Biomedical Physics & Engineering Express最新文献

Traumatic injury remains a leading cause of death worldwide, with traumatic bleeding being one of its most critical and fatal consequences. The use of whole-body computed tomography (WBCT) in trauma management has rapidly expanded. However, interpreting WBCT images within the limited time available before treatment is particularly challenging for acute care physicians. Our group has previously developed an automated bleeding detection method in WBCT images. However, further reduction of false positives (FPs) is necessary for clinical application. To address this issue, we propose a novel automated detection for traumatic bleeding in CT images using deep learning and multi-organ segmentation; Methods: The proposed method integrates a three-dimensional U-Net# model for bleeding detection with an FP reduction approach based on multi-organ segmentation. The multi-organ segmentation method targets the bone, kidney, and vascular regions, where FPs are primarily found during the bleeding detection process. We evaluated the proposed method using a dataset of delayed-phase contrast-enhanced trauma CT images collected from four institutions; Results: Our method detected 70.0% of bleedings with 76.2 FPs/case. The processing time for our method was 6.3 ± 1.4 min. Compared with our previous ap-proach, the proposed method significantly reduced the number of FPs while maintaining detection sensitivity.

Objective: Arterial sampling for PET imaging often involves continuously measuring the radiotracer activity concentration in blood using an automatic blood sampling system (ABSS). We proposed and validated an external delay and dispersion correction procedure needed when a change in flow rate occurs during data acquisition. We also measured the external dispersion constant of [11C]CURB, [18F]FDG, [18F]FEPPA, and [18F]SynVesT-1.

Approach: External delay and dispersion constants were measured for the flow rates of 350, 300, 180, and 150 mL/h, using 1-minute-long rectangular inputs (n = 10; 18F-fluoride in saline). Resulting constants were used to validate the external delay and dispersion corrections (n = 6; 18F-fluoride in saline; flow rate change: 350 to 150 mL/h and 300 to 180 mL/h); constants were modelled to transition linearly between flow rates. Corrected curves were assessed using the percent area-under-the-curve (AUC) ratio and a modified model selection criterion (MSC). External delay and dispersion constants were measured for various radiotracers using a blood analog (i.e., similar viscoelastic properties).

Main results: ABSS outputs were successfully corrected for external delay and dispersion using our proposed method accounting for a change in flow rate. AUC ratio reduced from ~10% for the uncorrected 350-150 mL/h output (~6% for the 300-180 mL/h) to < 1% after correction when compared to true input (511 keV energy window); approx. 5-fold increase in MSC. Assuming an internal dispersion constant of 5 seconds, the dispersion constant (internal + external) for [11C]CURB, [18F]FDG, [18F]FEPPA, and [18F]SynVesT-1 was 13, 9, 16, and 10 s, respectively.

Significance: This study presented an external delay and dispersion correction procedure needed when a change in flow rate occurs during ABSS data acquisition. Additionally, this is the first study to measure the external delay and dispersion constants using a blood analog solution, a suitable alternative to blood when estimating external dispersion.

Multiple sclerosis (MS) is a disorder in which the body's immune system attacks structures of the central nervous system, resulting in lesions that can occur throughout the brain and spinal cord. Cortical lesions, in particular, can contribute to motor dysfunction. Walking disability is reported as the main impairment by people with MS (pwMS), often due to limited ankle movement. This study explored the event-related desynchronization (ERD) onset latency of the sensorimotor rhythms during foot dorsiflexion in pwMS computed using an objective and independent of human criterion method, as an electroencephalogram (EEG) based biomarker. EEG signals were recorded in eight persons with neither neurological condition nor motor dysfunction and eight pwMS with relapsing-remitting, primary progressive or secondary progressive MS. Recordings were divided into three groups: control, more affected lower limb and less affected lower limb. The ERD-onset latency was determined using a method based on the percent of ERD time course and the cluster depth tests. The median and interquartile range of the ERD-onset latency were 1186.0 (1100.0, 1250.0) ms; 1064.0 (1031.0, 1127.0) ms for the more and less affected groups respectively, whereas the median and interquartile range for the control group was 656.0 (472.2, 950.0) ms. There was a significant delay in the ERD-onset latencies of the pwMS groups compared to the control group (p<0.001 for both comparisons). These findings suggest that the ERD-onset latency computed using the proposed method could be used as an EEG biomarker to evaluate disease progression or therapeutic interventions in pwMS.

Conventional x-ray radiography relies on attenuation differences in the object, which often results in poor contrast in soft tissues. X-ray phase imaging has the potential to produce higher contrast but can be difficult to utilize. Instead of grating-based techniques, analyzer-based imaging, also known as diffraction enhanced imaging (DEI), uses a monochromator crystal with an analyzer crystal after the object. Analyzer-based systems most commonly employ synchrotron sources to provide adequate intensity, and typically use higher photon energies. In this work, a simulation has been devised to assess the potential for a polycapillary-based system. A polycapillary collimating optic has previously been shown to greatly enhance the intensity of the beam diffracted from the monochromatizing crystal. Detailed simulation of the optic is computationally intensive and requires comprehensive knowledge of the internal shape of the optic, so a simple geometric model using easier to obtain optic output data was developed and compared to the more detailed simulation. After verification, refraction band visibility was used as a quality parameter to address the effectiveness of the polycapillary-based DEI system at x-ray photon energies of 8 and 17.5 keV. The result shows promise for a polycapillary-coupled analyzer-based system even at low x-ray photon energy.

Mild cognitive impairment (MCI) is a significant predictor of the early progression of Alzheimer's disease (AD) and can serve as an important indicator of disease progression. However, many existing methods focus mainly on the image when processing brain imaging data, ignoring other non-imaging data (e.g., genetic, clinical information, etc.) that may have underlying disease information. In addition, imaging data acquired from different devices may exhibit varying degrees of heterogeneity, potentially resulting in numerous noisy connections during network construction. To address these challenges, this study proposes a Multimodal Multiview Bilinear Graph Convolution (MMBGCN) framework for disease risk prediction. Firstly, grey matter (GM), white matter (WM) and cerebrospinal fluid (CSF) features are extracted from magnetic resonance imaging (MRI), and non-imaging information is combined with the features extracted from MRI to construct a multimodal shared adjacency matrix. The shared adjacency matrix is then used to construct the multiview network so that the effect of potential disease information in the non-imaging information on the model can be considered. Finally, the MRI features extracted by the multiview network are weighted to reduce noise, and then the spatial pattern is restored by bilinear convolution. The features of the recovered spatial patterns are then combined with the genetic information for disease prediction. The proposed method is tested on the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset. Extensive experiments demonstrate the superior performance of the proposed framework and its ability to outperform other related algorithms. The average classification accuracy in the binary classification task in this study is 89.6%. The experimental results demonstrate that the method proposed in this study facilitates research on MCI diagnosis using multimodal data.

For brain-dedicated positron emission tomography (PET) scanners, depth-of-interaction (DOI) information is essential to achieve uniform spatial resolution across the field-of-view (FOV) by minimizing parallax error. Time-of-flight (TOF) information can enhance the image quality. In this study, we proposed a novel monolithic U-shaped crystal design that had a tapered geometry to achieve good coincidence timing resolution (CTR) and DOI resolution simultaneously. We compared a novel tapered U-shaped crystal design with a conventional flat-top geometry for PET detectors. Each crystal had outer dimensions of 5.85 × 2.75 × 15 mm³, with a 0.2 mm central gap forming physically isolated bottom surfaces (2.85 × 2.75 mm²). The novel U-shape crystal design with tapered top roof resulted in the best CTR of 201±3 ps, and DOI resolution of 3.1±0.6 mm, which were better than flat top geometry. In the next study, we plan to optimize the crystal surface treatment and reflector to further improve the CTR and DOI resolution.

The optimal method for three-dimensional thermal imaging within cells involves collecting intracellular temperature responses while simultaneously obtaining corresponding 3D positional information. Current temperature measurement techniques based on the photothermal properties of quantum dots face several limitations, including high cytotoxicity and low fluorescence quantum yields. These issues affect the normal metabolic processes of tumor cells. This study synthesizes a low-toxicity cell membrane-targeted quantum dot temperature sensor by optimizing the synthesis method of CdTe/CdS/ZnS core-shell structured quantum dots. Compared to CdTe-targeted quantum dot temperature sensors, the cytotoxicity of CdTe/CdS/ZnS-targeted quantum dot temperature sensors is reduced by 40.79%. Additionally, a novel cepstrum-based spatial localization algorithm is proposed to achieve rapidly compute the three-dimensional positions of densely distributed quantum dot temperature sensors. Ultimately, both targeted and non-targeted CdTe/CdS/ZnS quantum dot temperature sensors were used simultaneously to label the internal and external regions of human osteosarcoma cells to obtain temperature data at these labeling positions. By combining this with the cepstrum-based spatial localization algorithm, the spatial coordinates of the quantum dot temperature sensors were obtained. Three-dimensional temperature field reconstruction of three local regions was achieved within a 12 μm axial range in living cells. The method described in this paper can be widely applied to the quantitative study of intracellular thermal responses.

Cancer grade classification is a challenging task identified from the cell structure of healthy and abnormal tissues. The practitioners learns about the malignant cell through the grading and plans the treatment strategy accordingly. A major portion of researchers used DL models for grade classification. However, the behavior of DL models is hidden type, it is unknown which features contribute to the accuracy and how the features are chosen for grading. To address the issue the study proposes a Grade Differentiation Integrated Model (GradeDiff-IM) to classify the grades G1, G2, and G3. In GradeDiff-IM, different ML models, are used for grade classification from clinical and pathological reports. The biological-significant features with ranking technique prioritize influential features are used to identify grades G. Subsequently, histopathological images are used by DL models for grade classification and compared with ML models. Instead of employing a single ML model, the GradeDiff-IM model uses the stack-ensembled approach to improve the grade G classification performance. The maximum accuracy is attained by stacking G1-98.2, G2-97.6, and G3-97.5. The proposed study shows that the ML ensemble model is more accurate than the DL models. As a result, the proposed model achieved higher accuracy for G by implementing the stacking technique than the other state-of-the-art models.

In fundus images, precisely segmenting retinal blood vessels is important for diagnosing eye-related conditions, such as diabetic retinopathy and hypertensive retinopathy or other eye-related disorders. In this work, we propose an enhanced U-shaped network with dual-attention, named DAU-Net, divided into encoder and decoder parts. Wherein, we replace the traditional convolutional layers with ConvNeXt Block and SnakeConv Block to strengthen its recognition ability for different forms of blood vessels while lightweight the model. Additionally, we designed two efficient attention modules, namely Local-Global Attention (LGA) and Cross-Fusion Attention (CFA). Specifically, LGA conducts attention calculations on the features extracted by the encoder to accentuate vessel-related characteristics while suppressing irrelevant background information; CFA addresses potential information loss during feature extraction by globally modeling pixel interactions between encoder and decoder features. Comprehensive experiments in terms of public datasets DRIVE, CHASE_DB1, and STARE demonstrate that DAU-Net obtains excellent segmentation results on all three datasets. The results show an AUC of 0.9818, ACC of 0.8299, and F1 score of 0.9585 on DRIVE; 0.9894, 0.8499, and 0.9700 on CHASE_DB1; and 0.9908, 0.8620, and 0.9712 on STARE, respectively. These results strongly demonstrate the effectiveness of DAU-Net in retinal vessel segmentation, highlighting its potential for practical clinical use.

Background and purpose. This study presents machine learning (ML) models that predict if deep inspiration breath hold (DIBH) is needed based on lung dose in right-sided breast cancer patients during the initial computed tomography (CT) appointment.Materials and methods. Anatomic distances were extracted from a single-institution dataset of free breathing (FB) CT scans from locoregional right-sided breast cancer patients. Models were developed using combinations of anatomic distances and ML classification algorithms (gradient boosting, k-nearest neighbors, logistic regression, random forest, and support vector machine) and optimized over 100 iterations using stratified 5-fold cross-validation. Models were grouped by the number of anatomic distances used during development; those with the highest validation accuracy were selected as final models. Final models were compared based on their predictive ability, measurement collection efficiency, and robustness to simulated user error during measurement collection.Results. This retrospective study included 238 patients treated between 2016 and 2021. Model development ended once eight anatomic distances were included, and the validation accuracy plateaued. The best performing model used logistic regression with four anatomic distances achieving 80.5% average testing accuracy, with minimal false negatives and positives (<27%). The anatomic distances required for prediction were collected within 3 min and were robust to simulated user error during measurement collection, changing accuracy by <5%.Conclusion. Our logistic regression model using four anatomic distances provided the best balance between efficiency, robustness, and ability to predict if DIBH was needed for locoregional right-sided breast cancer patients.