Development of a machine learning-based model for predicting risk of early postoperative recurrence of hepatocellular carcinoma

BACKGROUND Surgical resection is the primary treatment for hepatocellular carcinoma(HCC).However,studies indicate that nearly 70%of patients experience HCC recurrence within five years following hepatectomy.The earlier the recurrence,the worse the prognosis.Current studies on postoperative recurrence primarily rely on postoperative pathology and patient clinical data,which are lagging.Hence,developing a new pre-operative prediction model for postoperative recurrence is crucial for guiding individualized treatment of HCC patients and enhancing their prognosis.AIM To identify key variables in pre-operative clinical and imaging data using machine learning algorithms to construct multiple risk prediction models for early postoperative recurrence of HCC.METHODS The demographic and clinical data of 371 HCC patients were collected for this retrospective study.These data were randomly divided into training and test sets at a ratio of 8:2.The training set was analyzed,and key feature variables with predictive value for early HCC recurrence were selected to construct six different machine learning prediction models.Each model was evaluated,and the bestperforming model was selected for interpreting the importance of each variable.Finally,an online calculator based on the model was generated for daily clinical practice.RESULTS Following machine learning analysis,eight key feature variables(age,intratumoral arteries,alpha-fetoprotein,preoperative blood glucose,number of tumors,glucose-to-lymphocyte ratio,liver cirrhosis,and pre-operative platelets)were selected to construct six different prediction models.The XGBoost model outperformed other models,with the area under the receiver operating characteristic curve in the training,validation,and test datasets being 0.993(95%confidence interval:0.982-1.000),0.734(0.601-0.867),and 0.706(0.585-0.827),respectively.Calibration curve and decision curve analysis indicated that the XGBoost model also had good predictive performance and clinical application value.CONCLUSION The XGBoost model exhibits superior performance and is a reliable tool for predicting early postoperative HCC recurrence.This model may guide surgical strategies and postoperative individualized medicine.

Letter to editor‘Non-invasive model for predicting high-risk esophageal varices based on liver and spleen stiffness’

predicting high-risk esophageal varices based on liver and spleen stiffness".Acute bleeding caused by esophageal varices is a life-threatening complication in patients with liver cirrhosis.Due to the discomfort,contraindications,and associated complications of upper gastrointestinal endoscopy screening,it is crucial to identify an imaging-based non-invasive model for predicting high-risk esophageal varices in patients with cirrhosis.

Seeing Is Believing:Directly Imaging Contractions of Artificial Molecular Muscles with Platinum Atom Markers

Artificial molecular muscles undergo well-controlled contractile and extensile motions upon external stimulation,leading to remarkable length changes.Evaluating such length changes at the molecular level is essential to the design of integrated artificial molecular muscles that mimic biological muscles.Taking advantage of the strong contrast of platinum(Pt)atoms in high-angle annular dark-field scanning transmission electron microscopy images,we imaged Pt-containing molecular[c2]daisy chains directly by employing metal atom markers.The length changes and associated conformational transformations of these newly developed artificial molecular muscles have been measured experimentally in combination with theoretical calculations.The contraction ratios of these two molecular muscles with the TEMPO or pyrene anchoring group were calculated to be 21.0%or 15.7%respectively,suggesting a substantial anchoring effect.This study demonstrates the experimental measurement of the length changes of artificial molecular muscles and provides a new avenue for investigating the motion of artificial molecular machines.

基于响应值中心加权卷积特征的图像检索算法

提出一种新颖的基于交叉熵的损失,将该损失应用于经典的卷积神经网络上获得了更优的嵌入空间。设计了一种基于响应值中心加权的卷积特征聚合算法处理神经网络得到的三维卷积特征,该算法通过当前位置响应值以及高斯中心计算获得特征图的空间加权系数,减少了三维卷积特征图降成一维图像特征描述子时的信息丢失,并实现了目标区域的增强。最后,将得到的图像特征描述子用于检索任务。在CUB-200-2011数据集上通过消融实验分别验证了损失函数和响应值中心加权算法的有效性。本文算法在Paris6k、Oxford5k、CUB-200-2011、CARS196四个数据集上较当前已有的检索方法获得了更高的准确率和召回率。

Structurally Diverse Sesquiterpenoids from Chloranthus henryi var.hupehensis and Their Anti-inflammatory Activities

In this study,27 new sesquiterpenoids:twenty monomers and seven dimers,with diverse structures,along with one known chlora-japonilide F(25)were isolated from the aerial parts of Chloranthus henryi var.hupehensis.Structurally,chlorahupetolides A(1)and B(2),two eudesmane-type merosesquiterpenoids with an undescribed C18 carbon framework,and chlorahupetene E(18)are the first example of dimers comprising two eudesmane sesquiterpenoids bridged by a four-membered ring in the Chloranthus.Their struc-tures were characterized by nuclear magnetic resonance(NMR),electronic circular dichroism(ECD),and X-ray diffraction analysis.In the RAW264.7 macrophages model of inflammation induced by LPS,compounds 12,18,and 25 significantly reduced NO production in a dose dependent manner and without cytotoxicity.The mRNA expression of COX-2 was considerably inhibited by treatments with compounds 12,18,and 25.