Image Slices Research Articles

Low Energy Electron Induced Dissociation of Acetylacetone.

We report a detailed investigation of the dissociative electron attachment (DEA) to acetylacetone, an important molecule for industrial and atmospheric chemistry. We find that the DEA process leads to several dissociation channels, leading to the formation of and other anions with m/z 39, 41, 43, 57, 83, and 85, with the H- and OH- being the most dominant channels. The absolute cross sections for both of these channels peak around 8.8 and 9 eV with cross-section values of 3.1 × 10-18 and 3.7 × 10-19 cm2, respectively. We also report the DEA dynamics for these channels based on the Velocity Slice Images that were obtained. We also compare the dissociation pattern from DEA with that from photodissociation reported for 248 and 266 nm wavelengths and propose the possibility of the neutral excited state's role as the underlying parent state for the negative ion resonance.

Low-Intensity Ultrasound Facilitation of Intranasal Drug Delivery to Olfactory Bulb and Trigeminal Nerves.

Nasal-to-brain (NtoB) delivery is a noninvasive approach that uses the nasal cavity as a pathway to transport therapeutic agents directly to the brain. This approach bypasses systemic circulation and avoids the blood-brain barrier (BBB). Transcranial ultrasound, coupled with microbubbles (MB), is a technique used to oscillate and generate acoustic cavitation to open the capillary tight junctions of BBB temporarily. Its efficacy in facilitating NtoB delivery has been demonstrated in vivo. However, while opening the BBB, sonication with MB poses the risk of cerebral microhemorrhage or brain tissue damage due to sonication-induced physical injury. This study aimed to assess the effectiveness of low-intensity ultrasound treatment to facilitate NtoB delivery in a mouse model without using MB. In this study, 10-kDa dextran was administered intranasally (IN), and transcranial planar US was applied to the entire mouse brain without MB assistance. Ex-vivo whole brain imaging via fluorescence macroscopy, brain slice analysis with fluorescence microscope, and quantification of dextran concentration in distinct brain regions were conducted to compare the IN-only, IN combined with US (IN+US), and sham groups. For the trigeminal nerves (TN), fluorescence macroscopy, microscopy, and TN concentration quantification were performed to compare the three groups. Whole brain imaging revealed that US facilitated the IN delivery of dextran to the olfactory bulb (OB) in the IN+US group compared with that in the IN-only and sham groups; however, this difference was not observed after a 24 h follow-up. Conversely, brain slice images showed that the tracer was delivered to the OB, cerebral cortex, striatum and brainstem in the IN+US group, but this finding was not observed in the IN-only group at the 4 h mark. The quantification of fluorescence intensity at two follow-up time points revealed no significant difference between the IN and IN+US groups in these specific regions. Dextran concentration analysis for distinct brain areas and TN showed that ultrasound significantly increased the tracer concentration delivered to the OB and TN in the IN+US group at the 4 h mark compared with that in the IN-only and sham groups; however, this effect was not sustained at 24 h. Confocal microscopy indicated that the dextran tracer accumulated in the perivascular space along the microvascular structures. We demonstrated the efficacy of low-intensity ultrasound without using MB, in enhancing nose-to-OB and nose-to-TN drug delivery, and proposed the potential for future clinical application. Thus, we showed that this approach was safe, without evidence of microhemorrhage or brain tissue damage.

Achieving theranostic probes targeting BRD3/BRD4 for imaging and therapy of tumor.

The BET family proteins play pleiotropic roles in the tumorigenesis and growth of various human malignancies that have aroused great interests as the cancer therapeutic targets. Therefore, it's significant to develop labeling toolkits for the dynamic monitoring of BET family proteins in living tumor cells and tissue slices. In particular, there are few small-molecule fluorescent probes based on BET family proteins developed for real-time imaging of these proteins in tumor cells and tissues and treatment of breast cancer. In general, antibodies of BET family proteins are chosen for imaging of these proteins. However, the cost of these antibodies is more expensive than small molecules and the operation is relatively complicated. Moreover, the antibodies for imaging are not capable of the therapy for breast cancer. Thus, it's essential to exploit a novel imaging system for BET family proteins which is easy-operating and economical, with achieving therapeutic effects simultaneously. Therefore, a series of fluorescent probes (17-21) targeting BET family proteins were developed. Through the evaluation studies, probe 17 showed advantages in docking studies, analysis of cell viability, and imaging studies. Importantly, probe 17 was capable of distinguishing tumor cells and tissue slices and made a distinction between them by labeling BRD3 and BRD4 proteins. Importantly, probe 17 showed higher resolution in mouse tumor and human tumor slice imaging than BRD3 and BRD4 antibodies. Moreover, compared with these antibodies, probe 17 was more stable, more economical and easier to operate in the imaging assays. In addition, it also played an anti-tumor role by inducing cell apoptosis, proliferation inhibition, and cycle arrest in tumor cells. All these features render probe 17 to perform as an effective labeling toolkit compatible for imaging studies of BRD3/BRD4, as well as an approach to diagnosis and treatment of breast cancer.

Assessing Titania Pigment Distribution in Water-Based Coatings by Image Analysis of Ion Beam Milled Sections.

By applying various image analysis methods, the distribution of titania pigments in water-based paint films is assessed in this work. Cross-sections of paint films containing titania are prepared using triple ion beam milling, and the milled cross-sections are imaged using scanning electron microscopy. The obtained morphology of the paint films with known difference in pigment distribution is then determined and quantified by means of image analysis. It is found that the inter-particle distance which is calculated by determining the distance between each individual particle can serve as good differentiator for assessing the quality of pigment distribution. Furthermore, a comparison is made between 2D (triple ion beam milling) and 3D (focused ion beam with slice and view) imaging methods on the same sample. The 2D method, with its ease of use, short experimental time and artefact free imaging, gives better results in the system studied in this work.

Longitudinal Changes in Medial Meniscal Extrusion After ACL Injury and Reconstruction and Its Relationship With Cartilage Degeneration Assessed Using MRI-Based T1ρ and T2 Analysis

Background: Anterior cruciate ligament (ACL) injury often leads to posttraumatic osteoarthritis (PTOA), despite ACL reconstruction (ACLR). Medial meniscal extrusion (MME) is implicated in PTOA progression but remains understudied after ACL injury and ACLR. Hypothesis/Purpose: It was hypothesized that MME would increase longitudinally after ACL injury and ACLR, with greater changes in the ipsilateral knee compared with the contralateral knee, leading to cartilage degeneration. The study aimed to assess MME 3 years after ACLR and its relationship with magnetic resonance imaging (MRI) T1ρ and T2 as cartilage degeneration markers. Study Design: Cohort study; Level of evidence, 2. Methods: MME and relative percentage of extrusion (RPE) were measured on 3 coronal slices of 3-dimensional fast spin-echo images and the mean values were used. T1ρ and T2 sequences were obtained and cartilage compositional measurements were performed using in-house developed software with MATLAB. Mixed models were used to assess the longitudinal changes and linear regression was used to assess the relationships between RPE and T1ρ and T2 values. Results: A total of 54 participants with unilateral ACL injuries underwent preoperative bilateral knee MRI. A total of 36 participants completed MR scans at 6 months and 3 years after ACLR. MME and RPE measurements demonstrated high reliability (ICC > 0.88 and > 0.91, respectively). The predicted values of MME and RPE from the mixed models showed that the ipsilateral side had significantly greater MME and RPE than the contralateral side at all 3 time points (P = .023 for MME; P = .013 for RPE at baseline; and P < .001 at 6 months and P < .001 at 3 years for both MME and RPE). The rate of change of MME and RPE on the ipsilateral side was significantly greater than that on the contralateral side (P < .001). Postoperative RPE was associated with T1ρ and T2 values in the posterior medial femoral condyle. Conclusion: MME and RPE obtained pre- and postoperatively after ACLR on the ipsilateral side were significantly greater than those on the contralateral side, and the longitudinal increases on the ipsilateral side were greater than those on the contralateral side. Postoperative RPE was significantly associated with cartilage degeneration in the posterior medial femoral condyle.

Deep learning-based prediction of mortality using brain midline shift and clinical information.

Brain midline shift (MLS) indicates the severity of mass effect from intracranial lesions such as traumatic brain injury, stroke, brain tumor, or hematoma. Brain MLS can be used to determine whether patients require emergency surgery and to predict patients' prognosis. Since brain MLS is usually emergent, it must be diagnosed immediately. Therefore, this study presents a computer-aided deep-learning method for detecting MLS, aiming to predict mortality in a prognosis-predicting cohort using brain MLS and clinical in-formation. The brain midline is a 3-dimensional structure, but computed tomography (CT) slices are 2-dimensional which limits brain MLS detection. Here we propose a keypoint detection method to detect brain midline on each CT slice, acquiring brain MLS distance and area in each slice. Combined with clinical information, patient mortality can be predicted using the multilayer perceptron (MLP) model. The accuracy, precision, sensitivity, specificity, and F1-score for slice selection with the proposed model are 0.966, 0.952, 0.991, 0.932, and 0.971, respectively. Both MLS distance and volume were precisely predicted at slice-level and case-level with only the slightest error. The detected midlines were clearly separated into left and right brain with a dice coefficient of 0.98. The accuracy and AUC of the MLP model were both above 0.8. The model detected large brain MLS cases well in the prediction of outcomes in the prognosis-predicting cohort. The method performs well on slice selection and brain MLS detection, and predictions of MLS distance and volume combined with clinical information predicts the patient's prognosis well.

SFCLI-Net: Spatial-frequency collaborative learning interpolation network for Computed Tomography slice synthesis

SFCLI-Net: Spatial-frequency collaborative learning interpolation network for Computed Tomography slice synthesis

Accuracy and user experience of dental diagnosis of a patient with cleidocranial dysplasia using immersive virtual reality and cone-beam computed tomography multiplanar reconstructions.

Accuracy and user experience of dental diagnosis for a patient with cleidocranial dysplasia (CCD) using immersive virtual reality (VR) and cone-beam computed tomography multiplanar reconstruction methods were evaluated. Dental students (n= 40) were randomly assigned to VR or MP groups. VR participants manipulated and visualized the rendered 3-dimensional model using VR hardware and software. The MP participants viewed cone-beam computed tomography slices using orthogonal planes and 3-dimensional rendered images on a computer. Participants identified erupted and unerupted primary, permanent, and supernumerary teeth in a patient with CCD and completed presurvey, postsurvey, Presence, and NASA Task Load Index questionnaires. The VR group was significantly more accurate in identifying supernumerary teeth (P<0.008), developing permanent teeth (P<0.020), and primary teeth (P<0.05) in the maxillary anterior region than the MP group. There was more variability in the accuracy of tooth identification for the MP participants (P= 0.005). The Presence Questionnaire showed that participants from the VR group had significantly greater feelings of control and sensory factors during the task (P<0.05). The NASA Task Load Index showed that the MP group participants worked harder and required more mental demands to accomplish the same task (P<0.001). This study showed that for novice clinicians, the VR method might offer a more accurate method of dental diagnosis of a patient with CCD who presents with retained primary teeth, multiple impacted, and supernumerary teeth. In addition, the VR participants showed increased engagement and a sense of presence.

Gross target volume contouring in canine extra-axial brain tumors: Effects of magnetic resonance image slice thickness and time between subsequent image sets.

Accurate determination of the gross target volume (GTV) is critical in radiation treatment planning, as errors could result in underdosing of the tumor or overdosing of nearby organs at risk. This multicenter retrospective observational serial measurement study evaluated the effects of variations in MRI slice thickness and a time delay between the diagnostic (MRI-1) and RT planning (MRI-2) MRIs GTV contouring in dogs with presumed meningiomas. The hypothesis was that the GTV would increase in size with time on T1-weighted sequences with contrast. Inclusion required paired MRI acquisition within 3 months. The GTV was contoured on each MRI. Forty-six dogs were included. Slice thickness was significantly different (P<.001) between MRIs: MRI-1 had a median of 3.9mm (range: 0.8-6mm; only two dogs <2mm), and MRI-2 had a median of 0.9mm (range: 0.6-4.5mm; only two dogs >2mm). The median time between MRIs was 22 days (range: 8-74 days). The MRI-1 GTV was significantly different from MRI-2 GTV (P<.0001); thirty (65%) were larger, five were equal in size, and 12 were smaller than the MRI-2 GTV. This difference in GTV is likely due to the slice thickness differences between MRI acquisitions rather than changes in tumor size due to the short time interval between MRI-1 and MRI-2. This finding highlights the differences between diagnostic and RT treatment-planning MRIs. For brain tumor target contouring, an MRI at the same time as the RT planning CT with <1mm slice thickness, 3D acquisitions, and anisotropic voxel is recommended.

Lobectomy Increases Postoperative Pulmonary Artery Enlargement to a Greater Extent than Segmentectomy.

The underlying mechanism why segmentectomy has demonstrated the non-inferiority to lobectomy in several randomized trials remains unclear. Computed tomography (CT)-measured pulmonary artery (PA) enlargement reflects PA pressure and predicts the prognosis of certain respiratory diseases. We compared the preoperative and postoperative PA diameter to the ascending aorta diameter (PA/A) ratio, investigating its impact on right ventricular function in lung resection. This retrospective study was conducted in patients with lower-lobe lung tumors who underwent anatomical lung resection between 2017 and 2022. The PA diameter at the bifurcation and the ascending aorta diameter at the same CT image slice were measured preoperatively and postoperatively. We calculated the enlargement of PA/A ratio (PA/A change) and compared lobectomy and segmentectomy. This analysis included 279 patients (235 with lobectomy and 44 with segmentectomy). The PA/A change was significantly greater in patients with lobectomy than segmentectomy (104% vs. 102%, P = 0.02). In the multivariable analysis, airflow obstruction (yes, P = 0.04) and the type of surgery (segmentectomy, P = 0.04) were independent prognostic factors for PA/A change. The PA/A change was greater in lobectomy than in segmentectomy. This change could reflect a burden on right ventricular function after lobectomy.

Inclusive, exclusive and hierarchical atlas of NFATc1+/PDGFR-α+ cells in dental and periodontal mesenchyme.

Platelet-derived growth factor receptor alpha (PDGFR-α) activity is crucial in the process of dental and periodontal mesenchyme regeneration facilitated by autologous platelet concentrates (APCs), such as platelet-rich fibrin (PRF), platelet-rich plasma (PRP) and concentrated growth factors (CGF), as well as by recombinant PDGF drugs. However, it is largely unclear about the physiological patterns and cellular fate determinations of PDGFR-α+ cells in the homeostasis maintaining of adult dental and periodontal mesenchyme. We previously identified NFATc1 expressing PDGFR-α+ cells as a subtype of skeletal stem cells (SSCs) in limb bone in mice, but their roles in dental and periodontal remain unexplored. To this end, in the present study we investigated the spatiotemporal atlas of NFATc1+ and PDGFR-α+ cells residing in dental and periodontal mesenchyme in mice, their capacity for progeny cell generation, and their inclusive, exclusive and hierarchical relations in homeostasis. We utilized CRISPR/Cas9-mediated gene editing to generate two dual recombination systems, which were Cre-loxP and Dre-rox combined intersectional and exclusive reporters respectively, to concurrently demonstrate the inclusive, exclusive, and hierarchical distributions of NFATc1+ and PDGFR-α+ cells and their lineage commitment. By employing the state-of-the-art transgenic lineage tracing techniques in cooperating with tissue clearing-based advanced imaging and three-dimensional slices reconstruction, we systematically mapped the distribution atlas of NFATc1+ and PDGFR-α+ cells in dental and periodontal mesenchyme and tracked their in vivo fate trajectories in mice. Our findings extend current understanding of NFATc1+ and PDGFR-α+ cells in dental and periodontal mesenchyme homeostasis, and furthermore enhance our comprehension of their sustained therapeutic impact for future clinical investigations.

Noise Reduction in Brain CT: A Comparative Study of Deep Learning and Hybrid Iterative Reconstruction Using Multiple Parameters.

We evaluated the noise reduction effects of deep learning reconstruction (DLR) and hybrid iterative reconstruction (HIR) in brain computed tomography (CT). CT images of a 16 cm dosimetry phantom, a head phantom, and the brains of 11 patients were reconstructed using filtered backprojection (FBP) and various levels of DLR and HIR. The slice thickness was 5, 2.5, 1.25, and 0.625 mm. Phantom imaging was also conducted at various tube currents. The noise reduction ratio was calculated using FBP as the reference. For patient imaging, overall image quality was visually compared between DLR and HIR images that exhibited similar noise reduction ratios. The noise reduction ratio increased with increasing levels of DLR and HIR in phantom and patient imaging. For DLR, noise reduction was more pronounced with decreasing slice thickness, while such thickness dependence was less evident for HIR. Although the noise reduction effects of DLR were similar between the head phantom and patients, they differed for the dosimetry phantom. Variations between imaging objects were small for HIR. The noise reduction ratio was low at low tube currents for the dosimetry phantom using DLR; otherwise, the influence of the tube current was small. In terms of visual image quality, DLR outperformed HIR in 1.25 mm thick images but not in thicker images. The degree of noise reduction using DLR depends on the slice thickness, tube current, and imaging object in addition to the level of DLR, which should be considered in the clinical use of DLR. DLR may be particularly beneficial for thin-slice imaging.

Modernizing histopathological analysis: a fully automated workflow for the digital image analysis of the intestinal microcolony survival assay.

Manual analysis of histopathological images is often not only time-consuming and painstaking but also prone to error from subjective evaluation criteria and human error. To address these issues, we created a fully automated workflow to enumerate jejunal crypts in a microcolony survival assay to quantify gastrointestinal damage from radiation. After abdominal irradiation of mice, jejuna were obtained and prepared on histopathologic slides, and crypts were counted manually by trained individuals. The automated workflow (AW) involved obtaining images of jejunal slices from the irradiated mice, followed by cropping and normalizing the individual slice images for resolution and color; using deep learning-based semantic image segmentation to detect crypts on each slice; using a tailored algorithm to enumerate the crypts; and tabulating and saving the results. A graphical user interface (GUI) was developed to allow users to review and correct the automated results. Crypts counted manually exhibited a mean absolute percent deviation of (34 ± 26)% between individuals vs the group mean across counters, which was reduced to (11 ± 6)% across the 3 most-experienced counters. The AW processed a sample image dataset from 60 mice in a few hours and required only a few minutes of active user effort. AW counts deviated from experts' mean counts by (10 ± 8)%. The AW thereby allowed rapid, automated evaluation of the microcolony survival assay with accuracy comparable to that of trained experts and without subjective inter-observer variation. We fully automated the digital image analysis of a microcolony survival assayAnalyzing 540 images takes a few hours with only minutes of active user effortThe automated workflow (AW) is just as accurate as trained expertsThe AW eliminates subjective inter-observer variation and human errorHuman review possible with built-in graphical user interface.

Deep Learning for Detecting and Subtyping Renal Cell Carcinoma on Contrast-Enhanced CT Scans Using 2D Neural Network with Feature Consistency Techniques

Abstract Objective The aim of this study was to explore an innovative approach for developing deep learning (DL) algorithm for renal cell carcinoma (RCC) detection and subtyping on computed tomography (CT): clear cell RCC (ccRCC) versus non-ccRCC using two-dimensional (2D) neural network architecture and feature consistency modules. Materials and Methods This retrospective study included baseline CT scans from 196 histopathologically proven RCC patients: 143 ccRCCs and 53 non-ccRCCs. Manual tumor annotations were performed on axial slices of corticomedullary phase images, serving as ground truth. After image preprocessing, the dataset was divided into training, validation, and testing subsets. The study tested multiple 2D DL architectures, with the FocalNet-DINO demonstrating highest effectiveness in detecting and classifying RCC. The study further incorporated spatial and class consistency modules to enhance prediction accuracy. Models' performance was evaluated using free-response receiver operating characteristic curves, recall rates, specificity, accuracy, F1 scores, and area under the curve (AUC) scores. Results The FocalNet-DINO architecture achieved the highest recall rate of 0.823 at 0.025 false positives per image (FPI) for RCC detection. The integration of spatial and class consistency modules into the architecture led to 0.2% increase in recall rate at 0.025 FPI, along with improvements of 0.1% in both accuracy and AUC scores for RCC classification. These enhancements allowed detection of cancer in an additional 21 slices and reduced false positives in 126 slices. Conclusion This study demonstrates high performance for RCC detection and classification using DL algorithm leveraging 2D neural networks and spatial and class consistency modules, to offer a novel, computationally simpler, and accurate DL approach to RCC characterization.

Automated Characterization of Abdominal MRI Exams Using Deep Learning.

Advances in magnetic resonance imaging (MRI) have revolutionized disease detection and treatment planning. However, as the volume and complexity of MRI data grow with increasing heterogeneity between institutions in imaging protocol, scanner technology, and data labeling, there is a need for a standardized methodology to efficiently identify, characterize, and label MRI sequences. Such a methodology is crucial for advancing research efforts that incorporate MRI data from diverse populations to develop robust machine learning models. This research utilizes convolutional neural networks (CNNs) to automatically classify sequence, orientation, and contrast, specifically tailored for abdominal MRI. Three distinct CNN models with similar backbone architectures were trained to classify single image slices into one of 12 sequences, 4 orientations, and 2 contrast classes. Results derived from this method demonstrate high levels of performance for the three specialized CNN models, with model accuracies for sequence, orientation, and contrast of 96.9%, 97.4%, and 97.3%, respectively.

Delta opioid receptor agonists activate PI3K-mTORC1 signaling in parvalbumin-positive interneurons in mouse infralimbic prefrontal cortex to exert acute antidepressant-lie effects.

The delta opioid receptor (DOP) is a promising target for novel antidepressants due to its potential for rapid action with minimal adverse effects; however, the functional mechanism underlying acute antidepressant actions remains elusive. We report that subcutaneous injection of the selective DOP agonist KNT-127 reduced immobility in the forced swimming test, and that this antidepressant-like response was reversed by intracerebroventricular injection of the selective mechanistic (or mammalian) target of rapamycin (mTOR) inhibitor rapamycin or the phosphatidylinositol-3 kinase (PI3K) inhibitor LY294002. KNT-127 also alleviated social avoidance and reduced sucrose consumption (anhedonia) among chronic vicarious social defeat stress model mice, which were similarly reversed by PI3K and mTOR inhibitors. In addition, KNT-127 increased phosphorylation levels of the mTOR signaling-related proteins Akt and p70S6 kinase in medial prefrontal cortex as revealed by immunoblotting. In the forced swimming test, a microinfusion of KNT-127 and another DOP agonist SNC80 in the infralimbic prefrontal cortex (IL-PFC) attenuated the immobility, which were blocked by rapamycin and LY294002. Perfusion of KNT-127 onto IL-PFC slices increased miniature excitatory postsynaptic current frequency and reduced miniature inhibitory postsynaptic current frequency in pyramidal neurons as measured by whole-cell patch-clamping, and both responses were reversed by rapamycin. Imaging of brain slices from transgenic mice with DOP-promoter-driven green fluorescent protein revealed that most DOPs were expressed in parvalbumin-positive interneurons in the IL-PFC. These findings suggest that DOP agonists exert antidepressant-like actions by suppressing GABA release from parvalbumin-positive interneurons via the PI3K-Akt-mTORC1-p70S6 kinase pathway, thereby enhancing IL-PFC pyramidal neuron excitation.

Micro-computed tomographic evaluation of dentinal microcracks in response to progressive taper enlargement in root canals of maxillary molars.

To investigate the development of dentinal microcracks resulting from the progressive enlargement of the buccal canals of maxillary molars with and without the MB2 canal employing instruments with the same tip and four different tapers. Twenty maxillary molars with (n = 10) and without (n = 10) the MB2 canal underwent micro-CT scanning. Their mesiobuccal and distobuccal canals were sequentially enlarged using nickel-titanium instruments with sizes 25/.03, 25/.05, 25/.06 and 25/.08v. Subsequent scans were conducted after each canal enlargement. The preoperative scans were coregistered with their respective datasets, and the cross-sectional images were carefully examined to identify dentinal defects. This process, conducted twice at two-week intervals, involved two examiners. Intra- and inter-examiner calibrations were validated using the Kappa index. The incidence of microcracks was reported as a percentage frequency. The intra- and inter-examiner Kappa values were 0.85 and 0.88, respectively, indicating excellent agreement. Dentinal microcracks were detected in 1206 out of 42 975 cross-sectional image slices (2.8%). Molars with MB2 canal exhibited more slices (4.5%) than teeth without MB2 (0.9%). Every dentinal microcrack observed in the images after root canal preparation was already present in the corresponding images takenbefore root canal preparation. The progressive enlargement of root canals using instruments with different tapers did not induce the formation of new dentinal microcracks in maxillary molars, regardless of the presence of the MB2 canal.

Depleting parenchymal border macrophages alleviates cerebral edema and neuroinflammation following status epilepticus

BackgroundStatus epilepticus (SE) is a common severe neurological emergency. Cerebral edema caused by SE is unavoidable and may exacerbate epilepsy. Recent studies have identified cerebrospinal fluid (CSF) as a crucial fluid source of initial cerebral edema following ischemic stroke and cardiac arrest. Moreover, synchronized neuronal firings drive CSF influx into interstitial fluid (ISF). Parenchymal border macrophages (PBMs) have been found to play a role in regulating CSF flow dynamics. However, the involvement of CSF and PBMs in cerebral edema during SE remains unclear. Here, we investigated the fluid source of cerebral edema in the initial phase of SE with the role of PBMs involved.MethodsLithium chloride-pilocarpine was used to induce SE in C57BL/6 J mice. Electroencephalogram (EEG) was acquired to assess changes in relative EEG power pre- and post-seizure onset. Apparent diffusion coefficient (ADC) maps reconstructed from diffusion-weighted imaging (DWI) were utilized to evaluate cytotoxic edema. Blood–brain barrier (BBB) permeability was examined using sodium fluorescein (NaFl). CSF tracer influx into the brain was assessed by transcranial imaging and brain slices. PBMs were depleted using clodronate liposomes. Immunohistochemistry was used to evaluate PBM depletion, severity of vasogenic edema, inflammation, and neuronal damage.ResultsDuring the initial stage of SE, relative EEG power sharply increased and ADC values significantly decreased. Concurrently, CSF tracer influx into the cortex significantly elevated, though NaFl leakage from blood to brain parenchyma did not evidently alter. Following depletion of PBM, CSF influx declined but AQP4 expression and polarization remained unaffected. Post-PBM depletion, there was no significant alteration in relative EEG power, yet CSF influx decreased substantially during the initial stage of SE. The degree of ADC decline lessened, IgG extravasation after SE decreased, activated microglia and proliferating astrocytes count fell, and neuronal damage post-SE alleviated.ConclusionsCSF appeared to contribute to cerebral edema in SE. Depletion of PBM alleviated cytotoxic edema in the initial phase of SE, and subsequent vasogenic edema, inflammatory response and neurological damage were reduced. These findings may provide potential novel strategies for treating cerebral edema following SE.

Medical Transformer: Universal Encoder for 3-D Brain MRI Analysis.

Transfer learning has attracted considerable attention in medical image analysis because of the limited number of annotated 3-D medical datasets available for training data-driven deep learning models in the real world. We propose Medical Transformer, a novel transfer learning framework that effectively models 3-D volumetric images as a sequence of 2-D image slices. To improve the high-level representation in 3-D-form empowering spatial relations, we use a multiview approach that leverages information from three planes of the 3-D volume, while providing parameter-efficient training. For building a source model generally applicable to various tasks, we pretrain the model using self-supervised learning (SSL) for masked encoding vector prediction as a proxy task, using a large-scale normal, healthy brain magnetic resonance imaging (MRI) dataset. Our pretrained model is evaluated on three downstream tasks: 1) brain disease diagnosis; 2) brain age prediction; and 3) brain tumor segmentation, which are widely studied in brain MRI research. Experimental results demonstrate that our Medical Transformer outperforms the state-of-the-art (SOTA) transfer learning methods, efficiently reducing the number of parameters by up to approximately 92% for classification and regression tasks and 97% for segmentation task, and it also achieves good performance in scenarios where only partial training samples are used.

Enhancing low radio-dosage thyroid imaging using deep learning and Monte Carlo-based absorbed dose measurement in nuclear medicine

The acquisition of improved quality via the administration of low radioactivity is an important challenge in the field of nuclear medicine. Thus, we evaluated the image quality generated under low radioactivity using deep-learning algorithms. In addition, internal dosimetry for various organs was measured based on Monte Carlo simulation. The dataset comprised 500 slices of thyroid nuclear medicine images acquired at 370 MBq as high radioactivity and 18.5 MBq as low radioactivity. The residual neural network (ResNet) with 16 residual blocks and a U-Net model were used with various hyperparameters for learning. In addition, we measured the internal dosimetry in peripheral organs based on a Monte Carlo simulation. The peak signal-to-noise ratio (PSNR) results were 18.24 for the ResNet model and 18.85 for the U-Net model on average. The structural similarity index measurement (SSIM) results were also 0.995 and 0.994 for the ResNet and U-Net models, respectively. The average absorbed dose for radionuclides with 37 MBq was 3.98 times higher than that of the radionuclide with 18.5 MBq. In conclusion, it is possible to make a diagnosis based on the administration of low radioactivity when scanning thyroid nuclear medicine images.