Background:Free air (FA) in the abdominal cavity is a critical finding requiring prompt surgical intervention. We developed an AI-based segmentation model, Free Air-Net (FA-NET), to detect FA in abdominal computed tomography (CT) scans and further refined it with negative training to create FA-NET-NT, aiming to reduce false positives. Materials and methods:FA-NET-NT was developed using a retrospective dataset from a single institution (n = 162). To evaluate its generalizability, the model was validated through both a temporal internal cohort (n = 215) and an independent external cohort from a different hospital (n = 237), which included various CT manufacturers and protocols. The model evaluation was threefold: (1) the Dice score coefficient, (2) image-wise, and (3) patient-wise sensitivity and specificity using representative CT segments (segments 4 through 8, out of 20 equally divided sections of the total axial series). If the model detected at least two images having FA among the representative images, the patient was regarded as having FA. Results:Both models achieved high Dice scores (0.87). FA-NET-NT improved specificity (96%) while maintaining high sensitivity (85%) in image-wise analysis. In patient-wise analysis, FA-NET-NT achieved 95-96% sensitivity for ulcer perforation and 82-92% specificity for non-FA conditions (cholecystitis, pancreatitis, and appendicitis). Specificity for ileus remained moderate (62%). In the external validation, the model demonstrated a patient-wise sensitivity of 95% for ulcer perforation. High specificity was maintained against differential diagnoses, including appendicitis (88%), pancreatitis (88%), cholecystitis (82%), and ileus (80%). Most false-positive findings were attributable to physiological bowel gas mimicking FA. Conclusion:FA-NET-NT is a robust decision-support tool for detecting FA, with its generalizability confirmed through multi-institutional validation. To provide definitive evidence of its clinical superiority, further prospective multi-center trials are necessary in emergency settings.
Backgrounds:Given the distinct pathogenic mechanisms of early-onset Alzheimer's disease (EOAD) and late-onset Alzheimer's disease (LOAD), identifying disease-specific therapeutic targets for each subtype is particularly critical. Methods:We performed proteome-wide Mendelian randomization (MR), colocalization analysis, summary-data-based MR, and Heterogeneity in Dependent Instruments (HEIDI) tests to identify the causal roles of candidate proteins in EOAD and LOAD. Further analyses included protein-protein interaction network analysis, GO/KEGG enrichment analyses, and single-cell RNA sequencing annotation. Druggability evaluation of the target proteins, Phenome-wide association study was conducted to systematically evaluate the potential adverse effects associated with druggable proteins. High-throughput molecular docking and molecular dynamics simulations were conducted to target the top therapeutic targets. Results:Genetically predicted levels of three proteins (APOE, NECTIN2, and PVR) were associated with EOAD risk, nine proteins (APOE, NECTIN2, PVR, EPHB4, SEMA3F, RNASET2, BTN1A1, PSAPL1, and GRN) were associated with LOAD risk, three proteins (APOE, NECTIN2, and PVR) were colocalized with EOAD, and six proteins (APOE, NECTIN2, PVR, SEMA3F, BTN1A1, and EPHB4) were colocalized with LOAD. Three of the proteins (APOE, NECTIN2, and PVR) serve as common targets for both EOAD and LOAD. EPHB4 and BTN1A1 were prioritized for LOAD with the most convincing evidence. The small molecule cucurbit[8]uril exhibits excellent binding affinity with both EPHB4 and BTN1A1 target proteins for the treatment of Alzheimer's disease. Conclusions:This study pinpointed APOE, NECTIN2, and PVR as shared therapeutic targets for EOAD and LOAD, and EPHB4 and BTN1A1 singled out as a priority target for LOAD.
Background:Lung cancer remains the leading cause of cancer-related death, and earlier detection has increased the demand for accurate histological diagnosis While percutaneous transthoracic needle biopsy (PCNB) and surgical biopsy are widely used, concerns remain about PCNB-related tumor dissemination and recurrence. These risks are influenced by tumor location and morphology, which affect procedural feasibility and oncologic outcomes. This study investigated the association between biopsy method and recurrence-free survival (RFS) in surgically resected lung cancer and evaluated how tumor characteristics may identify subgroups at higher risk. Materials and Methods:Medical records of 363 patients with surgically resected primary lung cancer who underwent preoperative PCNB (n = 221) or intraoperative surgical biopsy (n = 142) at a tertiary center between 2015 and 2020 were retrospectively reviewed. Patients were grouped according to biopsy methods and compared. Demographics, preoperative chest CT findings such as consolidation to tumor ratio (CTR) and intralobar tumor location, pathologic findings, and recurrences were evaluated. Results:In the propensity-matched cohort (103 patients per group), multivariate analysis showed male (HR 2.11), current smoking (HR 2.12), central tumors (HR 5.47), and higher CTR (HR 16.6) as independent predictors of reduced RFS, whereas PCNB itself was not (P = 0.16). CTR and tumor location remained important risk factors in both groups. Notably, locoregional and pleural recurrences were more frequent in the PCNB group, although not statistically significant. Five-year RFS was significantly lower in the PCNB group (85.0% vs. 74.3%, P = 0.013), with the negative impact of higher CTR and central tumors being more pronounced in the PCNB group. Conclusion:PCNB was not an independent risk factor for recurrence. However, its association with higher rates of locoregional and pleural recurrence along with a trend toward decreased RFS in tumors with adverse CT features suggests careful clinical consideration when choosing biopsy method.
Background:The global incidence of tumors is rising sharply. Immune checkpoint inhibitors (ICIs) represent the mainstream of immunotherapeutic approaches. Despite the remarkable clinical benefits of immunotherapy in various tumors, ICIs resistance remains an urgent challenge to be addressed. The transcription factors cAMP response element-binding protein 1(CREB1) and c-Jun converge to regulate oncogenesis and immune modulation. Methods:Based on the 'CREBP1CJUN_01' gene set from the Molecular Signatures Database and single-cell RNA sequencing data, we identified 116 CREB-c-Jun transcription factor target-related differential genes (CR.Sig) and refined to 20 CREB-c-Jun transcription factor target-related important feature genes (Hub-CR.Sig) through multi-algorithm machine learning. This Hub-CR.Sig enabled prognostic risk modeling, molecular subtyping of bladder cancer, and somatic mutation stratification. Results:Among these 20 signatures, RAD51C emerged as a top-ranked driver that was correlated with advanced T/N/M stages, associated with poor survival outcomes and is highly expressed in bladder cancer tissues. Functional assessment revealed that RAD51C knockdown suppressed bladder cancer cell proliferation, invasion, migration, and DNA repair capacity in vitro, while also reducing tumor formation in vivo. RNA sequencing analysis implicated the enrichment of JAK-STAT signaling pathways. Meanwhile, we identified an additional type of RAD51C+ fibroblast that was significantly associated with the depth of tumor invasion. Conclusions:The study establishes the Hub-CR.Sig as an immunotherapy response classifier and nominates RAD51C targeting as a promising therapeutic strategy for bladder cancer.
Background:Failure to incise ventral to the Rouvière's sulcus and maintain the dissection plane on the gallbladder (GB) surface predisposes patients to bile duct injuries (BDIs), particularly during trainee-performed laparoscopic cholecystectomy (LC). No existing artificial intelligence (AI) tools offer Tokyo Guidelines 2018 (TG 2018)-anchored real-time guidance. We developed and externally validated an AI navigation system that highlights the alert zone (AZ) - the hepatoduodenal-ligament tissues lying below an imaginary line from the roof of the Rouvière's sulcus to the base of segment 4 and the infundibulum-cystic duct junction - and GB surface specified in the TG 2018. Materials and Methods:Seventy-three LC videos (January 2022-March 2024) were used to train the DeepLab v3 + segmentation model. The AZ and GB surface were manually annotated. The performance was tested on 10 independent videos (100 frames) with an intersection-over-union (IoU) metric. A two-arm pilot usability study randomized 10 fifth- or sixth-year postgraduate surgeons to answer video-based safety questions with or without AI assistance (20 tasks each). Results:The AI achieved a mean IoU of 0.703 (AZ) and 0.735 (GB) compared to the developer ground truth and 0.706 (AZ) and 0.730 (GB) compared to the external ground truth. With AI navigation, the correct selection of a safe incision point increased from 58% to 90%, and contour recognition of the GB surface increased from 70% to 92% (both P < 0.05). Conclusion:The AI navigation system based on the TG 2018 reliably delineated critical landmarks and markedly improved intraoperative trainee decision-making. Prospective real-time trials should determine whether this technology reduces BDIs.
Background:Enhanced recovery after cesarean delivery (ERAC) protocols have been designed to optimize maternal recovery. Although its potential has been demonstrated, high-quality evidence supporting the efficacy of ERAC protocols remains limited. This study aimed to evaluate the effect of anesthesiologist-led bundled ERAC interventions on perioperative adverse events and recovery quality. Methods:In this two-center, double-blind, randomized controlled trial, 122 women undergoing elective cesarean delivery under spinal anesthesia were allocated to receive either conventional care or a bundled ERAC protocol. The ERAC bundle included: preoperative personalized and information-based education; intraoperative prophylactic phenylephrine infusion, mandatory non-opioid analgesia (intravenous acetaminophen and ketorolac), prophylaxis for postoperative nausea and vomiting (ondansetron/dexamethasone), and active fluid warming; and postoperative scheduled ketorolac. The primary outcome was a composite of perioperative adverse events (hypotension, nausea/vomiting, shivering, pruritus, hypothermia, and moderate-to-severe pain). Secondary outcomes included opioid consumption, length of hospital stay, pain scores at rest and with movement, the ObsQoR-11 recovery score, and anxiety scores. Results:Among 116 participants who completed the study (57 conventional care, 59 ERAC), the incidence of the primary composite outcome was significantly lower in the ERAC group (45% vs. 71%; absolute risk difference -31.4%, 95% CI -48.1% to -14.7%; odds ratio 0.25, 95% CI 0.11 to 0.56; P = 0.001). On postoperative day 1, the ERAC group demonstrated superior recovery, with lower worst pain scores (P<0.01), lower pain scores with movement (P<0.01), fewer patient-controlled analgesia boluses (P<0.01), and higher ObsQoR-11 scores (P<0.01). No significant differences were found in pain at rest, length of stay, or anxiety scores. Conclusion:An anesthesiologist-led, bundled ERAC protocol significantly reduced perioperative adverse events and enhanced the quality of recovery after elective cesarean delivery compared to conventional care.
Perihilar cholangiocarcinoma (PHCC) is a malignant tumor arising from bile ducts at the hilar area. It was first described by Altemeyer et al in 1957 in Cincinnati and later in 1965 by Gerald Klatskin at Yale University. The surgical management with R0 resection represents the only potentially curative treatment for PHCC patients. The ideal surgical treatment represents a challenge for liver surgeons across the globe, and expertise in vascular reconstruction and transplantation is required. The anatomy of the liver hilum is a complex area where the proximity of portal vein bifurcation and hepatic artery branches, especially the right hepatic artery, gets an intimal relation with the bile duct confluence area. This explains frequent vascular involvement in such tumors. The modern surgical approaches in PHCC involve several topics of discussion and controversies among hepatobiliary surgeons. Some key points will be discussed in this comprehensive discussion.
Background and aims:Colonoscopy remains as an invaluable diagnostic and therapeutic tool used for the screening and evaluation of colonic pathologies. Periodic endoscopic screening allowing for early detection and removal of pre-malignant lesions, resulting in significant overall decreased incidence and mortality from colorectal cancer. Despite its added costs, artificial intelligence (AI) has been increasingly utilized in the field of endoscopy, boasting benefits of improving adenoma detection rates while reducing endoscopists' mental fatigue intra-procedurally. As such, further evaluation of the financial impact of such AI-usage in screening colonoscopies was performed to assess if routine use of AI-guidance in screening colonoscopies was justifiable. Methods:This study conducted a cost effectiveness analysis of the GI Genius™ CADx Intelligent Endoscopy Module (US-DG-2000309, 2021 Medtronic) system in our local Singaporean tertiary public healthcare institution. A decision tree model was used to calculate Incremental Cost Effectiveness Ratio (ICER), comparing cancer-related costs for patients who underwent conventional colonoscopy versus colonoscopy with AI guidance. Results:With a calculated ICER value of 0.72, our study suggests that AI-assisted colonoscopy was indeed cost effective. Quality-Adjusted Life Years (QALY) values derived for each subgroup of patients further demonstrated overall improved quality of life reported by patients who underwent AI-assisted colonoscopy, as compared to conventional colonoscopy. Conclusion:Our study supports the incorporation of AI guidance into routine colonoscopic evaluation in view of increased cost effectiveness and improved overall quality of life achieved. However, further studies and analysis must be undertaken before reliably determining the cost-effectiveness of AI on regional or international scales.
