Drug-induced liver injury in children: A nationwide cohort study from China.

Background & Aims: Currently, there is limited knowledge on the clinical profile of drug-induced liver injury (DILI) in Chinese children. We aimed to assess the clinical characteristics, suspected drugs, and outcomes associated with pediatric DILI in China. Methods: This nationwide, multicenter, retrospective study, conducted between 2012 and 2014, analyzed 25,927 cases of suspected DILI at 308 medical centers using the inpatient medical register system. Utilizing the Roussel Uclaf causality assessment method score, only patients with scores ≥6 or diagnosed with DILI by three experts after scoring <6 were included in the analysis. Among them, 460 cases met the EASL biochemical criteria. The study categorized children into three age groups: toddlers (≥30 days to <6 years old), school-age children (6 to <12 years old), and adolescents (12 to <18 years old). Results: Hepatocellular injury was the predominant clinical classification, accounting for 63% of cases, with 34% of these cases meeting Hy's law criteria. Adolescents comprised the majority of children with moderate/severe DILI (65%). Similarly, adolescents faced a significantly higher risk of severe liver injury compared to younger children (adjusted odd ratios 4.75, p = 0.002). The top three most frequently prescribed drug classes across all age groups were antineoplastic agents (25.9%), antimicrobials (21.5%), and traditional Chinese medicine (13.7%). For adolescents, the most commonly suspected drugs were antitubercular drugs (22%) and traditional Chinese medicine (23%). Conclusion: Adolescents are at a greater risk of severe and potentially fatal liver injury compared to younger children. Recognizing the risk of pediatric DILI is crucial for ensuring safe medical practices. Impact and implications: Drug-induced liver injury, a poorly understood yet serious cause of pediatric liver disease, encompasses a spectrum of clinical presentations, ranging from asymptomatic liver enzyme elevation to acute liver failure. This retrospective study, utilizing a large Chinese cohort of pediatric liver injury cases from 308 centers nationwide, characterized the major clinical patterns and suspected drugs in detail, revealing that adolescents are at a greater risk of severe liver injury compared to younger children. Vigilant care and careful surveillance of at-risk pediatric patients are crucial for physicians, researchers, patients, caregivers, and policymakers. Additional multicenter prospective studies are needed to evaluate the risk of hepatotoxicity in outpatients and hospitalized pediatric patients.

An Unbiased Proteomic Platform for Activity-based Arginylation Profiling.

Protein arginylation is an essential posttranslational modification (PTM) catalyzed by arginyl-tRNA-protein transferase 1 (ATE1) in mammalian systems. Arginylation features a post-translational conjugation of an arginyl to a protein, making it extremely challenging to differentiate from translational arginine residues with the same mass in a protein sequence. Here we present a general activity-based arginylation profiling (ABAP) platform for the unbiased discovery of arginylation substrates and their precise modification sites. This method integrates isotopic arginine labeling into an ATE1 assay utilizing biological lysates (ex vivo) rather than live cells, thus eliminating translational bias derived from the ribosomal activity and enabling bona fide arginylation identification using isotopic features. ABAP has been successfully applied to an array of peptide, protein, cell, patient, and animal tissue samples using 20 µg sample input, with 229 unique arginylation sites revealed from human proteomes. Representative sites were validated and followed up for their biological functions. The developed platform is globally applicable to the aforementioned sample types and therefore paves the way for functional studies of this difficult-to-characterize protein modification.

Locally advanced breast cancer patients should be cautious about the immediate breast reconstruction after mastectomy: a pooling analysis of safety and efficacy.

BACKGROUND: The purpose of this study was to compare safety and efficacy outcomes between immediate breast reconstruction (IBR) and mastectomy alone in locally advanced breast cancer patients. METHODS: We conducted a comprehensive literature search of PUBMED, EMBASE, and Cochrane databases. The primary outcomes evaluated were overall survival, disease-free survival, and local recurrence. The secondary outcome was the incidence of surgical complications. All data were analyzed using Review Manager 5.3. RESULTS: Sixteen studies, involving 15,364 participants were included in this meta-analysis. Pooled data demonstrated that patients underwent IBR were more likely to experience surgical complications than those underwent mastectomy alone (HR: 3.96, 95%CI [1.07,14.67], p = 0.04). No significant difference was found in overall survival (HR: 0.94, 95%CI [0.73,1.20], p = 0.62), disease-free survival (HR: 1.03, 95%CI [0.83,1.27], p = 0.81), or breast cancer specific survival (HR: 0.93, 95%CI [0.71,1.21], p = 0.57) between IBR group and Non-IBR group. CONCLUSIONS: Our study demonstrates that IBR after mastectomy does not affect the overall survival and disease-free survival of locally advanced breast cancer patients. However, IBR brings with it a nonnegligible higher risk of complications and needs to be fully evaluated and carefully decided.

Laparoscopic spleen-preserving total pancreatectomy for the treatment of low-grade malignant pancreatic tumors: Two case reports and review of literature.

BACKGROUND: Function-preserving pancreatectomy can improve the long-term quality of life of patients with benign or low-grade malignant tumors, such as intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms. However, there is limited literature on laparoscopic spleen-preserving total pancreatectomy (L-SpTP) due to technical difficulties. CASE SUMMARY: Patient 1 was a 51-year-old male diagnosed with IPMN based on preoperative imaging, showing solid nodules in the pancreatic head and diffuse dilation of the main pancreatic duct with atrophy of the distal pancreas. We performed L-SpTP with preservation of the splenic vessels, and the postoperative pathology report revealed IPMN with invasive carcinoma. Patient 2 was a 60-year-old male with multiple cystic lesions in the pancreatic head and body. L-SpTP was performed, and intraoperatively, the splenic vein was injured and required ligation. Postoperative pathology revealed a mucinous cystic tumor of the pancreas with low-grade dysplasia. Both patients were discharged on postoperative day 7, and there were no major complications during the perioperative period. CONCLUSION: We believe that L-SpTP is a safe and feasible treatment for low-grade malignant pancreatic tumors, but more case studies are needed to evaluate its safety, efficacy, and long-term outcomes.

Segmentation Guided Crossing Dual Decoding Generative Adversarial Network for Synthesizing Contrast-Enhanced Computed Tomography Images.

Although contrast-enhanced computed tomography (CE-CT) images significantly improve the accuracy of diagnosing focal liver lesions (FLLs), the administration of contrast agents imposes a considerable physical burden on patients. The utilization of generative models to synthesize CE-CT images from non-contrasted CT images offers a promising solution. However, existing image synthesis models tend to overlook the importance of critical regions, inevitably reducing their effectiveness in downstream tasks. To overcome this challenge, we propose an innovative CE-CT image synthesis model called Segmentation Guided Crossing Dual Decoding Generative Adversarial Network (SGCDD-GAN). Specifically, the SGCDD-GAN involves a crossing dual decoding generator including an attention decoder and an improved transformation decoder. The attention decoder is designed to highlight some critical regions within the abdominal cavity, while the improved transformation decoder is responsible for synthesizing CE-CT images. These two decoders are interconnected using a crossing technique to enhance each other's capabilities. Furthermore, we employ a multi-task learning strategy to guide the generator to focus more on the lesion area. To evaluate the performance of proposed SGCDD-GAN, we test it on an in-house CE-CT dataset. In both CE-CT image synthesis tasks-namely, synthesizing ART images and synthesizing PV images-the proposed SGCDD-GAN demonstrates superior performance metrics across the entire image and liver region, including SSIM, PSNR, MSE, and PCC scores. Furthermore, CE-CT images synthetized from our SGCDD-GAN achieve remarkable accuracy rates of 82.68%, 94.11%, and 94.11% in a deep learning-based FLLs classification task, along with a pilot assessment conducted by two radiologists.

Translocation of two-dimensional active polymers through nanopores using Langevin dynamics simulations.

The translocation of polymers through nanopores is a complex process influenced by various factors. In this study, the translocation behavior of a two-dimensional active polymer chain, comprised of a head active Brownian particle (ABP) and a tail passive polymer chain, through a nanopore is studied using Langevin dynamics simulations. Results show that the effect of the self-propulsion force of the ABP on the translocation differs significantly from the driving force inside the pore for traditional polymer translocations. Specifically, the translocation time τ initially increases with increasing the magnitude fs of the self-propulsion force and then decreases with a further increase in fs. A small fs lowers the potential barrier for the translocation and thus promotes slow translocations, whereas a large fs directly pulls the polymer chain through the nanopore following the scaling relation τ ∝ fs-1. Moreover, two asymptotic scaling relations between τ and polymer length N, τ ∝ Nα, are found, with the exponent α of about 2.5 for small fs or long N and the exponent α of about 1.4 for short active polymers with large fs. We discover that the slow rotation of the ABP accelerates the translocation process.

Harnessing molecular mechanism for precision medicine in dilated cardiomyopathy caused by a mutation in troponin T.

Familial dilated cardiomyopathy (DCM) is frequently caused by autosomal dominant point mutations in genes involved in diverse cellular processes, including sarcomeric contraction. While patient studies have defined the genetic landscape of DCM, genetics are not currently used in patient care, and patients receive similar treatments regardless of the underlying mutation. It has been suggested that a precision medicine approach based on the molecular mechanism of the underlying mutation could improve outcomes; however, realizing this approach has been challenging due to difficulties linking genotype and phenotype and then leveraging this information to identify therapeutic approaches. Here, we used multiscale experimental and computational approaches to test whether knowledge of molecular mechanism could be harnessed to connect genotype, phenotype, and drug response for a DCM mutation in troponin T, deletion of K210. Previously, we showed that at the molecular scale, the mutation reduces thin filament activation. Here, we used computational modeling of this molecular defect to predict that the mutant will reduce cellular and tissue contractility, and we validated this prediction in human cardiomyocytes and engineered heart tissues. We then used our knowledge of molecular mechanism to computationally model the effects of a small molecule that can activate the thin filament. We demonstrate experimentally that the modeling correctly predicts that the small molecule can partially rescue systolic dysfunction at the expense of diastolic function. Taken together, our results demonstrate how molecular mechanism can be harnessed to connect genotype and phenotype and inspire strategies to optimize mechanism-based therapeutics for DCM.

Activation of the MEK/ERK Pathway Mediates the Inhibitory Effects of Silvestrol on Nasopharyngeal Carcinoma Cells via RAP1A, HK2, and GADD45A.

BACKGROUND: Nasopharyngeal carcinoma (NPC) is a malignant tumor associated with Epstein-Barr virus (EBV) infection. Chemoradiotherapy is the mainstream treatment for locally advanced NPC, and chemotherapeutic drugs are an indispensable part of NPC treatment. However, the toxic side-effects of chemotherapy drugs limit their therapeutic value, and new chemotherapy drugs are urgently needed for NPC. Silvestrol, an emerging natural plant anticancer molecule, has shown promising antitumor activity in breast cancer, melanoma, liver cancer, and other tumor types by promoting apoptosis in cancer cells to a greater extent than in normal cells. However, the effects of silvestrol on NPC and its possible molecular mechanisms have yet to be fully explored. METHODS: Cell counting kit-8 (CCK-8), cell scratch, flow cytometry, 5-ethynyl-2'-deoxyuridine (EdU), and Western blot (WB) assays were used to evaluate the effects of silvestrol on the cell viability, cell cycle, apoptosis, and migration of NPC cells. RNA sequencing (RNA-Seq) was used to study the effect of extracellular signal-regulated kinase (ERK) inhibitors on the cell transcriptome, and immunohistochemistry (IHC) to assess protein expression levels in patient specimens. RESULTS: Silvestrol inhibited cell migration and DNA replication of NPC cells, while promoting the expression of cleaved caspase-3, apoptosis, and cell cycle arrest. Furthermore, silvestrol altered the level of ERK phosphorylation. The ERK-targeted inhibitor LY3214996 attenuated silvestrol-mediated inhibition of NPC cell proliferation but not migration. Analysis of RNA-Seq data and WB were used to identify and validate the downstream regulatory targets of silvestrol. Expression of GADD45A, RAP1A, and hexokinase-II (HK2) proteins was inhibited by silvestrol and LY3214996. Finally, IHC revealed that GADD45A, RAP1A, and HK2 protein expression was more abundant in cancer tissues than in non-tumor tissues. CONCLUSIONS: Silvestrol inhibits the proliferation of NPC cells by targeting ERK phosphorylation. However, the inhibition of NPC cell migration by silvestrol was independent of the Raf-MEK-ERK pathway. RAP1A, HK2, and GADD45A may be potential targets for the action of silvestrol.

Integrated proteomics reveals autophagy landscape and an autophagy receptor controlling PKA-RI complex homeostasis in neurons.

Autophagy is a conserved, catabolic process essential for maintaining cellular homeostasis. Malfunctional autophagy contributes to neurodevelopmental and neurodegenerative diseases. However, the exact role and targets of autophagy in human neurons remain elusive. Here we report a systematic investigation of neuronal autophagy targets through integrated proteomics. Deep proteomic profiling of multiple autophagy-deficient lines of human induced neurons, mouse brains, and brain LC3-interactome reveals roles of neuronal autophagy in targeting proteins of multiple cellular organelles/pathways, including endoplasmic reticulum (ER), mitochondria, endosome, Golgi apparatus, synaptic vesicle (SV) for degradation. By combining phosphoproteomics and functional analysis in human and mouse neurons, we uncovered a function of neuronal autophagy in controlling cAMP-PKA and c-FOS-mediated neuronal activity through selective degradation of the protein kinase A - cAMP-binding regulatory (R)-subunit I (PKA-RI) complex. Lack of AKAP11 causes accumulation of the PKA-RI complex in the soma and neurites, demonstrating a constant clearance of PKA-RI complex through AKAP11-mediated degradation in neurons. Our study thus reveals the landscape of autophagy degradation in human neurons and identifies a physiological function of autophagy in controlling homeostasis of PKA-RI complex and specific PKA activity in neurons.

Reduced Volume Expansion of Micron-Sized SiOx via Closed-Nanopore Structure Constructed by Mg-Induced Elemental Segregation.

The inherently huge volume expansion during Li uptake has hindered the use of Si-based anodes in high-energy lithium-ion batteries. While some pore-forming and nano-architecting strategies show promises to effectively buffer the volume change, other parameters essential for practical electrode fabrication, such as compaction density, are often compromised. Here we propose a new in situ Mg doping strategy to form closed-nanopore structure into a micron-sized SiOx particle at a high bulk density. The doped Mg atoms promote the segregation of O, so that high-density magnesium silicates form to generate closed nanopores. By altering the mass content of Mg dopant, the average radii (ranged from 5.4 to 9.7 nm) and porosities (ranged from 1.4 % to 15.9 %) of the closed pores are precisely adjustable, which accounts for volume expansion of SiOx from 77.8 % to 22.2 % at the minimum. Benefited from the small volume variation, the Mg-doped micron-SiOx anode demonstrates improved Li storage performance towards realization of a 700-(dis)charge-cycle, 11-Ah-pouch-type cell at a capacity retention of >80 %. This work offers insights into reasonable design of the internal structure of micron-sized SiOx and other materials that undergo conversion or alloying reactions with drastic volume change, to enable high-energy batteries with stable electrochemistry.

Association of adherence to the enhanced recovery after surgery pathway and outcomes after laparoscopic total gastrectomy.

OBJECTIVE: The current study used a composite outcome to investigate whether applying the ERAS protocol would enhance the recovery of patients undergoing laparoscopic total gastrectomy (LTG). EXPOSURES: Laparoscopic total gastrectomy and perioperative interventions were the exposure. An ERAS clinical pathway consisting of 14 items was implemented and assessed. Patients were divided into either ERAS-compliant or non-ERAS-compliant group according the adherence above 9/14 or not. MAIN OUTCOMES AND MEASURES: The primary study outcome was a composite outcome called 'optimal postoperative recovery' with the definition as below: discharge within 6 days with no sever complications and no unplanned re-operation or readmission within 30 days postoperatively. Univariate logistic regression analysis and multivariate logistic regression analysis were used to model optimal postoperative recovery and compliance, adjusting for patient-related and disease-related characteristics. RESULTS: A total of 252 patients were included in this retrospective study, 129 in the ERAS compliant group and 123 in the non-ERAS-compliant group. Of these, 79.07% of the patients in ERAS compliant group achieved optimal postoperative recovery, whereas 61.79% of patients in non-ERAS-compliant group did (P = 0.0026). The incidence of sever complications was lower in the ERAS-compliant group (1.55% vs. 6.5%, P = 0.0441). No patients in ERAS compliant group had unplanned re-operation, whereas 5.69% (7/123) of patients in non-ERAS-compliant group had (p = 0.006). The median length of the postoperative hospital stay was shorter in the in the ERAS compliant group (5.51 vs. 5.68 days, P = 0.01). Both logistic (OR 2.01, 95% CI 1.21-3.34) and stepwise regression (OR 2.07, 95% CI 1.25-3.41) analysis showed that high overall compliance with the ERAS protocol facilitated optimal recovery in such patients. In bivariate analysis of compliance for patients who had an optimal postoperative recovery, carbohydrate drinks (p = 0.0196), early oral feeding (P = 0.0043), early mobilization (P = 0.0340), and restrictive intravenous fluid administration (P < 0.0001) were significantly associated with optimal postoperative recovery. CONCLUSIONS AND RELEVANCE: Patients with higher ERAS compliance (almost 70% of the accomplishment) suffered less severe postoperative complications and were more likely to achieve optimal postoperative recovery.

Misdiagnosis of ß-Thalassemia Major Due to Chinese Gγ+(Aγδß)0-Thalassemia Combined with ß0-Thalassemia.

δß-thalassemia is a rare type of thalassemia characterized by increased Hb F levels, including mainly Chinese Gγ(Aγδß)0-thalassemia, Yunnanese Gγ(Aγδß)0-thalassemia, Cantonese Gγ(Aγδß)0-thalassemia in China. Due to the low rate of δß-thalassemia carriers, there are few reports of δß-thalassemia combined with ß-thalassemia causing ß-thalassemia major. Herein, we described the combination of Chinese Gγ(Aγδß)0-thalassemia and ß-thalassemia leading to ß-thalassemia major in a Chinese patient. Hemoglobin analysis was performed by capillary electrophoresis (CE). Routine genetic analysis was carried out by gap-polymerase chain reaction (Gap-PCR) and PCR and reverse dot blot (PCR-RDB). Multiple ligation-dependent probe amplification (MLPA) was used to detect the large deletion, and Gap-PCR confirmed the deletion. A CE result showed an elevated Hb F level of 98.7% and 11.7% in the proband and her mother, but the proband was diagnosed with ßCD17M/ßCD17M using routine genetic analysis. However, her father was heterozygous for CD17 in ß-globin, and her mother was detected as SEA heterozygous. The further analysis presented that the proband had actually missed the diagnosis of Chinese Gγ(Aγδß)0-thalassemia by MLPA and PCR-RDB. Finally, the genotype of the proband was corrected from ßCD17M/ßCD17M to ßCD17M/ßGγ(Aγδß)0. This is the first report of Chinese Gγ(Aγδß)0-thalassemia combined with ß-thalassemia resulting in ß-thalassemia major in China. Screening for δß-thalassemia by Hb analysis could be an effective method.

The compensatory increase of Gli-similar 3 inhibited neuronal apoptosis through regulating Mps one binder kinase activator 1b (MOB1b): a possible strategy for the functional recovery after spinal cord injury.

Spinal cord injury (SCI) is a devastating disease characterized by neuronal apoptosis. Gli-similar 3 (GLIS3), a transcriptional factor, was involved in cell apoptosis and associated with the transcription of downstream target genes related to neuronal function. However, the function of GLIS3 in SCI remains unknown. Therefore, we used the mouse model of SCI to explore the role of GLIS3 in SCI. The results showed that GLIS3 expression was significantly increased in spinal cord tissues of SCI mice, and GLIS3 overexpression promoted the functional recovery, reserved histological changes, and inhibited neuronal apoptosis after SCI. Through online tools, the potential target genes of GLIS3 were analyzed and we found that Mps one binder kinase activator 1b (Mob1b) had a strong association with SCI among these genes. MOB1b is a core component of Hippo signaling pathway, which was reported to inhibit cell apoptosis. MOB1b expression was significantly increased in mice at 7 days post-SCI and GLIS3 overexpression further increased its expression. Dual-luciferase reporter assay revealed that GLIS3 bound to the promoter of Mob1b and promoted its transcription. In conclusion, our findings reveal that the compensatory increase of GLIS3 promotes functional recovery after SCI through inhibiting neuronal apoptosis by transcriptionally regulating MOB1b. Our study provides a novel target for functional recovery after SCI.

Adsorption of active polymers on attractive nanoparticles.

The adsorption of active polymers on an attractive nanoparticle (NP) is studied using Langevin dynamics simulations. The active polymers consist of an active Brownian particle (ABP) at the head and a subsequent passive polymer chain. The ABP experiences an active force of magnitude Fa. The interactions between the active polymer and NP are modeled as Lennard-Jones potential with a strength εpn. We find the critical adsorption point εpn* increases with increasing the active force Fa. The increment of εpn*, denoted as Δεpn*, due to Fa can be expressed approximately as Δεpn* ∝ Fa2.5 for the restricted rotating active polymer (RRAP) where the rotation of the head ABP is restricted and Δεpn* ∝ Fa1.7 for the freely rotating active polymer (FRAP) where the ABP rotates freely. Meanwhile, the conformation of the adsorbed polymer, such as adsorbed trains on NP and the tail near the ABP, are also dependent on Fa. When the tail near the ABP is short, the adsorption is significantly affected by the active force. However, when the tail is long, the whole polymer can be viewed as a long tail stretched by the active force and unperturbed adsorption monomers. Simulation results show that the active force has a direct and significant effect on εpn* and the structure of the adsorbed active polymers.

Spatial Attention-Guided Generative Adversarial Network for Synthesizing Contrast-enhanced Computed Tomography Images.

Compared to non-contrast computed tomography (NC-CT) scans, contrast-enhanced (CE) CT scans provide more abundant information about focal liver lesions (FLLs), which play a crucial role in the FLLs diagnosis. However, CE-CT scans require patient to inject contrast agent into the body, which increase the physical and economic burden of the patient. In this paper, we propose a spatial attention-guided generative adversarial network (SAG-GAN), which can directly obtain corresponding CE-CT images from the patient's NC-CT images. In the SAG-GAN, we devise a spatial attention-guided generator, which utilize a lightweight spatial attention module to highlight synthesis task-related areas in NC-CT image and neglect unrelated areas. To assess the performance of our approach, we test it on two tasks: synthesizing CE-CT images in arterial phase and portal venous phase. Both qualitative and quantitative results demonstrate that SAG-GAN is superior to existing GANs-based image synthesis methods.

Identifying Sex-Specific Serum Patterns of Alzheimer's Mice through Deep TMT Profiling and a Concentration-Dependent Concatenation Strategy.

Alzheimer's disease (AD) is the most prevalent form of dementia, disproportionately affecting women in disease prevalence and progression. Comprehensive analysis of the serum proteome in a common AD mouse model offers potential in identifying possible AD pathology- and gender-associated biomarkers. Here, we introduce a multiplexed, nondepleted mouse serum proteome profiling via tandem mass-tag (TMTpro) labeling. The labeled sample was separated into 475 fractions using basic reversed-phase liquid chromatography (RPLC), which were categorized into low-, medium-, and high-concentration fractions for concatenation. This concentration-dependent concatenation strategy resulted in 128 fractions for acidic RPLC-tandem mass spectrometry (MS/MS) analysis, collecting ∼5 million MS/MS scans and identifying 3972 unique proteins (3413 genes) that cover a dynamic range spanning at least 6 orders of magnitude. The differential expression analysis between wild type and the commonly used AD model (5xFAD) mice exhibited minimal significant protein alterations. However, we detected 60 statistically significant (FDR < 0.05), sex-specific proteins, including complement components, serpins, carboxylesterases, major urinary proteins, cysteine-rich secretory protein 1, pregnancy-associated murine protein 1, prolactin, amyloid P component, epidermal growth factor receptor, fibrinogen-like protein 1, and hepcidin. The results suggest that our platform possesses the sensitivity and reproducibility required to detect sex-specific differentially expressed proteins in mouse serum samples.

A Boundary-Enhanced Liver Segmentation Network for Multi-Phase CT Images with Unsupervised Domain Adaptation.

Multi-phase computed tomography (CT) images have gained significant popularity in the diagnosis of hepatic disease. There are several challenges in the liver segmentation of multi-phase CT images. (1) Annotation: due to the distinct contrast enhancements observed in different phases (i.e., each phase is considered a different domain), annotating all phase images in multi-phase CT images for liver or tumor segmentation is a task that consumes substantial time and labor resources. (2) Poor contrast: some phase images may have poor contrast, making it difficult to distinguish the liver boundary. In this paper, we propose a boundary-enhanced liver segmentation network for multi-phase CT images with unsupervised domain adaptation. The first contribution is that we propose DD-UDA, a dual discriminator-based unsupervised domain adaptation, for liver segmentation on multi-phase images without multi-phase annotations, effectively tackling the annotation problem. To improve accuracy by reducing distribution differences between the source and target domains, we perform domain adaptation at two levels by employing two discriminators, one at the feature level and the other at the output level. The second contribution is that we introduce an additional boundary-enhanced decoder to the encoder-decoder backbone segmentation network to effectively recognize the boundary region, thereby addressing the problem of poor contrast. In our study, we employ the public LiTS dataset as the source domain and our private MPCT-FLLs dataset as the target domain. The experimental findings validate the efficacy of our proposed methods, producing substantially improved results when tested on each phase of the multi-phase CT image even without the multi-phase annotations. As evaluated on the MPCT-FLLs dataset, the existing baseline (UDA) method achieved IoU scores of 0.785, 0.796, and 0.772 for the PV, ART, and NC phases, respectively, while our proposed approach exhibited superior performance, surpassing both the baseline and other state-of-the-art methods. Notably, our method achieved remarkable IoU scores of 0.823, 0.811, and 0.800 for the PV, ART, and NC phases, respectively, emphasizing its effectiveness in achieving accurate image segmentation.

A Lightweight Network for Contextual and Morphological Awareness for Hepatic Vein Segmentation.

Accurate segmentation of the hepatic vein can improve the precision of liver disease diagnosis and treatment. Since the hepatic venous system is a small target and sparsely distributed, with various and diverse morphology, data labeling is difficult. Therefore, automatic hepatic vein segmentation is extremely challenging. We propose a lightweight contextual and morphological awareness network and design a novel morphology aware module based on attention mechanism and a 3D reconstruction module. The morphology aware module can obtain the slice similarity awareness mapping, which can enhance the continuous area of the hepatic veins in two adjacent slices through attention weighting. The 3D reconstruction module connects the 2D encoder and the 3D decoder to obtain the learning ability of 3D context with a very small amount of parameters. Compared with other SOTA methods, using the proposed method demonstrates an enhancement in the dice coefficient with few parameters on the two datasets. A small number of parameters can reduce hardware requirements and potentially have stronger generalization, which is an advantage in clinical deployment.

Risk stratification of clinically relevant delayed gastric emptying after pancreaticoduodenectomy.

BACKGROUND: Delayed gastric emptying (DGE) remains one of the major complications after pancreaticoduodenectomy (PD), with discrepant reports of its contributing factors. This study aimed to develop a nomogram to identify potential predictors and predict the probability of DGE after PD. METHODS: This retrospective study enrolled 422 consecutive patients who underwent PD from January 2019 to December 2021 at our institution. The LASSO algorithm and multivariate logistic regression were performed to identify independent risk and protective factors associated with clinically relevant delayed gastric emptying (CR-DGE). A nomogram was established based on the selected variables. Then, the calibration curve, ROC curve, decision curve analysis (DCA), and clinical impact curve (CIC) were applied to evaluate the predictive performance of our model. Finally, an independent cohort of 45 consecutive patients from January 2022 to March 2022 was enrolled to further validate the nomogram. RESULTS: Among 422 patients, CR-DGE occurred in 94 patients (22.2%). A previous history of chronic gastropathy, intraoperative plasma transfusion ≥ 400 ml, end-to-side gastrointestinal anastomosis, intra-abdominal infection, incisional infection, and clinically relevant postoperative pancreatic fistula (CR-POPF) were identified as risk predictors. Minimally invasive pancreaticoduodenectomy (MIPD) was demonstrated to be a protective predictor of CR-DGE. The areas under the curve (AUCs) were 0.768 (95% CI, 0.706-0.830) in the development cohort, 0.766 (95% CI, 0.671-0.861) in the validation cohort, and 0.787 (95% CI, 0.633-0.940) in the independent cohort. Then, we built a simplified scale based on our nomogram for risk stratification. CONCLUSIONS: Our study identified seven predictors and constructed a validated nomogram that effectively predicted CR-DGE for patients who underwent PD.