Optimal Limestone Content on Hydration Properties of Ordinary Portland Cement with 5% Ground Granulated Blast-Furnace Slag.

In this study, the effect of limestone content on the mechanical performance and the heat of hydration of ordinary Portland cement (OPC) was investigated. Changes in the phase assemblage were analyzed through XRD and thermodynamic modeling. The purpose of the study was to identify the optimal limestone content in OPC. As a result of the experiment, all samples were found to have equal fluidity. Increasing the limestone content accelerated the hydration of the cement before approximately 13 h and shortened the setting time due to the acceleration of the initial hydration reaction. The compressive strength of the cement mortar showed a dilution effect, with lower compressive strength compared to the reference sample at an early age, but it gradually recovered at a later age. This is because, as shown in the XRD and thermodynamic modeling results, the carboaluminate phases formed due to the chemical effect of limestone contributed to the development of compressive strength. As a result, within the scope of this study, it is believed that maintaining the limestone content in OPC within 10% is optimal to minimize quality degradation.

Extracellular C1qbp inhibits myogenesis by suppressing NFATc1.

Aging and lack of exercise are the most important etiological factors for muscle loss. We hypothesized that new factors that contribute to muscle loss could be identified from ones commonly altered in expression in aged and exercise-limited skeletal muscles. Mouse gastrocnemius muscles were subjected to mass spectrometry-based proteomic analysis. The muscle proteomes of hindlimb-unloaded and aged mice were compared to those of exercised and young mice, respectively. C1qbp expression was significantly upregulated in the muscles of both hindlimb-unloaded and aged mice. In vitro myogenic differentiation was not affected by altering intracellular C1qbp expression but was significantly suppressed upon recombinant C1qbp treatment. Additionally, recombinant C1qbp repressed the protein level but not the mRNA level of NFATc1. NFATc1 recruited the transcriptional coactivator p300, leading to the upregulation of acetylated histone H3 levels. Furthermore, NFATc1 silencing inhibited p300 recruitment, downregulated acetylated histone H3 levels, and consequently suppressed myogenic differentiation. The expression of C1qbp was inversely correlated with that of NFATc1 in the gastrocnemius muscles of exercised or hindlimb-unloaded, and young or aged mice. These findings demonstrate a novel role of extracellular C1qbp in suppressing myogenesis by inhibiting the NFATc1/p300 complex. Thus, C1qbp can serve as a novel therapeutic target for muscle loss.

Odontogenic cystic lesion segmentation on cone-beam CT using an auto-adapting multi-scaled UNet.

Objectives: Precise segmentation of Odontogenic Cystic Lesions (OCLs) from dental Cone-Beam Computed Tomography (CBCT) is critical for effective dental diagnosis. Although supervised learning methods have shown practical diagnostic results in segmenting various diseases, their ability to segment OCLs covering different sub-class varieties has not been extensively investigated. Methods: In this study, we propose a new supervised learning method termed OCL-Net that combines a Multi-Scaled U-Net model, along with an Auto-Adapting mechanism trained with a combined supervised loss. Anonymous CBCT images were collected retrospectively from one hospital. To assess the ability of our model to improve the diagnostic efficiency of maxillofacial surgeons, we conducted a diagnostic assessment where 7 clinicians were included to perform the diagnostic process with and without the assistance of auto-segmentation masks. Results: We collected 300 anonymous CBCT images which were manually annotated for segmentation masks. Extensive experiments demonstrate the effectiveness of our OCL-Net for CBCT OCLs segmentation, achieving an overall Dice score of 88.84%, an IoU score of 81.23%, and an AUC score of 92.37%. Through our diagnostic assessment, we found that when clinicians were assisted with segmentation labels from OCL-Net, their average diagnostic accuracy increased from 53.21% to 55.71%, while the average time spent significantly decreased from 101s to 47s (P<0.05). Conclusion: The findings demonstrate the potential of our approach as a robust auto-segmentation system on OCLs in CBCT images, while the segmented masks can be used to further improve OCLs dental diagnostic efficiency.

Bactericidal Effects and Quality Impact of Peroxyacetic Acid and Sodium Hypochlorite on Chicken Carcasses.

There is an urgent need to develop efficient and environmentally friendly decontaminants for poultry products. In this study, we aimed to evaluate the practical application of peroxyacetic acid (PAA) as a replacement for sodium hypochlorite (SH) to sterilize fresh chicken carcasses, using microbial, color, and electronic-nose analyses. We evaluated the decontamination effects of different concentrations of PAA and SH on chicken carcasses. The bactericidal effects of PAA at pH 3, 7, and 9, and SH at pH 10, at concentrations ranging from 100 to 500 ppm on coliform bacteria, total bacteria, and Salmonella spp. were evaluated. PAA induced a similar bactericidal effect at lower concentrations than SH. Therefore, at the same concentration and treatment time, PAA showed better bactericidal effects than SH. Although treatment with PAA (pH 3) and SH (pH 10) resulted in considerable discoloration, the degree of discoloration decreased when the pH of PAA was increased to 7 and 9. Therefore, by increasing the pH of PAA, the discoloration effect on chicken carcasses can be reduced without altering the microbial-reduction effect. Electronic-nose analysis showed that the flavor of the chicken was almost unaffected by volatile components at a treatment time < 30 min. Therefore, this study experimentally identified the optimal PAA concentration for the decontamination of chicken carcasses. The study findings provide a theoretical basis for the replacement of traditional bactericides, such as SH, with PAA for the production of poultry products.

Enzymatic extraction of anthocyanins from the pomace of aronia (Aronia melanocarpa).

Efficient extraction of natural pigments is a key focus in enhancing the utilization of by-products for applications in the food industry. In this study, an enzymatic extraction method using Pectinex Ultra SP-L, Pectinex XXL, Novoshape, and Celluclast was used to investigate natural pigment production from the pomace of aronia, a commercially important plant. The method's performance was monitored using high-performance liquid chromatography with diode-array detection by measuring total and individual anthocyanin levels. Pectinex XXL (0.5%) yielded the highest total anthocyanin extraction (2082.41 ± 85.69 mg/100 g) in the single enzyme treatment, followed by Pectinex Ultra SP-L (0.05%), Celluclast (0.01%), and Novoshape (0.1%). Combining Pectinex XXL (0.25%) with Celluclast (0.01%) increased the extraction ratio of total anthocyanins (2 323.04 ± 61.32 mg/100 g) by ∼50.7% compared with that obtained using the solvent extraction method. This study demonstrated an effective enzymatic extraction method for application in the food industry.

Age-related cholesterol and colorectal cancer progression: Validating squalene epoxidase for high-risk cases.

As people age, the risk and progression of colorectal cancer (CRC), along with cholesterol levels, tend to increase. Nevertheless, epidemiological studies on serum lipids and CRC have produced conflicting results. We previously demonstrated that the reduction of squalene epoxidase (SQLE) due to accumulated cholesterol within cells accelerates CRC progression through the activation of the ß-catenin pathway. This study aimed to investigate the mechanism by which age-related cholesterol accumulation within tissue accelerates CRC progression and to assess the clinical significance of SQLE in older individuals with elevated CRC risk. Using machine learning-based digital image analysis with fluorescence-immunohistochemistry, we assessed SQLE, GSK3ßpS9 (GSK3ß activity inhibition through serine 9 phosphorylation at GSK3ß), p53 wild-type (p53WT), and p53 mutant (p53MT) levels in CRC tissues. Our analysis revealed a significant reduction in SQLE, p53WT, and p53MT and increase in GSK3ßpS9 levels, all associated with the substantial accumulation of intra-tissue cholesterol in aged CRCs. Cox analysis underscored the significant influence of SQLE on overall survival and progression-free survival in grade 2-3 CRC patients aged over 50. SQLE and GSK3ßpS9 consistently exhibited outstanding prognostic and diagnostic performance, particularly in older individuals. Furthermore, combining SQLE with p53WT, p53MT, and GSK3ßpS9 demonstrated a robust diagnostic ability in the older population. In conclusion, we have identified that individuals aged over 50 face an increased risk of CRC progression due to aging-linked cholesterol accumulation within tissue and the subsequent reduction in SQLE levels. This study also provides valuable biomarkers, including SQLE and GSK3ßpS9, for older patients at elevated risk of CRC.

CM1, a Chrysin Derivative, Protects from Endotoxin-Induced Lethal Shock by Regulating the Excessive Activation of Inflammatory Responses.

Sepsis, a leading cause of death worldwide, is a harmful inflammatory condition that is primarily caused by an endotoxin released by Gram-negative bacteria. Effective targeted therapeutic strategies for sepsis are lacking. In this study, using an in vitro and in vivo mouse model, we demonstrated that CM1, a derivative of the natural polyphenol chrysin, exerts an anti-inflammatory effect by inducing the expression of the ubiquitin-editing protein TNFAIP3 and the NAD-dependent deacetylase sirtuin 1 (SIRT1). Interestingly, CM1 attenuated the Toll-like receptor 4 (TLR4)-induced production of inflammatory cytokines by inhibiting the extracellular-signal-regulated kinase (ERK)/MAPK and nuclear factor kappa B (NF-κB) signalling pathways. In addition, CM1 induced the expression of TNFAIP3 and SIRT1 on TLR4-stimulated primary macrophages; however, the anti-inflammatory effect of CM1 was abolished by the siRNA-mediated silencing of TNFAPI3 or by the genetic or pharmacologic inhibition of SIRT1. Importantly, intravenous administration of CM1 resulted in decreased susceptibility to endotoxin-induced sepsis, thereby attenuating the production of pro-inflammatory cytokines and neutrophil infiltration into the lung compared to control mice. Collectively, these findings demonstrate that CM1 has therapeutic potential for diverse inflammatory diseases, including sepsis.

Panoramic imaging errors in machine learning model development: a systematic review.

OBJECTIVES: To investigate the management of imaging errors from panoramic radiography (PAN) datasets used in the development of machine learning (ML) models. METHODS: This systematic literature followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses and used three databases. Keywords were selected from relevant literature. ELIGIBILITY CRITERIA: PAN studies that used ML models and mentioned image quality concerns. RESULTS: Out of 400 articles, 41 papers satisfied the inclusion criteria. All the studies used ML models, with 35 papers using deep learning (DL) models. PAN quality assessment was approached in 3 ways: acknowledgement and acceptance of imaging errors in the ML model, removal of low-quality radiographs from the dataset before building the model, and application of image enhancement methods prior to model development. The criteria for determining PAN image quality varied widely across studies and were prone to bias. CONCLUSIONS: This study revealed significant inconsistencies in the management of PAN imaging errors in ML research. However, most studies agree that such errors are detrimental when building ML models. More research is needed to understand the impact of low-quality inputs on model performance. Prospective studies may streamline image quality assessment by leveraging DL models, which excel at pattern recognition tasks.

BIDCell: Biologically-informed self-supervised learning for segmentation of subcellular spatial transcriptomics data.

Recent advances in subcellular imaging transcriptomics platforms have enabled high-resolution spatial mapping of gene expression, while also introducing significant analytical challenges in accurately identifying cells and assigning transcripts. Existing methods grapple with cell segmentation, frequently leading to fragmented cells or oversized cells that capture contaminated expression. To this end, we present BIDCell, a self-supervised deep learning-based framework with biologically-informed loss functions that learn relationships between spatially resolved gene expression and cell morphology. BIDCell incorporates cell-type data, including single-cell transcriptomics data from public repositories, with cell morphology information. Using a comprehensive evaluation framework consisting of metrics in five complementary categories for cell segmentation performance, we demonstrate that BIDCell outperforms other state-of-the-art methods according to many metrics across a variety of tissue types and technology platforms. Our findings underscore the potential of BIDCell to significantly enhance single-cell spatial expression analyses, enabling great potential in biological discovery.

Multi-Level Swin Transformer Enabled Automatic Segmentation and Classification of Breast Metastases.

Detection of metastatic breast cancer lesions is a challenging task in breast cancer treatment. The recent advancements in deep learning gained attention owing to its robustness, particularly in addressing automated segmentation and classification issues in medical images. In this paper, we proposed a modified Swin Transformer model (mST) integrated with a novel Multi-Level Adaptive Feature Fusion (MLAFF) Module. We constructed a modified Swin Transformer network comprising of a Local Transferable MSA (LT-MSA) and a Global Transferable MSA (GT-MSA) in addition to a Feed Forward Network (FFN). Our novel Multi-Level Adaptive Feature Fusion (MLAFF) module iteratively combines the features throughout multiple transformers. We utilized a pre-trained deep learning model U-Net and trained it on mammography utilizing Transfer Learning for automated segmentation. The proposed method, mST-MLAFF, is used for breast cancer classification into normal, benign, and malignant classes. Our model outperformed comparison methods based on U-Net and Swin Transformer in breast metastatic lesion segmentation on the seven benchmark datasets, namely INBreast, DDSM, MIAS, CBIS-DDSM, MIMBCD-UI, KAU-BCMD, and Mammographic Masses. Our model achieved 98% Dice-Similarity coefficient (DSC) for segmentation and an average of 94.5% accuracy for classification, whereas U-Net based model achieved 92% DSC and Swin Transformer achieved 93% DSC. Extensive performance evaluation of our model on benchmark datasets shows the potential of our model for breast cancer classification.Clinical relevance- This research work is focused on assisting the radiologist in the early detection and classification of breast cancer. A single mammography image is analyzed in less than a minute for automated segmentation and classification into malignant and benign classes.

A Deep-Learning Enabled Automatic Fetal Thalamus Diameter Measurement Algorithm.

The analysis of maternal factors that impact the normal development of the fetal thalamus is an emerging field of research and requires the retrospective measurement of fetal thalamus diameter (FTD). Unfortunately, FTD is not measured in routine 2D ultrasound (2D-US) screenings of fetuses. Manual measurement of FTD is a laborious, difficult, and error-prone process because the thalamus lacks well-defined boundaries in 2D-US images of the fetal brain as it has a similar echogenicity to the surrounding brain tissue. Traditional methods based on statistical shape models (SSMs) perform poorly in measuring FTD due to the noisy textures and fuzzy edges of the fetal thalamus in 2D-US images of the fetal brain. To overcome these difficulties, we propose a deep learning-based automatic FTD measurement algorithm, FTDNet. FTDNet measures FTD by learning to directly detect the measurement landmarks through supervised learning. The algorithm first detects the region of the brain that contains the thalamus structure, and then focuses on processing that region for FTD landmark detection. Our FTD dataset, developed through a consensus between two ultrasonographers, contains 1,111 pairs of landmark coordinates for measuring FTD and verified bounding boxes surrounding the fetal thalamus. To assess FTDNet's measurement consistency compared to the ground truth, we used the intraclass correlation coefficient (ICC). FTDNet achieved an ICC score of 0.734, significantly outperforming the prior SSM method and other baseline comparison methods. Our findings are an important step forward in understanding the maternal factors which influence fetal brain development.Clinical relevance- This work proposes an end-to-end thalamus detection and measurement algorithm for measuring fetal thalamus diameter. Our work represents a significant step in the research of how maternal factors can impact fetal thalamus development. The development of an automatic and accurate method for measuring FTD through deep learning has the potential to greatly advance this field of study.

Exploring the Performance of Geometry-Based Markerless Registration in a Simulated Surgical Environment: A Comparative Study of Registration Algorithms in Medical Augmented Reality.

Augmented Reality (AR) has been utilized in multiple applications in the medical field, such as augmenting Computed Tomography (CT) images onto the patient's body during surgery. However, one of the challenges in its utilization is to register the pre-operative CT images to the patient's body accurately. The current registration process requires prior attachment of tracking markers, and their localization within the body and CT images. This process can be cumbersome, error-prone, and dependent on the surgeon's experience. Moreover, there are cases where medical instruments, drapes, or the body may occlude the markers. In light of these limitations, markerless registration algorithms have the potential to aid the registration process in the clinical setting. While those algorithms have been successfully used in other sectors, such as multimedia, they have not yet been thoroughly investigated in a clinical setting, especially in surgery, where there are more challenging cases with different positions of the patients in the image and the surgical environment. In this paper, we benchmarked and evaluated the performance of 6 state-of-the-art markerless registration algorithms from the multimedia sector by registering a CT image onto the whole-body phantom dataset acquired from a simulated surgical environment. We also analyzed the suitability of these algorithms for use in the surgical setting and discussed their potential for the advancement of AR-assisted surgery.Clinical Relevance-Our study provides insight into the potential of AR-assisted surgery and helps practitioners in choosing the most suitable registration algorithm for their needs to improve patient outcomes, reduce the risk of surgical errors and shorten the time of preoperative planning.

SSPT-bpMRI: A Self-supervised Pre-training Scheme for Improving Prostate Cancer Detection and Diagnosis in Bi-parametric MRI.

Prostate cancer (PCa) is one of the most prevalent cancers in men. Early diagnosis plays a pivotal role in reducing the mortality rate from clinically significant PCa (csPCa). In recent years, bi-parametric magnetic resonance imaging (bpMRI) has attracted great attention for the detection and diagnosis of csPCa. bpMRI is able to overcome some limitations of multi-parametric MRI (mpMRI) such as the use of contrast agents, the time-consuming for imaging and the costs, and achieve detection performance comparable to mpMRI. However, inter-reader agreements are currently low for prostate MRI. Advancements in artificial intelligence (AI) have propelled the development of deep learning (DL)-based computer-aided detection and diagnosis system (CAD). However, most of the existing DL models developed for csPCa identification are restricted by the scale of data and the scarcity in labels. In this paper, we propose a self-supervised pre-training scheme named SSPT-bpMRI with an image restoration pretext task integrating four different image transformations to improve the performance of DL algorithms. Specially, we explored the potential value of the self-supervised pre-training in fully supervised and weakly supervised situations. Experiments on the publicly available PI-CAI dataset demonstrate that our model outperforms the fully supervised or weakly supervised model alone.

Hyper-Connected Transformer Network for Multi-Modality PET-CT Segmentation.

[18F]-Fluorodeoxyglucose (FDG) positron emission tomography - computed tomography (PET-CT) has become the imaging modality of choice for diagnosing many cancers. Co-learning complementary PET-CT imaging features is a fundamental requirement for automatic tumor segmentation and for developing computer aided cancer diagnosis systems. In this study, we propose a hyper-connected transformer (HCT) network that integrates a transformer network (TN) with a hyper connected fusion for multi-modality PET-CT images. The TN was leveraged for its ability to provide global dependencies in image feature learning, which was achieved by using image patch embeddings with a self-attention mechanism to capture image-wide contextual information. We extended the single-modality definition of TN with multiple TN based branches to separately extract image features. We also introduced a hyper connected fusion to fuse the contextual and complementary image features across multiple transformers in an iterative manner. Our results with two clinical datasets show that HCT achieved better performance in segmentation accuracy when compared to the existing methods.Clinical Relevance-We anticipate that our approach can be an effective and supportive tool to aid physicians in tumor quantification and in identifying image biomarkers for cancer treatment.

CG-3DSRGAN: A classification guided 3D generative adversarial network for image quality recovery from low-dose PET images.

Positron emission tomography (PET) is the most sensitive molecular imaging modality routinely applied in our modern healthcare. High radioactivity caused by the injected tracer dose is a major concern in PET imaging and limits its clinical applications. However, reducing the dose leads to inadequate image quality for diagnostic practice. Motivated by the need to produce high quality images with minimum 'low-dose', convolutional neural networks (CNNs) based methods have been developed for high quality PET synthesis from its low-dose counterparts. Previous CNNs-based studies usually directly map low-dose PET into features space without consideration of different dose reduction level. In this study, a novel approach named CG-3DSRGAN (Classification-Guided Generative Adversarial Network with Super Resolution Refinement) is presented. Specifically, a multi-tasking coarse generator, guided by a classification head, allows for a more comprehensive understanding of the noise-level features present in the low-dose data, resulting in improved image synthesis. Moreover, to recover spatial details of standard PET, an auxiliary super resolution network - Contextual-Net - is proposed as a second-stage training to narrow the gap between coarse prediction and standard PET. We compared our method to the state-of-the-art methods on whole-body PET with different dose reduction factors (DRF). Experiments demonstrate our method can outperform others on all DRF.Clinical Relevance- Low-Dose PET, PET recovery, GAN, task driven image synthesis, super resolution.

Diagnosis of Crohn's disease and ulcerative colitis using the microbiome.

BACKGROUND: Inflammatory bowel disease (IBD) is a multifactorial chronic inflammatory disease resulting from dysregulation of the mucosal immune response and gut microbiota. Crohn's disease (CD) and ulcerative colitis (UC) are difficult to distinguish, and differential diagnosis is essential for establishing a long-term treatment plan for patients. Furthermore, the abundance of mucosal bacteria is associated with the severity of the disease. This study aimed to differentiate and diagnose these two diseases using the microbiome and identify specific biomarkers associated with disease activity. RESULTS: Differences in the abundance and composition of the microbiome between IBD patients and healthy controls (HC) were observed. Compared to HC, the diversity of the gut microbiome in patients with IBD decreased; the diversity of the gut microbiome in patients with CD was significantly lower. Sixty-eight microbiota members (28 for CD and 40 for UC) associated with these diseases were identified. Additionally, as the disease progressed through different stages, the diversity of the bacteria decreased. The abundances of Alistipes shahii and Pseudodesulfovibrio aespoeensis were negatively correlated with the severity of CD, whereas the abundance of Polynucleobacter wianus was positively correlated. The severity of UC was negatively correlated with the abundance of A. shahii, Porphyromonas asaccharolytica and Akkermansia muciniphilla, while it was positively correlated with the abundance of Pantoea candidatus pantoea carbekii. A regularized logistic regression model was used for the differential diagnosis of the two diseases. The area under the curve (AUC) was used to examine the performance of the model. The model discriminated UC and CD at an AUC of 0.873 (train set), 0.778 (test set), and 0.633 (validation set) and an area under the precision-recall curve (PRAUC) of 0.888 (train set), 0.806 (test set), and 0.474 (validation set). CONCLUSIONS: Based on fecal whole-metagenome shotgun (WMS) sequencing, CD and UC were diagnosed using a machine-learning predictive model. Microbiome biomarkers associated with disease activity (UC and CD) are also proposed.

Experimental Validation of the Cementation Mechanism of Wood Pellet Fly Ash Blended Binder in Weathered Granite Soil.

In response to climate change, wood pellets have been increasingly utilized as a sustainable energy source. However, their growing utilization increases the production of wood pellet fly ash (WA) by-products, necessitating alternative recycling technologies due to a shortage of discharging landfills. Thus, this research seeks to utilize WA by developing a new sustainable construction material, called wood pellet fly ash blended binder (WABB), and to validate its stabilizing performance in natural soils, namely weathered granite soil (WS). WABB is made from 50% WA, 30% ground granulated blast-furnace slag (GGBS), and 20% cement by dry mass. WS was mixed with 5%, 15%, and 25% WABB and was tested for a series of unconfined compressive strength (qu), pH, and suction tests at 3, 7, 14, and 28 days. For the microstructural analyses, XRD, SEM, and EDS were employed. As the WABB dosage rate increased, the average qu increased by 1.88 to 11.77, which was higher than that of compacted WS without any binder. Newly cementitious minerals were also confirmed. These results suggest that the effects of the combined hydration mechanism of WABB are due to cement's role in facilitating early strength development, GGBS's latent hydraulic properties, and WA's capacity to stimulate the alkaline components of WABB and soil grains. Thus, this research validates a new sustainable binder, WABB, as a potential alternative to conventional soil stabilizers.

Multi-task deep learning-based radiomic nomogram for prognostic prediction in locoregionally advanced nasopharyngeal carcinoma.

PURPOSE: Prognostic prediction is crucial to guide individual treatment for locoregionally advanced nasopharyngeal carcinoma (LA-NPC) patients. Recently, multi-task deep learning was explored for joint prognostic prediction and tumor segmentation in various cancers, resulting in promising performance. This study aims to evaluate the clinical value of multi-task deep learning for prognostic prediction in LA-NPC patients. METHODS: A total of 886 LA-NPC patients acquired from two medical centers were enrolled including clinical data, [18F]FDG PET/CT images, and follow-up of progression-free survival (PFS). We adopted a deep multi-task survival model (DeepMTS) to jointly perform prognostic prediction (DeepMTS-Score) and tumor segmentation from FDG-PET/CT images. The DeepMTS-derived segmentation masks were leveraged to extract handcrafted radiomics features, which were also used for prognostic prediction (AutoRadio-Score). Finally, we developed a multi-task deep learning-based radiomic (MTDLR) nomogram by integrating DeepMTS-Score, AutoRadio-Score, and clinical data. Harrell's concordance indices (C-index) and time-independent receiver operating characteristic (ROC) analysis were used to evaluate the discriminative ability of the proposed MTDLR nomogram. For patient stratification, the PFS rates of high- and low-risk patients were calculated using Kaplan-Meier method and compared with the observed PFS probability. RESULTS: Our MTDLR nomogram achieved C-index of 0.818 (95% confidence interval (CI): 0.785-0.851), 0.752 (95% CI: 0.638-0.865), and 0.717 (95% CI: 0.641-0.793) and area under curve (AUC) of 0.859 (95% CI: 0.822-0.895), 0.769 (95% CI: 0.642-0.896), and 0.730 (95% CI: 0.634-0.826) in the training, internal validation, and external validation cohorts, which showed a statistically significant improvement over conventional radiomic nomograms. Our nomogram also divided patients into significantly different high- and low-risk groups. CONCLUSION: Our study demonstrated that MTDLR nomogram can perform reliable and accurate prognostic prediction in LA-NPC patients, and also enabled better patient stratification, which could facilitate personalized treatment planning.

CT-based Radiogenomics Framework for COVID-19 Using ACE2 Imaging Representations.

Coronavirus disease 2019 (COVID-19) is caused by Severe Acute Respiratory Syndrome Coronavirus 2 which enters the body via the angiotensin-converting enzyme 2 (ACE2) and altering its gene expression. Altered ACE2 plays a crucial role in the pathogenesis of COVID-19. Gene expression profiling, however, is invasive and costly, and is not routinely performed. In contrast, medical imaging such as computed tomography (CT) captures imaging features that depict abnormalities, and it is widely available. Computerized quantification of image features has enabled 'radiogenomics', a research discipline that identifies image features that are associated with molecular characteristics. Radiogenomics between ACE2 and COVID-19 has yet to be done primarily due to the lack of ACE2 expression data among COVID-19 patients. Similar to COVID-19, patients with lung adenocarcinoma (LUAD) exhibit altered ACE2 expression and, LUAD data are abundant. We present a radiogenomics framework to derive image features (ACE2-RGF) associated with ACE2 expression data from LUAD. The ACE2-RGF was then used as a surrogate biomarker for ACE2 expression. We adopted conventional feature selection techniques including ElasticNet and LASSO. Our results show that: i) the ACE2-RGF encoded a distinct collection of image features when compared to conventional techniques, ii) the ACE2-RGF can classify COVID-19 from normal subjects with a comparable performance to conventional feature selection techniques with an AUC of 0.92, iii) ACE2-RGF can effectively identify patients with critical illness with an AUC of 0.85. These findings provide unique insights for automated COVID-19 analysis and future research.

K-PathVQA: Knowledge-Aware Multimodal Representation for Pathology Visual Question Answering.

Pathology imaging is routinely used to detect the underlying effects and causes of diseases or injuries. Pathology visual question answering (PathVQA) aims to enable computers to answer questions about clinical visual findings from pathology images. Prior work on PathVQA has focused on directly analyzing the image content using conventional pretrained encoders without utilizing relevant external information when the image content is inadequate. In this paper, we present a knowledge-driven PathVQA (K-PathVQA), which uses a medical knowledge graph (KG) from a complementary external structured knowledge base to infer answers for the PathVQA task. K-PathVQA improves the question representation with external medical knowledge and then aggregates vision, language, and knowledge embeddings to learn a joint knowledge-image-question representation. Our experiments using a publicly available PathVQA dataset showed that our K-PathVQA outperformed the best baseline method with an increase of 4.15% in accuracy for the overall task, an increase of 4.40% in open-ended question type and an absolute increase of 1.03% in closed-ended question types. Ablation testing shows the impact of each of the contributions. Generalizability of the method is demonstrated with a separate medical VQA dataset.