INTRODUCTION: An interdisciplinary panel, comprising professionals from medicine, AI and data science, law and ethics, and patient advocacy, convened to discuss key principles on regulation, implementation and evaluation of AI models in healthcare for Singapore. METHOD: The panel considered 14 statements split across 4 themes: "The Role and Scope of Regulatory Entities," "Regulatory Processes," "Pre-Approval Evaluation of AI Models" and "Medical AI in Practice". Moderated by a thematic representative, the panel deliberated on each statement and modified it until a majority agreement threshold is met. The roundtable meeting was convened in Singapore on 1 July 2024. While the statements reflect local perspectives, they may serve as a reference for other countries navigating similar challenges in AI governance in healthcare. RESULTS: Balanced testing approaches, differentiated regulatory standards for autonomous and assistive AI, and context-sensitive requirements are essential in regulating AI models in healthcare. A hybrid approach-integrating global standards with local needs to ensure AI comple-ments human decision-making and enhances clinical expertise-was recommended. Additionally, the need for patient involvement at multiple levels was underscored. There are active ongoing efforts towards development and refinement of AI governance guidelines and frameworks balancing between regulation and freedom. The statements defined therein provide guidance on how prevailing values and viewpoints can streamline AI implementation into healthcare. CONCLUSION This roundtable discussion is among the first in Singapore to develop a structured set of state-ments tailored for the regulation, implementation and evaluation of AI models in healthcare, drawing on interdisciplinary expertise from medicine, AI, data science, law, ethics and patient advocacy.
Singapore ; Humans ; Artificial Intelligence/standards* ; Delivery of Health Care/organization & administration*

Objective To predict and preliminarily validate potential shared key genes involved in cardiac complications caused by influenza A virus(IAV)and severe acute respiratory syndrome coronavirus 2(SARS-CoV-2)infections.Methods Differentially expressed genes(DEGs)associated with cardiac complications were obtained from the Gene Expression Omnibus(GEO)database.A hierarchical intersection strategy was applied.First,cardiac complication related DEGs were overlapped with 2 independent virus related gene sets:3 454 human genes linked to IAV infection in GeneCards and 333 human protein-coding genes interacting with SARS-CoV-2 in the Human Protein Atlas.The 2 overlap results were then intersected to yield 22 hub genes.Lasso regression,random forest(RF)and support vector machine algorithms(SVM)were employed to refine this list.Predicted genes were validated in vitro in H1N1-infected human cardiomyocyte AC16 cells and in vivo in IFITM3 knockout mice challenged with H1N1,assessing transcriptional changes.Results A total of 22 hub genes were identified through integrative bioinformatics analysis.Application of the 3 machine learning algorithms resulted in 5 common key genes:ACE2,TBK1,NUP210,PUSL1,and MEPCE.In vitro infection of AC16 cells with H1N1 revealed dynamic transcriptional changes in all 5 genes post-infection(P<0.05).In vivo experiments using H1N1-infected IFITM3 knockout mice confirmed the dynamic mRNA expression changes of these 5 genes,consistent with the in vitro results(P<0.05).Conclusion By combining multilayered bioinformatics analysis with 3 machine learning approaches,5 common key genes are identified:ACE2,TBK1,NUP210,PUSL1 and MEPCE.Validation in H1N1 infection models confirms their relevance to IAV-induced cardiac complications.

Background and Aims:Laparoscopic pancreaticoduodenectomy(LPD)has become a preferred approach for periampullary tumors,yet delayed gastric emptying(DGE)remains a frequent complication that hampers postoperative recovery.The nasojejunal feeding tube(NJT)is commonly used for early enteral nutrition,but its impact on DGE is controversial.This study aimed to evaluate whether intraoperative NJT placement increases the risk of DGE after LPD and to assess its influence on postoperative recovery outcomes.Methods:A retrospective cohort of 319 patients who underwent LPD at Provincial Hospital Affiliated to Shandong First Medical University from April 2017 to November 2023 was analyzed.Patients were divided into two groups based on intraoperative NJT placement(NJT group,n=200;non-NJT group,n=119).The incidence of DGE and postoperative outcomes were compared.Multivariate logistic regression and propensity score matching(PSM)were performed to identify independent risk factors for DGE.Results:The incidence of grade B/C DGE was significantly higher in the NJT group than in the non-NJT group(36.5%vs.21.8%,P=0.006).NJT placement was associated with longer postoperative hospital stay and higher hospitalization costs(both P<0.05).Multivariate analysis revealed intraoperative NJT placement(OR=1.960,95%CI=1.142-3.363,P=0.015)and intraoperative blood loss>400 mL(OR=1.921,95%CI=1.155-3.194,P=0.012)as independent risk factors for DGE.These findings were consistent after PSM.Conclusions:Prophylactic intraoperative NJT placement confers no additional benefit for postoperative recovery after LPD and is associated with a higher risk of DGE,prolonged hospitalization,and increased medical costs.Routine NJT placement should therefore be avoided,and individualized strategies should be adopted to minimize postoperative complications and enhance recovery.

Objective:To evaluate the performance of an artificial intelligence-based endoscopic virtual ruler (EVR) for non-invasive measurement of esophageal varices (EV) diameter.Methods:Patients with liver cirrhosis and EV hospitalized at the First Affiliated Hospital of Anhui Medical University between October 2022 and May 2023 were prospectively enrolled. EV diameter was measured using visual estimation, esophageal varix manometer (EVM), and EVR, with procedure times recorded. The intraclass correlation coefficient (ICC) was used to assess the consistency of EV diameter measurement obtained from the three methods, and repeated-measures ANOVA was used to compare differences in time measurements across three methods.Results:The study included 41 patients with liver cirrhosis and EV. Inter-observer ICC for visual estimation was 0.594, versus 0.840 for EVM and 0.884 for EVR. The ICC value between the EV diameters measured by EVR and EVM was higher than that of the visual assessment. The ICC value between EV diameter measurement by EVM and EVR was 0.991. Measurement times differed significantly across methods ( P<0.001): visual estimation 18.6±2.2 s (14.7-23.3 s), EVR 41.5±4.1 s (31.7-50.3 s), and EVM 170.8±26.4 s (129.3-229.3 s). Repeated measures analysis of variance (corrected using Greenhouse-Geisser) revealed significant differences in time across the three measurement methods [ F(1.033, 41.313)=1 233.800, P<0.001]. Subsequent Bonferroni post-hoc tests revealed significant differences in time between all method pairs ( P<0.001). Conclusion:EVR provides rapid, non-invasive EV diameter measurements with excellent agreement to EVM assessment, offering an efficient alternative to conventional techniques.

Background and Aims:Laparoscopic pancreaticoduodenectomy(LPD)has become a preferred approach for periampullary tumors,yet delayed gastric emptying(DGE)remains a frequent complication that hampers postoperative recovery.The nasojejunal feeding tube(NJT)is commonly used for early enteral nutrition,but its impact on DGE is controversial.This study aimed to evaluate whether intraoperative NJT placement increases the risk of DGE after LPD and to assess its influence on postoperative recovery outcomes.Methods:A retrospective cohort of 319 patients who underwent LPD at Provincial Hospital Affiliated to Shandong First Medical University from April 2017 to November 2023 was analyzed.Patients were divided into two groups based on intraoperative NJT placement(NJT group,n=200;non-NJT group,n=119).The incidence of DGE and postoperative outcomes were compared.Multivariate logistic regression and propensity score matching(PSM)were performed to identify independent risk factors for DGE.Results:The incidence of grade B/C DGE was significantly higher in the NJT group than in the non-NJT group(36.5%vs.21.8%,P=0.006).NJT placement was associated with longer postoperative hospital stay and higher hospitalization costs(both P<0.05).Multivariate analysis revealed intraoperative NJT placement(OR=1.960,95%CI=1.142-3.363,P=0.015)and intraoperative blood loss>400 mL(OR=1.921,95%CI=1.155-3.194,P=0.012)as independent risk factors for DGE.These findings were consistent after PSM.Conclusions:Prophylactic intraoperative NJT placement confers no additional benefit for postoperative recovery after LPD and is associated with a higher risk of DGE,prolonged hospitalization,and increased medical costs.Routine NJT placement should therefore be avoided,and individualized strategies should be adopted to minimize postoperative complications and enhance recovery.

Objective:To evaluate the performance of an artificial intelligence-based endoscopic virtual ruler (EVR) for non-invasive measurement of esophageal varices (EV) diameter.Methods:Patients with liver cirrhosis and EV hospitalized at the First Affiliated Hospital of Anhui Medical University between October 2022 and May 2023 were prospectively enrolled. EV diameter was measured using visual estimation, esophageal varix manometer (EVM), and EVR, with procedure times recorded. The intraclass correlation coefficient (ICC) was used to assess the consistency of EV diameter measurement obtained from the three methods, and repeated-measures ANOVA was used to compare differences in time measurements across three methods.Results:The study included 41 patients with liver cirrhosis and EV. Inter-observer ICC for visual estimation was 0.594, versus 0.840 for EVM and 0.884 for EVR. The ICC value between the EV diameters measured by EVR and EVM was higher than that of the visual assessment. The ICC value between EV diameter measurement by EVM and EVR was 0.991. Measurement times differed significantly across methods ( P<0.001): visual estimation 18.6±2.2 s (14.7-23.3 s), EVR 41.5±4.1 s (31.7-50.3 s), and EVM 170.8±26.4 s (129.3-229.3 s). Repeated measures analysis of variance (corrected using Greenhouse-Geisser) revealed significant differences in time across the three measurement methods [ F(1.033, 41.313)=1 233.800, P<0.001]. Subsequent Bonferroni post-hoc tests revealed significant differences in time between all method pairs ( P<0.001). Conclusion:EVR provides rapid, non-invasive EV diameter measurements with excellent agreement to EVM assessment, offering an efficient alternative to conventional techniques.

Drug testing involves many analytical instruments and test items, sample pretreatment is tedious, the industry's intelligence level remains low, making drug testing a labour-intensive job. However, in the era of Industry 4.0 intelligent manufacturing, intelligent transformation of the quality control (QC) laboratory has become the focus of industry. At the same time, driven by consistency evaluation of the quality and efficacy of generic drugs and the centralized procurement policies, pharmaceutical companies have intensified their competition, further stimulating the intrinsic demand for laboratory intelligence. Based on the current state and future trends of the pharmaceutical industry, this review discusses the development of a digital and automated QC laboratory. It points out the necessity of transitioning from the traditional centralized laboratory model to an intelligent, distributed quality control model to accommodate continuous manufacturing processes. At the same time, it also analyses the potential challenges in the implementation process and coping strategies, in order to provide relevant practitioners with ideas for building intelligent QC laboratories.

Natural compounds demonstrate unique therapeutic advantages for cancer treatment, primarily through direct tumor suppression or interference with the tumor microenvironment (TME). Glycyrrhizic acid (GL), a bioactive ingredient derived from the medicinal herb Glycyrrhiza uralensis Fisch., and its sapogenin glycyrrhetinic acid (GA), have been recognized for their ability to inhibit angiogenesis and remodel the TME. Consequently, the combination of GL with other therapeutic agents offers superior therapeutic benefits. Given GL's amphiphilic structure, self-assembly capability, and liver cancer targeting capacity, various GL-based nanoscale drug delivery systems have been developed. These GL-based nanosystems exhibit angiogenesis suppression and TME regulation properties, synergistically enhancing anti-cancer effects. This review summarizes recent advances in GL-based nanosystems, including polymer-drug micelles, drug-drug assembly nanoparticles (NPs), liposomes, and nanogels, for cancer treatment and tumor postoperative care, providing new insights into the anti-cancer potential of natural compounds. Additionally, the review discusses existing challenges and future perspectives for translating GL-based nanosystems from bench to bedside.
Animals ; Humans ; Antineoplastic Agents/therapeutic use* ; Glycyrrhizic Acid/therapeutic use* ; Liposomes/chemistry* ; Micelles ; Nanoparticles/chemistry* ; Neoplasms/pathology* ; Tumor Microenvironment/drug effects* ; Nanoparticle Drug Delivery System/therapeutic use*

With the advent of artificial intelligence (AI), machines are increasingly being used to complete complicated tasks, yielding remarkable results. Machine learning (ML) is the most relevant subset of AI in medicine, which will soon become an integral part of our everyday practice. Therefore, physicians should acquaint themselves with ML and AI, and their role as an enabler rather than a competitor. Herein, we introduce basic concepts and terms used in AI and ML, and aim to demystify commonly used AI/ML algorithms such as learning methods including neural networks/deep learning, decision tree and application domain in computer vision and natural language processing through specific examples. We discuss how machines are already being used to augment the physician's decision-making process, and postulate the potential impact of ML on medical practice and medical research based on its current capabilities and known limitations. Moreover, we discuss the feasibility of full machine autonomy in medicine.
Humans ; Artificial Intelligence ; Machine Learning ; Algorithms ; Neural Networks, Computer ; Medicine

Magnetic resonance imaging (MRI)-targeted prostate biopsy is the recommended investigation in men with suspicious lesion(s) on MRI. The role of concurrent systematic in addition to targeted biopsies is currently unclear. Using our prospectively maintained database, we identified men with at least one Prostate Imaging-Reporting and Data System (PI-RADS) ≥3 lesion who underwent targeted and/or systematic biopsies from May 2016 to May 2020. Clinically significant prostate cancer (csPCa) was defined as any Gleason grade group ≥2 cancer. Of 545 patients who underwent MRI fusion-targeted biopsy, 222 (40.7%) were biopsy naïve, 247 (45.3%) had previous prostate biopsy(s), and 76 (13.9%) had known prostate cancer undergoing active surveillance. Prostate cancer was more commonly found in biopsy-naïve men (63.5%) and those on active surveillance (68.4%) compared to those who had previous biopsies (35.2%; both P < 0.001). Systematic biopsies provided an incremental 10.4% detection of csPCa among biopsy-naïve patients, versus an incremental 2.4% among those who had prior negative biopsies. Multivariable regression found age (odds ratio [OR] = 1.03, P = 0.03), prostate-specific antigen (PSA) density ≥0.15 ng ml^-2 (OR = 3.24, P < 0.001), prostate health index (PHI) ≥35 (OR = 2.43, P = 0.006), higher PI-RADS score (vs PI-RADS 3; OR = 4.59 for PI-RADS 4, and OR = 9.91 for PI-RADS 5; both P < 0.001) and target lesion volume-to-prostate volume ratio ≥0.10 (OR = 5.26, P = 0.013) were significantly associated with csPCa detection on targeted biopsy. In conclusion, for men undergoing MRI fusion-targeted prostate biopsies, systematic biopsies should not be omitted given its incremental value to targeted biopsies alone. The factors such as PSA density ≥0.15 ng ml^-2, PHI ≥35, higher PI-RADS score, and target lesion volume-to-prostate volume ratio ≥0.10 can help identify men at higher risk of csPCa.
Male ; Humans ; Prostate/pathology* ; Prostatic Neoplasms/pathology* ; Prostate-Specific Antigen ; Magnetic Resonance Imaging/methods* ; Image-Guided Biopsy/methods* ; Retrospective Studies