new framework for VFI which is based on two advanced Fourier migration methods, namely, slant stack migration (SSM) and ultrasound Fourier slice beamform (UFSB). By carefully modifying the beamforming parameters, we successfully apply the cross-beam technique within the Fourier beamformers. The proposed Fourier-based VFI is validated in simulation studies, in vitro, and in vivo experiments. The velocity estimation is evaluated via bias and standard deviation and the results are compared with conventional time-domain VFI using the DAS beamformer. In the simulation, the bias is 6.4%, -6.2%, and 5.7%, and the standard deviation is 4.3%, 2.4%, and 3.9% for DAS, UFSB, and SSM, respectively. In vitro studies reveal a bias of 4.5%, -5.3%, and 4.3% and a standard deviation of 3.5%, 1.3%, and 1.6% from DAS, UFSB, and SSM, respectively. The in vivo imaging of the basilic vein and femoral bifurcation also generate similar results using all three methods. With the proposed Fourier beamformers, the computation time can be shortened by up to 9 times and 14 times using UFSB and SSM.
Asunto(s)
Algoritmos , Procesamiento de Imagen Asistido por Computador , Ultrasonografía/métodos , Fantasmas de Imagen , Simulación por Computador , Procesamiento de Imagen Asistido por Computador/métodosRESUMEN
Autism spectrum disorder (ASD) is characterized by poor social communication abilities and repetitive behaviors or restrictive interests, which has brought a heavy burden to families and society. In many attempts to understand ASD neurobiology, resting-state functional magnetic resonance imaging (rs-fMRI) has been an effective tool. However, current ASD diagnosis methods based on rs-fMRI have two major defects. First, the instability of rs-fMRI leads to functional connectivity (FC) uncertainty, affecting the performance of ASD diagnosis. Second, many FCs are involved in brain activity, making it difficult to determine effective features in ASD classification. In this study, we propose an interpretable ASD classifier DeepTSK, which combines a multi-output Takagi-Sugeno-Kang (MO-TSK) fuzzy inference system (FIS) for composite feature learning and a deep belief network (DBN) for ASD classification in a unified network. To avoid the suboptimal solution of DeepTSK, a joint optimization procedure is employed to simultaneously learn the parameters of MO-TSK and DBN. The proposed DeepTSK was evaluated on datasets collected from three sites of the Autism Brain Imaging Data Exchange (ABIDE) database. The experimental results showed the effectiveness of the proposed method, and the discriminant FCs are presented by analyzing the consequent parameters of Deep MO-TSK.
Asunto(s)
Trastorno del Espectro Autista , Trastorno Autístico , Humanos , Trastorno del Espectro Autista/diagnóstico por imagen , Encéfalo/diagnóstico por imagen , Mapeo Encefálico/métodos , Imagen por Resonancia Magnética/métodosRESUMEN
This paper reports on a low-cost, quantitative, point-of-care solution for the early detection of nitrite, a common biomarker for urinary tract infections (UTIs). In a healthy individual, nitrite is not found in the urine. However, a subject with a suspected UTI will produce nitrite in their urine since the bacteria present will convert nitrate into nitrite. Traditionally, nitrite is monitored by urinary dipsticks that are either read by eye or using a reflectance spectrophotometer. Both methods provide a semiquantitative positive or negative result at best. In this paper, we described a novel, affordable, portable transmission-based colorimeter for the quantitative measurement of nitrite. A unique permutation of the Griess reaction was optimized for the clinical detection of nitrite in urine and is reported. By using nitrite spiked in a salt buffer, artificial, and human urine samples, the performance of the colorimeter was evaluated against dipsticks read using two commercial dipstick analyzers, Urisys 1100 (Roche Diagnostics) and Clinitek Status+ (Siemens Medical Solutions). The colorimeter was able to detect the clinically relevant range of nitrite from 0.78 to 200 µM in a salt buffer. The detection limit in artificial urine was determined as 1.6 µM, which is â¼16× more sensitive than commercial dipstick reflectance analyzers, enabling the possibility for earlier detection of urinary infections. The colorimeter is assembled using off-the-shelf components (<$80) and controlled by a smartphone application via low-energy bluetooth. It has a built-in color correction algorithm and is designed to enable for a turbidity correction in samples containing bacteria or other cellular debris as well. The mobile application can display the nitrite concentration for a single sample or display the results over a period of time. Tracking urinalysis results longitudinally can help identify trends such as increases in nitrite concentrations over an individual's baseline and identify possible infections earlier. While the detection of nitrite was showcased here, this portable analyzer can be expanded to other colorimetric-based chemistries to detect a panel of biomarkers, which can improve the overall sensitivity and specificity of the desired assay.
RESUMEN
BACKGROUND: Microvascular invasion (MVI) has a significant effect on the prognosis of hepatocellular carcinoma (HCC), but its preoperative identification is challenging. Radiomics features extracted from medical images, such as magnetic resonance (MR) images, can be used to predict MVI. In this study, we explored the effects of different imaging sequences, feature extraction and selection methods, and classifiers on the performance of HCC MVI predictive models. METHODS: After screening against the inclusion criteria, 69 patients with HCC and preoperative gadoxetic acid-enhanced MR images were enrolled. In total, 167 features were extracted from the MR images of each sequence for each patient. Experiments were designed to investigate the effects of imaging sequence, number of gray levels (Ng), quantization algorithm, feature selection method, and classifiers on the performance of radiomics biomarkers in the prediction of HCC MVI. We trained and tested these models using leave-one-out cross-validation (LOOCV). RESULTS: The radiomics model based on the images of the hepatobiliary phase (HBP) had better predictive performance than those based on the arterial phase (AP), portal venous phase (PVP), and pre-enhanced T1-weighted images [area under the receiver operating characteristic (ROC) curve (AUC) =0.792 vs. 0.641/0.634/0.620, P=0.041/0.021/0.010, respectively]. Compared with the equal-probability and Lloyd-Max algorithms, the radiomics features obtained using the Uniform quantization algorithm had a better performance (AUC =0.643/0.666 vs. 0.792, P=0.002/0.003, respectively). Among the values of 8, 16, 32, 64, and 128, the best predictive performance was achieved when the Ng was 64 (AUC =0.792 vs. 0.584/0.697/0.677/0.734, P<0.001/P=0.039/0.001/0.137, respectively). We used a two-stage feature selection method which combined the least absolute shrinkage and selection operator (LASSO) and recursive feature elimination (RFE) gradient boosting decision tree (GBDT), which achieved better stability than and outperformed LASSO, minimum redundancy maximum relevance (mRMR), and support vector machine (SVM)-RFE (stability =0.967 vs. 0.837/0.623/0.390, respectively; AUC =0.850 vs. 0.792/0.713/0.699, P=0.142/0.007/0.003, respectively). The model based on the radiomics features of HBP images using the GBDT classifier showed a better performance for the preoperative prediction of MVI compared with logistic regression (LR), SVM, and random forest (RF) classifiers (AUC =0.895 vs. 0.850/0.834/0.884, P=0.558/0.229/0.058, respectively). With the optimal combination of these factors, we established the best model, which had an AUC of 0.895, accuracy of 87.0%, specificity of 82.5%, and sensitivity of 93.1%. CONCLUSIONS: Imaging sequences, feature extraction and selection methods, and classifiers can have a considerable effect on the predictive performance of radiomics models for HCC MVI.
RESUMEN
Magneto-acousto-electrical tomography (MAET) is designed to produce conductivity images with high spatial resolution for a conducting object. In a previous study, for an irregular conductor, transverse scanning and rotational methods with a focus transducer were combined to collect complete electrical information. This kind of method, however, is time-consuming because of the transverse scanning procedure. In this study, we proposed a novel imaging method based on plane ultrasound waves and a new aspect of projection in rotational MAET. In the proposed method, we achieved the projection in each rotation angle by using plane waves rather than mechanical scanning of the focus waves along the transverse direction. Thus, the imaging time was significantly saved. To verify the proposed method, we derived a measurement formula containing a lateral integration, which built the relationship between the measurement formula and the projection under each rotation angle. Next, we constructed two different numerical models to compute magneto-acousto-electrical signals by using a finite element method and reconstructed the corresponding conductivity parameter images based on a filtered back-projection algorithm. Then, simulated signals under different signal-to-ratios (6, 20, 40, and 60 dB) were generated to test the performance of the proposed algorithm. To improve the image quality, we further analysed the influence of the filters and the frequency scaling factors embedded in the filtered back-projection algorithm. Moreover, we computed the L2norm of the error in case of different frequency scaling factors and measurement noises. Finally, we conducted a phantom experiment with a 64-element linear phased array transducer (center frequency of 2.7 MHz) and reconstructed the conductivity parameter images of the circular phantom with an elliptical hole. The experimental results demonstrated the feasibility and time-efficiency of the proposed rapid rotational MAET.
Asunto(s)
Tomografía , Acústica , Algoritmos , Procesamiento de Imagen Asistido por Computador , Fantasmas de Imagen , Factores de Tiempo , Tomografía Computarizada por Rayos XRESUMEN
It is necessary to monitor the mechanical properties of arteries which directly related to cardiovascular diseases (CVDs) in the early stages. In this study, we proposed a new method based on deep learning (DL) to track the displacement of the vessel wall from the ultrasound radio-frequency (RF) signals, which is a key technique to achieve quantitative measurement of vascular biomechanics. In comparison with traditional method, both results on simulation and experimental carotid artery data demonstrated that the DL method has higher accuracy for motion tracking of artery walls. Hence, the DL method can be widely applied so that can predict the early pathology of cardiovascular system.
Asunto(s)
Aprendizaje Profundo , Arterias , Simulación por Computador , Ondas de Radio , UltrasonografíaRESUMEN
BACKGROUND: There is currently a debate about whether elastic stable intramedullary nails (ESIN) or external fixation (EF) is the best surgical method for treating pediatric femoral shaft fractures. We performed a meta-analysis to determine which surgical method leads to higher treatment satisfaction, lower complication rates, and reduced treatment time, to investigate whether ESIN is the preferred surgical method for treatment of pediatric femoral shaft fractures. PATIENTS AND METHODS: Relevant databases were searched for comparative studies of ESIN versus EF for the treating pediatric femoral shaft fractures. Literature reports and quality evaluations were extracted, followed by a systematic review using RevMan 5.3 software. Treatment satisfaction at the last follow-up, primary complications, secondary complications, and relevant time indicators (operation time, hospital stay, clinical healing time, bone healing time) were analyzed. RESULTS: A total of 22 reports were included in this meta-analysis. We found no statistical differences in the treatment satisfaction at the last follow-up between ESIN and EF for the treatment of pediatric femoral shaft fractures. A low rate of postoperative re-fracture (RR=3.58, 95% CI (1.85, 6.92), p=0.0001) and postoperative infection (RR=9.25, 95% CI (5.32, 16.11), p<0.00001), and a high risk of skin irritation (RR=0.15, 95% CI (0.06, 0.37), p<0.00001) were found in the ESIN group. No significant differences between the two approaches were found regarding malunion. A low rate of limb-length discrepancy (RR=2.41, 95% CI (1.40, 4.17), p=0.002), hospitalization (SMD=0.84, 95% CI (0.24, 1.43), p=0.006), clinical healing time (SMD=0.95, 95% CI (0.56, 1.33), p<0.00001) and bone healing time (SMD=0.89, 95% CI (0.39, 1.40), p=0.005) were found in the ESIN group, as compared to that in the EF group. No significant differences were found in fixation failure, activity limitation of the joint, and operation time between the two strategies. DISCUSSION: ESIN should be the primary choice for the treatment of pediatric femoral shaft fractures since it has a reliable curative effect and results in a shorter hospital stay, faster fracture healing, and fewer complications. EF is recommended for fractures with serious injury of the soft tissue to avoid intramedullary infection. Double-blind high-quality randomized studies with larger sample sizes are warranted to confirm our conclusions. LEVEL OF EVIDENCE: IV.
Asunto(s)
Fracturas del Fémur , Fijación Intramedular de Fracturas , Clavos Ortopédicos , Niño , Diáfisis , Fijadores Externos , Fracturas del Fémur/cirugía , Fijación de Fractura , Fijación Intramedular de Fracturas/efectos adversos , Curación de Fractura , Humanos , Ensayos Clínicos Controlados Aleatorios como Asunto , Resultado del TratamientoRESUMEN
Real time brain transcranial ultrasound imaging is extremely intriguing because of its numerous applications. However, the skull causes phase distortion and amplitude attenuation of ultrasound signals due to its density: the speed of sound is significantly different in bone tissue than in soft tissue. In this study, we propose an ultrafast transcranial ultrasound imaging technique with diverging wave (DW) transmission and a deep learning approach to achieve large field-of-view with high resolution and real time brain ultrasound imaging. DW transmission provides a frame rate of several kiloHz and a large field of view that is suitable for human brain imaging via a small acoustic window. However, it suffers from poor image quality because the diverging waves are all unfocused. Here, we adopted adaptive beamforming algorithms to improve both the image contrast and the lateral resolution. Both simulated and in situ experiments with a human skull resulted in significant image improvements. However, the skull still introduces a wavefront offset and distortion, which degrades the image quality even when adaptive beamforming methods are used. Thus, we also employed a U-Net neural network to detect the contour and position of the skull directly from the acquired RF signal matrix. This approach avoids the need for beamforming, image reconstruction, and image segmentation, making it more suitable for clinical use.
Asunto(s)
Encéfalo/diagnóstico por imagen , Aprendizaje Profundo , Cráneo/diagnóstico por imagen , Ultrasonografía/métodos , Humanos , Procesamiento de Imagen Asistido por Computador/métodos , Fantasmas de ImagenRESUMEN
PURPOSE: To investigate the impact of prostate weight on outcomes of nerve sparing laparoscopic radical prosta-tectomy (LRP) and assess its predictive value on postoperative continence and potency recovery. MATERIALS AND METHODS: We conducted a retrospective study on the clinical data of 165 patients with low risk prostate cancer (PCa) who underwent nerve sparing LRP. All the patients included had normal preoperative uri-nary and sexual function. The association of prostate weight with perioperative data was assessed using Spearman correlation coefficient. Univariate and multivariate Cox regression analyses were employed to identify prognostic predictors for continence and potency recovery. RESULTS: Increased prostate weight was significantly associated with older age, higher prostate-specific antigen (PSA), lower biopsy and pathological T stage and Gleason score, longer operative time, and higher estimated blood loss (P < .05). The continence rates at the 3rd, 6th, and 12th month after surgery were 63.6% (105/165), 87.9% (145/165), and 95.8% (158/165); and the potency rates were 44.8% (74/165), 62.4% (103/165) and 77.6% (128/165), respectively. Furthermore, multivariate Cox analysis showed that patient age (HR = 0.52, 95% CI: 0.35- 0.76) and prostate weight (HR = 0.54, 95% CI: 0.34-0.86) were independent predictors for continence recovery, while only patient age (HR = 0.66, 95% CI: 0.45-0.96) could independently predict potency recovery. CONCLUSION: Larger prostate size was correlated with older age, higher PSA, lower tumor stage and grade, longer operative time, and more intraoperative blood loss in low risk PCa patients. Increased prostate weight may inde-pendently predict poor continence recovery after nerve sparing LRP.