Automated breast imaging report generation based on the integration of multiple image features in a metadata format for shared decision-making.

Importance: Medical imaging increases the workload involved in writing reports. Given the lack of a standardized format for reports, reports are not easily used as communication tools. Objective: During medical team-patient communication, the descriptions in reports also need to be understood. Automatically generated imaging reports with rich and understandable information can improve medical quality. Design, setting, and participants: The image analysis theory of Panofsky and Shatford from the perspective of image metadata was used in this study to establish a medical image interpretation template (MIIT) for automated image report generation. Main outcomes and measures: The image information included digital imaging and communications in medicine (DICOM), reporting and data systems (RADSs), and image features used in computer-aided diagnosis (CAD). The utility of the images was evaluated by a questionnaire survey to determine whether the image content could be better understood. Results: In 100 responses, exploratory factor analysis revealed that the factor loadings of the facets were greater than 0.5, indicating construct validity, and the overall Cronbach's alpha was 0.916, indicating reliability. No significant differences were noted according to sex, age or education. Conclusions and relevance: Overall, the results show that MIIT is helpful for understanding the content of medical images.

Septic Arthritis Modeling Using Sonographic Fusion with Attention and Selective Transformation: a Preliminary Study.

Conventionally diagnosing septic arthritis relies on detecting the causal pathogens in samples of synovial fluid, synovium, or blood. However, isolating these pathogens through cultures takes several days, thus delaying both diagnosis and treatment. Establishing a quantitative classification model from ultrasound images for rapid septic arthritis diagnosis is mandatory. For the study, a database composed of 342 images of non-septic arthritis and 168 images of septic arthritis produced by grayscale (GS) and power Doppler (PD) ultrasound was constructed. In the proposed architecture of fusion with attention and selective transformation (FAST), both groups of images were combined in a vision transformer (ViT) with the convolutional block attention module, which incorporates spatial, modality, and channel features. Fivefold cross-validation was applied to evaluate the generalized ability. The FAST architecture achieved the accuracy, sensitivity, specificity, and area under the curve (AUC) of 86.33%, 80.66%, 90.25%, and 0.92, respectively. These performances were higher than using conventional ViT (82.14%) and significantly better than using one modality alone (GS 73.88%, PD 72.02%), with the p-value being less than 0.01. Through the integration of multi-modality and the extraction of multiple channel features, the established model provided promising accuracy and AUC in septic arthritis classification. The end-to-end learning of ultrasound features can provide both rapid and objective assessment suggestions for future clinic use.

LYNSU: automated 3D neuropil segmentation of fluorescent images for Drosophila brains.

The brain atlas, which provides information about the distribution of genes, proteins, neurons, or anatomical regions, plays a crucial role in contemporary neuroscience research. To analyze the spatial distribution of those substances based on images from different brain samples, we often need to warp and register individual brain images to a standard brain template. However, the process of warping and registration may lead to spatial errors, thereby severely reducing the accuracy of the analysis. To address this issue, we develop an automated method for segmenting neuropils in the Drosophila brain for fluorescence images from the FlyCircuit database. This technique allows future brain atlas studies to be conducted accurately at the individual level without warping and aligning to a standard brain template. Our method, LYNSU (Locating by YOLO and Segmenting by U-Net), consists of two stages. In the first stage, we use the YOLOv7 model to quickly locate neuropils and rapidly extract small-scale 3D images as input for the second stage model. This stage achieves a 99.4% accuracy rate in neuropil localization. In the second stage, we employ the 3D U-Net model to segment neuropils. LYNSU can achieve high accuracy in segmentation using a small training set consisting of images from merely 16 brains. We demonstrate LYNSU on six distinct neuropils or structures, achieving a high segmentation accuracy comparable to professional manual annotations with a 3D Intersection-over-Union (IoU) reaching up to 0.869. Our method takes only about 7 s to segment a neuropil while achieving a similar level of performance as the human annotators. To demonstrate a use case of LYNSU, we applied it to all female Drosophila brains from the FlyCircuit database to investigate the asymmetry of the mushroom bodies (MBs), the learning center of fruit flies. We used LYNSU to segment bilateral MBs and compare the volumes between left and right for each individual. Notably, of 8,703 valid brain samples, 10.14% showed bilateral volume differences that exceeded 10%. The study demonstrated the potential of the proposed method in high-throughput anatomical analysis and connectomics construction of the Drosophila brain.

The cerebellum computes frequency dynamics for motions with numerical precision and cross-individual uniformity.

Cross-individual variability is considered the essence of biology, preventing precise mathematical descriptions of biological motion1-7 like the physics law of motion. Here we report that the cerebellum shapes motor kinematics by encoding dynamic motor frequencies with remarkable numerical precision and cross-individual uniformity. Using in-vivo electrophysiology and optogenetics in mice, we confirmed that deep cerebellar neurons encoded frequencies via populational tuning of neuronal firing probabilities, creating cerebellar oscillations and motions with matched frequencies. The mechanism was consistently presented in self-generated rhythmic and non-rhythmic motions triggered by a vibrational platform, or skilled tongue movements of licking in all tested mice with cross-individual uniformity. The precision and uniformity allowed us to engineer complex motor kinematics with designed frequencies. We further validated the frequency-coding function of the human cerebellum using cerebellar electroencephalography recordings and alternating-current stimulation during voluntary tapping tasks. Our findings reveal a cerebellar algorithm for motor kinematics with precision and uniformity, the mathematical foundation for brain-computer interface for motor control.

Stereotactic Body Radiotherapy and Liver Transplant for Liver Cancer: A Nonrandomized Controlled Trial.

Importance: Whether stereotactic body radiotherapy (SBRT) as a bridge to liver transplant for hepatocellular carcinoma (HCC) is effective and safe is still unknown. Objective: To investigate the feasibility of SBRT before deceased donor liver transplant (DDLT) for previously untreated unresectable HCC. Design, Setting, and Participants: In this phase 2 nonrandomized controlled trial conducted between June 1, 2015, and October 18, 2019, 32 eligible patients within UCSF (University of California, San Francisco) criteria underwent dual-tracer (18F-fluorodeoxyglucose and 11C-acetate [ACC]) positron emission tomography with computed tomography (PET-CT) and magnetic resonance imaging (MRI) with gadoxetate followed by SBRT of 35 to 50 Gy in 5 fractions, and the same imaging afterward while awaiting DDLT. Statistical analysis was performed on an intention-to-treat basis between October 1 and 31, 2023. Intervention: Patients received SBRT followed by DDLT when matched deceased donor grafts were available. Main Outcomes and Measures: Coprimary end points were progression-free survival (PFS) and objective response rates (ORRs) by the Response Evaluation Criteria in Solid Tumors, version 1.1 (RECIST 1.1), modified RECIST (mRECIST), and PET Response Criteria in Solid Tumors (PERCIST). Secondary end points were local control rate, overall survival (OS), and safety. Results: A total of 32 patients (median age, 59 years [IQR, 54-63 years]; 22 men [68.8%]) with 56 lesions received SBRT. After a median follow-up of 74.6 months (IQR, 40.1-102.9 months), the median PFS was 17.6 months (95% CI, 6.6-28.6 months), and the median OS was 60.5 months (95% CI, 29.7-91.2 months). The 5-year PFS was 39.9% (95% CI, 19.9%-59.9%), and the 5-year OS was 51.3% (95% CI, 31.7%-70.9%). In terms of number of patients, ORRs were 62.5% ([n = 20] 95% CI, 54.2%-68.7%) by RECIST 1.1, 71.9% ([n = 23] 95% CI, 63.7%-79.0%) by mRECIST, and 78.1% ([n = 25] 95% CI, 73.2%-86.7%) by PERCIST. In terms of number of lesions, ORRs were 75.0% ([n = 42] 95% CI, 61.6%-80.8%) by RECIST 1.1, 83.9% ([n = 47] 95% CI, 74.7%-90.6%) by mRECIST, and 87.5% ([n = 49] 95% CI, 81.3%-98.6%) by PERCIST. Twenty patients with 36 lesions received DDLT, of whom 15 patients (75.0%) with 21 lesions (58.3%) exhibited pathologic complete response. Multivariable analyses revealed that pretreatment metabolic tumor volume (MTV) based on ACC (hazard ratio [HR], 1.06 [95% CI, 1.01-1.10]; P = .01) and complete metabolic response (CMR) by PERCIST (HR, 0.31 [95% CI, 0.10-0.96]; P = .04) were associated with PFS, while pretreatment MTV based on ACC (HR, 1.07 [95% CI, 1.03-1.16]; P = .01), total lesion activity based on ACC (HR, 1.01 [95% CI, 1.00-1.02]; P = .02), and CMR by PERCIST (HR, 0.21 [95% CI, 0.07-0.73]; P = .01) were associated with OS. Toxic effects associated with SBRT were reported for 9 patients (28.1%), with 1 grade 3 event. Conclusions and Relevance: This phase 2 nonrandomized controlled trial demonstrated promising survival and safety outcomes of SBRT before DDLT for unresectable HCC. Future randomized clinical trials are warranted.

Correlation of pathological examination with indocyanine green (ICG) intensity gradients: a prospective study in patients with liver tumor.

BACKGROUND: Intraoperative indocyanine green (ICG) fluorescence imaging has been shown to be a new and innovative way to illustrate the optimal resection margin in hepatectomy for hepatocellular carcinoma. This study investigated its accuracy in resection margin determination by looking into the correlation of ICG intensity gradients with pathological examination results of resected specimens. METHODS: This was a prospective, single-center, non-randomized controlled study. Patients who had liver tumors indicating liver resection were recruited. The hypothesis was that the use of intraoperative near-infrared/ICG fluorescence imaging would be a promising guiding tool for removing hepatocellular carcinoma with a better resection margin. Patients were given ICG (0.25 mg/kg) 1 day before operation. Resected specimens were inspected under a fluorescent imaging system. Biopsies were taken from tumors and normal tissue. Color signals obtained from ICG fluorescence imaging were compared with biopsies for analysis. RESULTS: Twenty-two patients were recruited for study. The median size of their tumors was 2.25 cm. One patient had resection margin involvement. Under ICG fluorescence, the tumors typically lighted up as yellow color, wrapped by a zone of green color. Tumors of 17 patients (77.3%) displayed yellow color and were confirmed malignancy, while tumors of 12 patients (54.5%) displayed green color and were confirmed malignancy. Receiver operating characteristic curve was used to measure the sensitivity and specificity of the green color to look for a clear resection margin. The area under the curve was 85.3% (p = 0.019, 95% confidence interval 0.696-1.000), with a sensitivity of 0.706 and specificity of 1.000. CONCLUSION: The use of ICG fluorescence can be helpful in determining resection margins. Resection of tumor should include complete resection of the green zone shown in the fluorescence image.

Fusion of memristor and digital compute-in-memory processing for energy-efficient edge computing.

Artificial intelligence (AI) edge devices prefer employing high-capacity nonvolatile compute-in-memory (CIM) to achieve high energy efficiency and rapid wakeup-to-response with sufficient accuracy. Most previous works are based on either memristor-based CIMs, which suffer from accuracy loss and do not support training as a result of limited endurance, or digital static random-access memory (SRAM)-based CIMs, which suffer from large area requirements and volatile storage. We report an AI edge processor that uses a memristor-SRAM CIM-fusion scheme to simultaneously exploit the high accuracy of the digital SRAM CIM and the high energy-efficiency and storage density of the resistive random-access memory memristor CIM. This also enables adaptive local training to accommodate personalized characterization and user environment. The fusion processor achieved high CIM capacity, short wakeup-to-response latency (392 microseconds), high peak energy efficiency (77.64 teraoperations per second per watt), and robust accuracy (<0.5% accuracy loss). This work demonstrates that memristor technology has moved beyond in-lab development stages and now has manufacturability for AI edge processors.

Stage-by-stage analysis of the effect of blood transfusion on survival after curative hepatectomy for hepatocellular carcinoma-a retrospective study.

OBJECTIVE: This study is to examine the impact of perioperative (intraoperative/postoperative) blood transfusion on the outcomes of curative hepatectomy for hepatocellular carcinoma. Hepatectomy is a well-established curative treatment for hepatocellular carcinoma, and blood transfusion cannot always be avoided in treating the disease. METHODS: A retrospective study of patients having curative hepatectomy for hepatocellular carcinoma from January 2010 to December 2019 at a single center was conducted. The patients were stratified by their disease stage. Patients with and without perioperative blood transfusion were matched by propensity-score matching and compared for each disease stage. Univariate and multivariate analyses were performed to identify prognostic factors for overall survival for each stage. RESULTS: A total of 846 patients were studied. Among them, 125 received perioperative blood transfusion and 720 did not. Patients with blood transfusion had worse disease-free and overall survival. After stratification and matching, the ratios of transfusion to non-transfusion were 33:165 (stage 1), 28:140 (stage 2), and 45:90 (stage 3). Perioperative blood transfusion was associated with a higher incidence of postoperative complications in all three disease stages (p = 0.004/0.006/0.017), and hence longer hospitalization (p < 0.001 in all stages), but had no significant impact on hospital mortality (p = 0.119/0.118/0.723), 90-day mortality (p = 0.259/0.118/0.723), disease-free survival (p = 0.128/0.826/0.511), or overall survival (p = 0.869/0.122/0.122) in any disease stage. Prognostic factors for overall survival included tumor size, tumor number, alpha-fetoprotein level, and postoperative complication of grade ≥ 3A. CONCLUSION: Perioperative blood transfusion was associated with a higher incidence of complications but had no significant impact on survival after curative hepatectomy for hepatocellular carcinoma.

Biliary complications after adult-to-adult living-donor liver transplantation: An international multicenter study of 3633 cases.

In living-donor liver transplantation, biliary complications including bile leaks and biliary anastomotic strictures remain significant challenges, with incidences varying across different centers. This multicentric retrospective study (2016-2020) included 3633 adult patients from 18 centers and aimed to identify risk factors for these biliary complications and their impact on patient survival. Incidences of bile leaks and biliary strictures were 11.4% and 20.6%, respectively. Key risk factors for bile leaks included multiple bile duct anastomoses (odds ratio, [OR] 1.8), Roux-en-Y hepaticojejunostomy (OR, 1.4), and a history of major abdominal surgery (OR, 1.4). For biliary anastomotic strictures, risk factors were ABO incompatibility (OR, 1.4), blood loss >1 L (OR, 1.4), and previous abdominal surgery (OR, 1.7). Patients experiencing biliary complications had extended hospital stays, increased incidence of major complications, and higher comprehensive complication index scores. The impact on graft survival became evident after accounting for immortal time bias using time-dependent covariate survival analysis. Bile leaks and biliary anastomotic strictures were associated with adjusted hazard ratios of 1.7 and 1.8 for graft survival, respectively. The study underscores the importance of minimizing these risks through careful donor selection and preoperative planning, as biliary complications significantly affect graft survival, despite the availability of effective treatments.

COVID-19 control measures unexpectedly increased the duration of stay at High Speed Rail stations during the first community outbreak in Taiwan.

During the COVID-19 pandemic, Taiwan has implemented strict border controls and community spread prevention measures. As part of these efforts, the government also implemented measures for public transportation. In Taiwan, there are two primary public transportation systems: Taiwan Railways (TR) is commonly utilized for local travel, while the Taiwan High-Speed Rail (THSR) is preferred for business trips and long-distance journeys due to its higher speed. In this study, we examined the impact of these disease prevention measures on the number of passengers and duration of stay in two major public transportation systems during the first community outbreak from April 29th to May 29th, 2021. Using data from a local telecommunications company, our study observed an expected decrease in the number of passengers after the cancellation of non-reserved seats at both TR and THSR stations across all 19 cities in the main island of Taiwan. Surprisingly, however, the duration of stay in some of the cities unexpectedly increased, especially at THSR stations. This unanticipated rise in the duration of stay has the potential to elevate contact probability among passengers and, consequently, the transmission rate. Our analysis shows that intervention policies may result in unforeseen outcomes, highlighting the crucial role of human mobility data as a real-time reference for policymakers. It enables them to monitor the impact of disease prevention measures and facilitates informed, data-driven decision-making.

Efficacy of outer membrane permeabilization in promoting aromatic isothiocyanates-mediated eradication of multidrug resistant Gram-negative bacteria and bacterial persisters.

Multidrug resistant (MDR) bacteria are recognized to be one of the most important problems in public health. The outer membrane permeability is a critical intrinsic mechanism of bacterial resistance. In addition, bacteria produce a small number of dormant persister cells causing multidrug tolerance that reduces antimicrobial efficacy. This study aimed to evaluate the inhibitory effects of the combination of aromatic isothiocyanates (ITCs) with membrane-active agents on bacterial persisters and MDR Gram-negative bacteria. Our study demonstrated that membrane-active agents, particularly ethylenediaminetetraacetic acid (EDTA) synergistically enhanced the inhibitory activity of aromatic benzyl ITC and phenethyl ITC against most Gram-negative bacteria strains with fractional inhibitory concentration index values ranging from 0.18 to 0.5 and 0.16 to 0.5, respectively, and contributed to an 8- to 64-fold minimal inhibitory concentration reduction compared with those of aromatic ITCs alone. The EDTA-aromatic ITCs combination effectively reduced the survival rates of tested bacteria and significantly eradicated bacterial persisters (p = 0.033 and 0.037, respectively). The growth kinetics analysis also supported the enhanced inhibitory effect of EDTA-aromatic ITCs combination against tested bacteria. Our results suggested an alternate treatment strategy against Gram-negative bacteria, promoting the entry of aromatic ITCs into bacterial cytoplasm to facilitate bacterial clearance and thus preventing the development of bacterial resistance.

Detecting microsatellite instability in colorectal cancer using Transformer-based colonoscopy image classification and retrieval.

Colorectal cancer (CRC) is a major global health concern, with microsatellite instability-high (MSI-H) being a defining characteristic of hereditary nonpolyposis colorectal cancer syndrome and affecting 15% of sporadic CRCs. Tumors with MSI-H have unique features and better prognosis compared to MSI-L and microsatellite stable (MSS) tumors. This study proposed establishing a MSI prediction model using more available and low-cost colonoscopy images instead of histopathology. The experiment utilized a database of 427 MSI-H and 1590 MSS colonoscopy images and vision Transformer (ViT) with different feature training approaches to establish the MSI prediction model. The accuracy of combining pre-trained ViT features was 84% with an area under the receiver operating characteristic curve of 0.86, which was better than that of DenseNet201 (80%, 0.80) in the experiment with support vector machine. The content-based image retrieval (CBIR) approach showed that ViT features can obtain a mean average precision of 0.81 compared to 0.79 of DenseNet201. ViT reduced the issues that occur in convolutional neural networks, including limited receptive field and gradient disappearance, and may be better at interpreting diagnostic information around tumors and surrounding tissues. By using CBIR, the presentation of similar images with the same MSI status would provide more convincing deep learning suggestions for clinical use.

Interactive content-based image retrieval with deep learning for CT abdominal organ recognition.

Objective.Recognizing the most relevant seven organs in an abdominal computed tomography (CT) slice requires sophisticated knowledge. This study proposed automatically extracting relevant features and applying them in a content-based image retrieval (CBIR) system to provide similar evidence for clinical use.Approach.A total of 2827 abdominal CT slices, including 638 liver, 450 stomach, 229 pancreas, 442 spleen, 362 right kidney, 424 left kidney and 282 gallbladder tissues, were collected to evaluate the proposed CBIR in the present study. Upon fine-tuning, high-level features used to automatically interpret the differences among the seven organs were extracted via deep learning architectures, including DenseNet, Vision Transformer (ViT), and Swin Transformer v2 (SwinViT). Three images with different annotations were employed in the classification and query.Main results.The resulting performances included the classification accuracy (94%-99%) and retrieval result (0.98-0.99). Considering global features and multiple resolutions, SwinViT performed better than ViT. ViT also benefited from a better receptive field to outperform DenseNet. Additionally, the use of hole images can obtain almost perfect results regardless of which deep learning architectures are used.Significance.The experiment showed that using pretrained deep learning architectures and fine-tuning with enough data can achieve successful recognition of seven abdominal organs. The CBIR system can provide more convincing evidence for recognizing abdominal organs via similarity measurements, which could lead to additional possibilities in clinical practice.

Indications of pro-inflammatory cytokines in laparoscopic and open liver resection for early-stage hepatocellular carcinoma.

BACKGROUND: Our clinical practice of laparoscopic liver resection (LLR) had achieved better short-term and long-term benefits for patients with hepatocellular carcinoma (HCC) over open liver resection (OLR), but the underlying mechanisms are not clear. This study was to find out whether systemic inflammation plays an important role. METHODS: A total of 103 patients with early-stage HCC under liver resection were enrolled (LLR group, n = 53; OLR group, n = 50). The expression of 9 inflammatory cytokines in patients at preoperation, postoperative day 1 (POD1) and POD7 was quantified by Luminex Multiplex assay. The relationships of the cytokines and the postoperative outcomes were compared between LLR and OLR. RESULTS: Seven of the circulating cytokines were found to be significantly upregulated on POD1 after LLR or OLR compared to their preoperative levels. Compared to OLR, the POD1 levels of granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-6 (IL-6), IL-8, and monocyte chemoattractant protein-1 (MCP-1) in the LLR group were significantly lower. Higher POD1 levels of these cytokines were significantly correlated with longer operative time and higher volume of blood loss during operation. The levels of these cytokines were positively associated with postoperative liver injury, and the length of hospital stay. Importantly, a high level of IL-6 at POD1 was a risk factor for HCC recurrence and poor disease-free survival after liver resection. CONCLUSIONS: Significantly lower level of GM-CSF, IL-6, IL-8, and MCP-1 after liver resection represented a milder systemic inflammation which might be an important mechanism to offer better short-term and long-term outcomes in LLR over OLR.

Predictive stroke risk model with vision transformer-based Doppler features.

BACKGROUND: Acute stroke is the leading cause of death and disability globally, with an estimated 16 million cases each year. The progression of carotid stenosis reduces blood flow to the intracranial vasculature, causing stroke. Early recognition of ischemic stroke is crucial for disease treatment and management. PURPOSE: A computer-aided diagnosis (CAD) system was proposed in this study to rapidly evaluate ischemic stroke in carotid color Doppler (CCD). METHODS: Based on the ground truth from the clinical examination report, the vision transformer (ViT) features extracted from all CCD images (513 stroke and 458 normal images) were combined in machine learning classifiers to generate the likelihood of ischemic stroke for each image. The pretrained weights from ImageNet reduced the time-consuming training process. The accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve were calculated to evaluate the stroke prediction model. The chi-square test, DeLong test, and Bonferroni correction for multiple comparisons were applied to deal with the type-I error. Only p values equal to or less than 0.00125 were considered to be statistically significant. RESULTS: The proposed CAD system achieved an accuracy of 89%, a sensitivity of 94%, a specificity of 84%, and an area under the receiver operating characteristic curve of 0.95, outperforming the convolutional neural networks AlexNet (82%, p < 0.001), Inception-v3 (78%, p < 0.001), ResNet101 (84%, p < 0.001), and DenseNet201 (85%, p < 0.01). The computational time in model training was only 30 s, which would be efficient and practical in clinical use. CONCLUSIONS: The experiment shows the promising use of CCD images in stroke estimation. Using the pretrained ViT architecture, the image features can be automatically and efficiently generated without human intervention. The proposed CAD system provides a rapid and reliable suggestion for diagnosing ischemic stroke.

Arachidonic acid activates NLRP3 inflammasome in MDSCs via FATP2 to promote post-transplant tumour recurrence in steatotic liver grafts.

Background & Aims: The steatotic grafts have been applied in liver transplantation frequently owing to the high incidence of non-alcoholic fatty liver disease. However, fatty livers are vulnerable to graft injury. Myeloid-derived suppressor cell (MDSC) recruitment during liver graft injury promotes tumour recurrence. Lipid metabolism exerts the immunological influence on MDSCs in tumour progression. Here, we aimed to explore the role and mechanism of inflammasome activation in MDSCs induced by lipid metabolism during fatty liver graft injury and the subsequent effects on tumour recurrence. Methods: MDSC populations and nucleotide-binding oligomerisation domain-like receptor family pyrin domain containing 3 (NLRP3) inflammasome levels were investigated in a clinical cohort and a rat liver transplantation model. The mechanism of NLRP3 activation by specific fatty acids was explored in mouse hepatic ischaemia/reperfusion injury (IRI) with tumour recurrence model and in vitro studies. Results: MDSC populations and NLRP3 levels were increased with higher tumour recurrent rate in patients using steatotic grafts. NLRP3 was upregulated in MDSCs with lipid accumulation post mouse fatty liver IRI. Mechanistically, arachidonic acid was discovered to activate NLRP3 inflammasome in MDSCs through fatty acid transport protein 2 (FATP2), which was identified by screening lipid uptake receptors. The mitochondrial dysfunction with enhanced reactive oxygen species bridged arachidonic acid uptake and NLRP3 activation in MDSCs, which subsequently stimulated CD4+ T cells producing more IL-17 in fatty liver IRI. Blockade of FATP2 inhibited NLRP3 activation in MDSCs, IL-17 production in CD4+ T cells, and the tumour recurrence post fatty liver IRI. Conclusions: During fatty liver graft injury, arachidonic acid activated NLRP3 inflammasome in MDSCs through FATP2, which subsequently stimulated CD4+ T cells producing IL-17 to promote tumour recurrence post transplantation. Impact and implications: The high incidence of non-alcoholic fatty liver disease resulted in the frequent application of steatotic donors in liver transplantation. Our data showed that the patients who underwent liver transplantation using fatty grafts experienced higher tumour recurrence. We found that arachidonic acid activated NLRP3 inflammasome in MDSCs through FATP2 during fatty liver graft injury, which led to more IL-17 secretion of CD4+ T cells and promoted tumour recurrence post transplantation. The inflammasome activation by aberrant fatty acid metabolism in MDSCs bridged the acute-phase fatty liver graft injury and liver tumour recurrence.

Analyzing drama metadata through machine learning to gain insights into social information dissemination patterns.

TV drama, through synchronization with social phenomena, allows the audience to resonate with the characters and desire to watch the next episode. In particular, drama ratings can be the criterion for advertisers to invest in ad placement and a predictor of subsequent economic efficiency in the surrounding areas. To identify the dissemination patterns of social information about dramas, this study used machine learning to predict drama ratings and the contribution of various drama metadata, including broadcast year, broadcast season, TV stations, day of the week, broadcast time slot, genre, screenwriters, status as an original work or sequel, actors and facial features on posters. A total of 800 Japanese TV dramas broadcast during prime time between 2003 and 2020 were collected for analysis. Four machine learning classifiers, including naïve Bayes, artificial neural network, support vector machine, and random forest, were used to combine the metadata. With facial features, the accuracy of the random forest model increased from 75.80% to 77.10%, which shows that poster information can improve the accuracy of the overall predicted ratings. Using only posters to predict ratings with a convolutional neural network still obtained an accuracy rate of 71.70%. More insights about the correlations between drama metadata and social information dissemination patterns were explored.

High prevalence of de novo metabolic dysfunction-associated fatty liver disease after liver transplantation and the role of controlled attenuation parameter.

BACKGROUND & AIMS: Although non-alcoholic fatty liver disease (NAFLD) remains an uncommon indication for liver transplantation (LT) in the Chinese, the prevalence of NAFLD is increasing. We aimed to determine the prevalence of de novo steatosis and metabolic dysfunction-associated fatty liver disease (MAFLD) after LT. METHODS: Transient elastography assessment for liver stiffness and controlled attenuation parameter (CAP) were performed after LT in 549 patients at median time of 77 months from LT. CAP was compared with implant liver biopsy, and also validated in 42 patients with post-LT liver biopsy. Longitudinal history including diabetes mellitus (DM), dyslipidemia, hypertension, and immunosuppressive regimen were recorded. RESULTS: The optimal cut-off level of CAP for diagnosing at least mild (≥ S1) and moderate-to-severe steatosis (≥ S2/3) was 266 and 293 dB/m respectively, with AUROC of 0.740 and 0.954 respectively. Using this newly derived cut-off, 28.9% patients have de novo NAFLD, of which 95.6% fulfilled the criteria for MAFLD. After multivariate analysis, BMI (HR 1.34), DM (HR 2.01), hypertension (HR 2.03), HDL-cholesterol (HR 0.25), LDL-cholesterol (HR 1.5) and cryptogenic cirrhosis (HR 4.85) were associated with the development of S2/3 graft steatosis. de novo NAFLD was associated with higher incidence of new-onset hypertension (p < 0.001), graft dysfunction (defined as ALT > 40 U/L; p = 0.008), but not associated with graft fibrosis (defined as liver stiffness > 12 kPa; p = 0.761). CONCLUSION: Although NAFLD remains an uncommon primary liver disease indication for LT in Chinese patients, post-transplant de novo graft steatosis is common and the majority is classified as MAFLD. Development of graft steatosis is not associated with an increase in graft fibrosis but was associated with worse metabolic control and graft dysfunction. Routine CAP measurement to detect de novo graft steatosis should be considered after LT regardless of the primary indication of LT.