目的 构建并验证机械通气患儿肠内营养支持发生误吸的风险预测模型。方法 回顾性分析中山市博爱医院2021年3月—2023年3月儿童重症监护病房330例行机械通气并进行肠内营养的患儿临床资料,通过二元Logistic回归,获取机械通气患儿肠内营养支持发生误吸的预测因素,绘制列线图模型,并进行模型评价及验证。结果 330例机械通气患儿中,104例患儿发生误吸、226例未发生误吸。两组患儿在意识状态、机械通气方式、管饲量、胃残留量、胃管置入深度、促胃动力药、镇静剂等方面对比差异具有统计学意义（P<0.05）。二元Logistic结果显示,胃残留量、机械通气方式、管饲量、意识状态、胃管置入深度、促胃动力药、镇静剂是机械通气患儿肠内营养支持发生误吸的影响因素（P<0.05）。建模组AUC为0.810（95%CI：0.760~0.860）,Hosmer-Lemesh结果显示,χ²=3.245,P=0.846;外部验证组AUC为0.873（95%CI：0.831~0.914）,Hosmer-Lemesh结果显示,χ²=3.567,P=0.875。建模组和训练组DCA曲线大部分落于Y=0上方。建模组与外部验证组校准曲线均与参考曲线高度贴合,预测概率与实际概率接近,校准度良好。结论 基于胃残留量、机械通气方式、管饲量、意识状态、胃管置入深度、促胃动力药、镇静剂等7项指标构建的风险预测模型具有一定的临床价值,可作为医护人员识别肠内营养机械通气误吸高危患儿的工具。

Objective To establish and verify the risk prediction model of enteral nutritional aspiration in children with mechanical ventilation．Methods The clinical data of 330 children who underwent mechanical ventilation and enteral nutrition in the PICU of Zhongshan Boai Hospital from March 2021 to March 2023 were retrospectively analyzed．The independent predictive factors of enteral nutrition support aspiration in children with mechanical ventilation were obtained by binary Logistic regression,and the nomographic model was drawn,and the model was evaluated and verified. Results Among 330 children with mechanical ventilation,104 had aspiration and 226 did not．There were statistically significant differences between the two groups in consciousness state,mechanical ventilation mode,tube feeding amount,gastric residual amount,gastric tube insertion depth,gastric motivity drugs,sedatives,etc.（P<0.05）．Binary Logistic results showed that gastric residual amount,mechanical ventilation mode,tube feeding amount,state of consciousness,depth of gastric tube insertion,gastric motonics and sedatives were the influential factors of enteral nutritional aspiration in children with mechanical ventilation（P<0.05）．The AUC of the modeling group was 0.810（95%CI：0.760-0.860）,and the Hosmer-Lemesh result showed that χ²=3.245,P=0.846．The AUC of the external verification group was 0.873（95%CI：0.831-0.914）,and the Hosmer-Lemesh result showed that χ²=3.567,P=0.875．The DCA curves of modeling group and training group mostly were above Y=0．The calibration curves of the modeling group and the external verification group are highly fit to the reference curves,and the prediction probability was close to the actual probability,and the calibration degree was good．Conclusion sThe risk prediction model based on 7 indexes,including stomach residual amount,mechanical ventilation mode,tube feeding amount,state of consciousness,depth of gastric tube insertion,gastric motivity drug and sedative,with certain clinical value,and can be used as a tool for medical staff to identify children at high risk of enteral nutritional mechanical aspiration．

目的 研究单核细胞/高密度脂蛋白比值（MHR）对免疫球蛋白A肾病（IgAN）患者肾小球硬化严重程度的预测价值。方法 回顾性分析2016年1月—2022年6月龙岩人民医院收治的296例IgAN患者的临床和肾脏病理资料,参照牛津分型将患者分为无肾小球节段性硬化组（S0组）、有肾小球节段性硬化组（S1组）及球性硬化组;根据Katafuchi肾小球积分将患者分为低三分位组、中三分位组及高三分位组。比较不同肾小球硬化程度和不同Katafuchi肾小球积分患者的MHR水平,对MHR与Katafuchi肾小球积分的关系进行相关性分析,绘制受试者操作特征（ROC）曲线分析MHR对肾小球硬化程度的预测效能。结果 S1组和球性硬化组的单核细胞计数分别为（0.41±0.11）×10⁹ /L、（0.45±0.10）×10⁹ /L,均高于S0组的（0.30±0.06）×10⁹ /L,对比差异有统计学意义（t₁=10.381,P₁<0.001;t₂=12.169,P₂<0.001）,球性硬化组的HDL水平为（1.14±0.16）mmoL/L,低于S0组（1.26±0.24）mmoL/L（t₂=2.992,P₂=0.003）。S1组和球性硬化组的MHR为（0.36±0.04）、（0.44±0.05）,大于S0组的（0.24±0.02）,对比差异有统计学意义（t₁=37.852,P₁<0.001;t₂=42.037,P₂<0.001）,球性硬化组的MHR大于S1组（t₃=9.673,P₃<0.001）。中三分位组和高三分位组的单核细胞计数为（0.34±0.06）×10⁹ /L、（0.48±0.10）×10⁹ /L,高于低三分位组的（0.27±0.05）×10⁹ /L,对比差异有统计学意义（t₁=9.017,P₁<0.001;t₂=20.080,P₂<0.001）,高三分位组的单核细胞计数高于中三分位组（t₃=8.855,P₃<0.001）。高三分位组的HDL水平为（0.96±0.12）mmoL/L,低于低三分位组的（1.23±0.21）mmoL/L和中三分位组的（1.19±0.16）mmoL/L,对比差异有统计学意义（t₂=8.132,P₂<0.001;t₃=7.954,P₃<0.001）。高三分位组的MHR为（0.49±0.07）,大于低三分位组的（0.25±0.03）和中三分位组（0.26±0.08）,对比差异有统计学意义（t₂=35.382,P₂<0.001;t₃=15.146,P₃<0.001）。相关性分析显示,单核细胞与Katafuchi肾小球积分呈正相关（r=0.58,P<0.05）,HDL与Katafuchi肾小球积分呈负相关（r=-0.52,P<0.05）,MHR与Katafuchi肾小球积分呈正相关（r=0.66,P<0.05）。MHR预测肾小球节段性硬化的曲线下面积（AUC）为0.609（95%CI：0.325~0.917）,此时截断值为0.29,灵敏度为68.42%,特异度为65.45%。MHR预测球性硬化的AUC为0.735（95%CI：0.527~1.001）,此时截断值为0.40,灵敏度为73.08%,特异度为66.14%。结论 MHR在预测IgAN患者肾小球硬化程度方面具有潜能。

Objective To explore predictive value of the monocyte / high-density lipoprotein ratio（MHR）on the severity of glomerulosclerosis in patients with immunoglobulin A nephropathy（IgAN）．Methods The clinical and renal pathological data of 296 IgAN patients admitted to Longyan People's Hospital from January 2016 to June 2022 were analyzed retrospectively,and the patients were divided into no segmental sclerosis group（S0）,segment sclerosis group（S1）and glomerular sclerosis group according to Oxford classification;the patients were divided into low group,middle group and high group according to Katafuchi score．MHR levels in patients with different degrees of glomerular sclerosis and different Katafuchi score were compared,the relationship between MHR and Katafuchi glomerular integration was analyzed,and ROC curves were drawn to analyze the predictive efficacy of MHR on the degree of glomerular sclerosis．Results Monocyte counts in the S1 and glomerular sclerosis groups [（0.41±0.11）10⁹ /L,（0.45±0.10）10⁹ /L] were all significantly higher than the S0 group（0.30±0.06）10⁹ /L,with statistically significant difference（t₁=10.381,P₁<0.001,t₂=12.169,P₂<0.001）．The HDL level（1.14±0.16）mmoL / L was significantly lower than that in the S0 group（1.26±0.24）mmoL / L（t₂=2.992,P₂=0.003）．The MHR in S1 and glomerular sclerosis group[（0.36±0.04）,（0.44±0.05）] were significantly greater than S0（0.24±0.02）（t₁=37.852,P₁<0.001,t₂=42.037,P₂<0.001）,and the MHR in glomerular sclerosis group was significantly greater than that of S1（t₃=9.673,P₃<0.001）．The monocyte counts of middle and high group[（0.34±0.06）10⁹ /L,（0.48±0.10）10⁹ /L] were significantly higher than the low group（0.27±0.05）10⁹ /L（t₁=9.017,P₁<0.001;t₂=20.080,P₂<0.001）and high group was significantly higher than middle group（t₃=8.855,P₃<0.001）．The HDL level of high group（0.96±0.12）mmoL/L was significantly lower than the low group（1.23±0.21）mmoL/L and middle group（1.19±0.16）mmoL/L,with statistically significance（t₂=8.132,P₂<0.001,t₃=7.954,P₃<0.001）．The MHR（0.49±0.07）in the high group was significantly greater than the low group（0.25±0.03）and middle group（0.26±0.08）,with statistically significance（t₂=35.382,P₂<0.001,t₃=15.146,P₃<0.001）．Correlation analysis showed that monocytes were positively correlated with Katafuchi score（r=0.58,P<0.05）,HDL negatively with Katafuchi score（r=-0.52,P<0.05）,and MHR positively with Katafuchi score（r=0.66,P<0.05）．The AUC of MHR predicting segmental sclerosis was 0.609（95%CI：0.325~0.917）,when the cut-off was 0.29,sensitivity was 68.42% and specificity of 65.45%．The AUC of MHR predicting glomerulosclerosis was 0.735（95%CI：0.527~1.001）,when the cut-off was 0.40,sensitivity was 73.08% and specificity was 66.14%．Conclusions MHR has the potential in predicting the degree of glomerulosclerosis in IgAN patients．

目的 探讨动态对比增强磁共振成像（DCE-MRI）多参数定量特征对乳腺癌腋窝淋巴结转移（ALNM）风险的预测价值。方法 回顾性收集2020年3月—2022年11月在佛山市高明区人民医院经手术病理确诊的155例乳腺癌患者临床资料,根据患者是否发生ALNM分为ALNM 组（n=39）和无ALNM 组（n=116）。采用单因素分析和多因素Logistic回归分析乳腺癌发生ALNM的影响因素。结果 ALNM组和无ALNM 组患者的肿块质地、肿块直径、肿块部位、肿块形状、肿块内部强化特征等指标比较差异无统计学意义（t/χ²=2.249、0.977、1.369、0.524、2.158,P>0.05）。两组患者肿块表观扩散系数（ADC）值、腋窝淋巴结（ALN）短径、肿块边缘、动态增强时间-信号强度曲线（TIC）曲线等指标比较,差异有统计学意义（t/χ²=6.573、9.873、29.441、2.031,P<0.05）。二元Logistic回归模型结果显示,肿块ADC值、ALN 短径（≥5 mm）、TIC曲线（流出型）为乳腺癌ALNM发生的危险因素（OR=0.251、0.106、0.002,P<0.05）。结论 DCE-MRI多参数定量特征中,乳腺癌患者的肿块ADC值低、ALN 短径（≥5 mm）、TIC曲线（流出型）为乳腺癌ALNM发生的危险因素。

Objective To investigate the predictive value of multi-parameter quantitative features of dynamic contrast-enhanced magnetic resonance imaging（DCE-MRI）in the risk of axillary lymph node metastasis（ALNM）in breast cancer．Methods Clinical data of 155 patients with breast cancer diagnosed by surgery and pathology in Foshan Gaoming District People's hospital from March 2020 to November 2022 were retrospectively collected．According to whether the patients had ALNM,they were divided into ALNM group（n=39）and non-ALNM group（n=116）．Univariate analysis and multiple Logistic regression models were used to explore the influencing factors of ALNM in breast cancer．Results There was no significant difference in mass texture,mass diameter,mass location,mass shape and internal enhancement between the ALNM group and the non-ALNM group（t/χ²=2.249,0.977,1.369,0.524,2.158,P>0.05）．There were significant differences in ADC value,ALN short diameter,tumor margin and TIC curve between the two groups（t/χ²=6.573,9.873,29.441,2.031,P<0.05）．Binary Logistic regression model showed that ADC value,ALN short diameter（≥5 mm）and tumor margin（blur）were risk factors for the occurrence of breast cancer ALNM（OR=0.251,0.106,0.002,P<0.05）．Conclusions Among the multi-parameter quantitative features of DCE-MRI,the ADC value of breast cancer,the short diameter of ALN（≥5 mm）,and the edge of the tumor（blur）are the risk factors for the occurrence of ALNM in breast cancer．

目的 探讨血沉(ESR)对川崎病(KD)冠状动脉病变(CAL)的预测价值。方法 收集 2017 年 5 月— 2021 年 6 月收治入院的KD患儿的临床资料,分析ESR对KD患儿CAL发生的预测作用。结果 纳入272例KD患儿,70例KD患儿合并CAL,202例KD患儿无CAL。KD患儿的ESR升高。单因素分析提示CAL组的ESR低于非CAL组,差异有统计学意义(P＜0.05)。ROC曲线下的面积为0.586,提示ESR可作为CAL的预测因子(P＜0.05)。ESR预测川崎病冠脉病变发生的临界值为76.5 mm/h。二分类Logistic回归分析显示,ESR是KD患儿发生CAL的独立危险因素(P＜0.05),当ESR＜76.5 mm·h^-1时,KD患儿CAL发生风险增加 (OR=2.38,95% CI: 1.25~4.53)。结论 KD急性期的ESR水平可用于预测KD患儿CAL的发生,ESR＜76.5 mm·h^-1时,提示KD患儿可能会出现CAL。

Objective To investigate the predictive value of erythrocyte sedimentation rate (ESR) on coronary artery lesions (CAL) in Kawasaki disease (KD). Methods Clinical data of children with KD admitted to the hospital from May 2017 to June 2021 were collected to analyze the predictive value of ESR on the occurrence of CAL in children with KD. Results Of the 272 enrolled children with KD, 70 children with CAL and 202 children without CAL. ESR was significantly higher in children with KD. Univariate analysis suggested that ESR was lower in the CAL group than in the non-CAL group, and the difference was statistically significant (P<0.05). The area under the ROC curve was 0.586, suggesting that ESR could be a predictor of CAL (P<0.05). The critical value of ESR for predicting the occurrence of CAL in KD was 76.5 mm/h. Dichotomous Logistic regression analysis showed that ESR was an independent risk factor for the development of CAL in children with KD (P<0.05), and the risk of CAL in children with KD was significantly increased when ESR was<76.5 mm/h. (OR = 2.38, 95% CI: 1.25-4.53). Conclusions ESR levels in the acute phase of KD can be used to predict the development of CAL in children with KD, and ESR <76.5 mm/h suggests that children with KD may develop CAL.

目的 研究CXC趋化因子配体4（CXCL4）、基质金属蛋白酶-9（MMP-9）、微小RNA-24（miR-24）对急性脑梗死（ACI）早期神经功能恶化（END）的预测价值。方法 分别选择2020年1月—2021年6月我院收治的30例ACI早期END患者（ACI+END组）,30例单纯ACI患者（ACI组）,同时期30例健康人群作为对照组,检测受试者CXCL4、MMP-9、miR-24表达情况及存在的相关性,分析血清CXCL4、MMP-9、miR-24表达情况与ACI早期发生END的关系。结果 CXCL4、MMP-9水平在对照组、ACI组、ACI+END组中依次升高,miR-24相对表达量依次降低（P<0.05）。血清CXCL4、MMP-9、miR-24水平在轻度、重度患者中呈升高趋势,miR-24相对表达呈降低趋势（P<0.05）。Logistic回归分析血清CXCL4、MMP-9、miR-24表达异常与ACI早期发生END独立相关（P<0.05）。经Pearson相关性分析发现,CXCL4与MMP-9之间呈正相关（r=0.584,P=0.001）;CXCL4、miR-24之间呈负相关（r=-0.569,P=0.001）,MMP-9、miR-24之间呈负相关（r=-0.567,P=0.001）。ROC曲线显示,与CXCL4、MMP-9、miR-24单项预测相比,三项联合对ACI的关系及对早期END的预测价值较高（P<0.05）。结论 CXCL4、MMP-9、miR-24在ACI发生END时出现异常表达,检测CXCL4、MMP-9、miR-24水平对ACI早期END具有一定预测价值,可尽早制定相关措施干预,提高治疗效率。

Objective To study the predictive value of CXC chemokine ligand 4（CXCL4）,matrix metalloproteinase-9（MMP-9）,microRNA-24（miR-24）in early neurological deterioration（END）of acute cerebral infarction（ACI）．Methods A total of 30 patients with ACI early END（ACI+END group）and 30 patients with ACI only（ACI group）who were admitted to our hospital from January 2020 to June 2021 were selected,and 30 healthy people（control group）who underwent physical examination in our hospital during the same period were selected．Expressions of CXCL4,MMP-9,and miR-24 were detected and their correlations were analyzed,and the relationship between the expressions of serum CXCL4,MMP-9,miR-24 and the early occurrence of END in ACI were analyzed．Results The levels of CXCL4 and MMP-9 were increased in the control group,ACI group and ACI+END group in turn,and the relative expression of miR-24 was decreased in turn,and the differences among the groups were statistically significant（P<0.05）．The levels of serum CXCL4,MMP-9,and miR-24 increased in mild and severe patients,while the relative expression of miR-24 decreased,and the differences between groups were statistically significant（P<0.05）．Logistic regression analysis showed that abnormal expressions of serum CXCL4,MMP-9 and miR-24 were independently correlated with the early occurrence of END in ACI（P<0.05）．After Pearson correlation analysis,it was found that there was a positive correlation between CXCL4 and MMP-9（r=0.584,P=0.001）,a negative correlation between CXCL4 and miR-24（r=-0.569,P=0.001）;a negative correlation between MMP-9 and miR-24（r=-0.567,P=0.001）．The ROC curve showed that compared with the single prediction of CXCL4,MMP-9 and miR-24,the predictive value of the combined prediction on ACI and early END were higher（P<0.05）．Conclusions CXCL4,MMP-9,and miR-24 are abnormally expressed in ACI when END occurs．Detection of CXCL4,MMP-9,and miR-24 levels has certain predictive value for early END of ACI,and relevant measures can be formulated as soon as possible to improve treatment efficiency．

目的 探讨尿液前列腺特异性抗原（u-PSA）预测老年良性前列腺增生（BPH）发生急性尿潴留（AUR）的价值。方法 选取东莞市中医院100例老年BPH患者（2020年1月—2021年4月）进行回顾性研究,均口服盐酸坦索罗辛+非那雄胺片治疗,随访1年,记录AUR发生情况,据此分为AUR组、非AUR组。比较2组一般资料,Logistic回归模型分析老年BPH发生AUR的危险因素,受试者工作特征（ROC）分析前列腺体积（PV）、u-PSA对老年BPH发生AUR的预测价值。结果 100例老年BPH患者AUR发生率为26%;AUR组u-PSA水平高于非AUR组,PV大于非AUR组（P<0.05）;Logistic回归模型分析,u-PSA水平及PV增高是老年BPH患者发生AUR的独立危险因素（P<0.05）;ROC曲线分析,u-PSA预测AUR的AUC=0.897,高于AUC_PV（P<0.05）。结论 u-PSA可作为老年BPH继发AUR的量化评估指标,有利于临床早期筛查、诊断,采取针对性干预措施,改善预后。

Objective To investigate the value of urinary prostate-specific antigen（u-PSA）in predicting acute urinary retention（AUR）in elderly patients with benign prostatic hyperplasia（BPH）．Methods A total of 100 elderly patients with BPH in our hospital（from January 2020 to April 2021）were selected for a retrospective study,all of whom were treated with oral tamsulosin hydrochloride + finasteride tablets,followed up for 1 year,and the occurrence of AUR was recorded．The patients were divided into AUR group and non-AUR group．The general data of the two groups were compared．Logistic regression model was used to analyze the risk factors of AUR in elderly BPH patients,and receiver operating characteristic（ROC）was used to analyze the predictive value of prostate volume（PV）and u-PSA for AUR occurrence．Results The incidence of AUR in 100 elderly patients with BPH was 26%;the level of u-PSA in the AUR group was higher than that in the non-AUR group,and the PV was greater than that in the non-AUR group（P<0.05）．Increased PV was an independent risk factor for AUR in elderly patients with BPH（P<0.05）．ROC curve analysis showed that the AUC of u-PSA for predicting AUR was 0.897,which was higher than that of PV（P<0.05）．Conclusions u-PSA can be used as a quantitative evaluation index for AUR secondary to BPH in the elderly,which is conducive to early clinical screening and diagnosing,and taking targeted intervention measures to improve prognosis.

目的 基于影像组学方法,探讨多层螺旋CT（MSCT）四期增强扫描单一/不同期相及不同容积感兴趣区（VOI）的选择,在术前预测原发性肝细胞癌（HCC）微血管侵犯（MVI）中的价值。方法 回顾性收集88例经手术病理证实为HCC并行术前MSCT四期增强扫描的患者,其中包括47例MVI阳性患者和41例MVI阴性患者。在MSCT增强扫描的动脉早期、动脉晚期、门静脉期及延迟期图像中手动逐层勾画肿瘤ROI,获得瘤体容积感兴趣区VOI（Vt）,然后基于计算机自动膨胀算法将Vt外扩10 mm获得瘤体及瘤周VOI（Vt+Vp）。使用Pyradiomics软件分别从Vt和Vt+Vp中提取影像组学特征,随后采用15种特征选择方法和10种分类器构建150个预测模型,并通过十折交叉检验以验证模型的效能。使用准确度、敏感度、特异度、受试者工作特性曲线下面积（AUC）评估模型的效能,并比较性能最优的前三个预测模型。结果 MSCT四期增强扫描图像中预测HCC MVI状态的影像组学模型在门静脉期的表现优于其它期相及各期相的不同组合,其中最大的AUC值在Vt和Vt+Vp两种ROI中分别为0.768和0.782。此外,基于Vt+Vp的影像组学模型对MVI的预测效能优于基于Vt的影像组学模型,基于Vt+Vp性能最优的预测模型的AUC值、准确度、敏感度和特异度分别0.782、0.728、0.745和0.705。结论 采用影像组学方法术前无创性预测HCC MVI状态首选增强扫描的门静脉期,ROI首选瘤体联合瘤周10 mm区域。

Objective To investigate the value of single or different phases of contrast-enhanced multi-slice spiral CT（MSCT）in different volumetric regions of interest（ROI）to preoperatively predict the state of microvascular invasion in primary hepatocellular carcinoma（HCC）based on radiomics methods．Methods A total of 88 patients with HCC confirmed by surgical pathology who underwent preoperative MSCT quadruple-enhanced scan were retrospectively recruited,including 47 MVI-positive patients and 41 MVI-negative patients．The ROI was manually delineated slice-by-slice in the early arterial phase,late arterial phase,portal venous phase,and equilibrium phase of enhanced MSCT images to obtain the volume of tumor VOI（Vt）,and then Vt was expanded by 10 mm through the computer expansion algorithm automatically to obtain the volume of tumor and peritumor（Vt+Vp）．Pyradiomics software was used to extract radiomic features from Vt and Vt+Vp,followed by 150 discriminant models constructed with 150 feature selection methods and 10 classifiers,and then 10-fold cross-validation was used to evaluate the performance of these models．Using accuracy,sensitivity,specificity,area under the receiver operating characteristic curve（AUC）to assess model performance．The top three predictive models with the best performance were also compared．Results The radiomics model for predicting HCC MVI status in portal venous phase among quadruple-enhanced MSCT images outperformed other phases and different combinations of phases,achieving the highest AUC values of 0.768 and 0.782 in Vt and Vt+Vp respectively．In addition,the prediction performance of the radiomics model based on Vt+Vp was superior to models based on Vt．AUC value,accuracy,sensitivity,and specificity of the model with the best performance based on Vt+Vp were 0.782,0.728,0.745 and 0.705 respectively．Conclusions Radiomics models based on the portal venous phase of contrast-enhanced MSCT and tumor combined with the 10mm peritumoral area were more recommended to be employed to preoperative non-invasively predict the state of HCC MVI．

目的 基于随机森林方法构建甲状腺功能减退（简称甲减）患病风险预测模型。方法 从MIMIC-IV数据库纳入5 735名甲减患者为病例组,4 803名非甲减患者为对照组,基于随机森林模型进行建模。同时利用逻辑回归、贝叶斯正则化神经网络、XGBoost作为比较模型。最后用准确率、F1分数、精确率、召回率、特异性以及AUC值评价四个机器学习模型性能。结果 随机森林模型准确率为0.85,F1分数为0.84,精确率为0.84,召回率为0.84,特异性为0.86,AUC值为0.91。在该模型中,促甲状腺激素、年龄、绝对淋巴细胞计数、血液中红细胞数、中性白细胞、性别、碱性磷酸酶、丙氨酸氨基转移酶、嗜酸性粒细胞绝对计数、尿素氮为甲减患者诊断重要性排前10的指标。结论 采用随机森林方法构建的甲减患病预测模型为甲减的早期诊断有潜在应用价值。

Objective To construct a risk prediction model for hypothyroidism based on the random forest model．Methods A total of 5 735 hypothyroidism patients were included from the MIMIC-IV database as the case group, and 4 803 non-hypothyroidism patients were included as the control group．Random forest models were constructed for both groups, and logistic regression, Bayesian regularized neural network, and XGBoost were used as comparative models．The performance of the four machine learning models was evaluated using accuracy, F1 score, precision, recall, specificity, and AUC value．Results The random forest model had an accuracy of 0.85, an F1 score of 0.84, a precision of 0.84, a recall of 0.84, a specificity of 0.86, and an AUC value of 0.91．In this model, thyroid-stimulating hormone, age, absolute lymphocyte count, red blood cell count in blood, neutrophil, gender, alkaline phosphatase, aspartate aminotransferase, absolute eosinophil count, and blood urea nitrogen were the top 10 indicators for diagnosing hypothyroidism patients．Conclusions The hypothyroidism disease prediction model constructed using the random forest method has potential application value for the early diagnosis of hypothyroidism．

目的 探讨亚实性肺腺癌结节CT征象在预测侵袭风险中的价值。方法 选取经胸部CT检查发现并经病理证实为肺腺癌的亚实性结节共76个病灶,病灶分为侵袭前组与侵袭组2组,分析结节CT征象,行统计学分析,明确亚实性肺腺癌结节侵袭性的预测因素。结果 侵袭前组（包括原位腺癌、微浸润性腺癌）共31例;侵袭组[浸润性腺癌（IAC）]共45例。2组在最大平均直径、结节平均CT值、结节形状、分叶、毛刺、胸膜牵拉征、有无实性成分、肿瘤微血管CT成像征间比较差异有统计学意义（P<0.05）,而在空泡征、空气支气管征方面比较差异无统计学意义（P>0.05）。结节最大直径预测浸润性腺癌侵袭性界值为13.63 mm,敏感度、特异度分别为68.9%、96.8%,AUC为0.885;平均CT值预测IAC侵袭性界值为-528 HU,敏感度、特异度分别为84.4%、83.9%,AUC为0.867。回归分析显示病灶最大平均直径（OR=2.015、P=0.01）可以作为浸润性肺腺癌结节的独立预测因子。结论 亚实性肺腺癌结节侵袭前组和侵袭组在最大直径、平均CT值、形状、分叶、毛刺、微血管征存在差异;结节的最大平均直径是IAC的独立预测因子。

Objective To investigate the value of CT signs in predicting the risk of invasion of subsolid lung adenocarcinoma nodules．Methods A total of 76 subsolid pulmonary adenocarcinoma nodules detected by chest CT and pathologically confirmed were selected,and the lesions were divided into pre-invasion group and invasion group．CT signs of nodules were analyzed and statistical analysis was performed to determine the predictive factors of invasion of subsolid pulmonary adenocarcinoma nodules．Results There were 31 cases in the pre-invasion group（including adenocarcinoma in situ and minimally invasive adenocarcinoma）and 45 cases ininvasion group[invasive adenocarcinoma（IAC）]．There were significant differences in maximum mean diameter,mean CT value of nodule,nodule shape,lobule,burr,pleural stretch sign,solid component and microvascular CT imaging signs between the two groups（P<0.05）,but there were no significant differences in vacuole sign and air bronchial sign between the two groups（P>0.05）．The maximum diameter of nodules predicted the invasion boundary of invasive adenocarcinoma was 13.63 mm,the sensitivity and specificity were 68.9% and 96.8%,respectively,and the AUC was 0.885．The mean CT value predicted the invasive boundary of IAC was -528 HU,the sensitivity and specificity were 84.4% and 83.9%,respectively,and the AUC was 0.867．Regression analysis showed that the maximum mean lesion diameter（OR=2.015,P=0.01）was an independent predictor of infiltrating lung adenocarcinoma nodules．Conclusions There were differences in the maximum diameter,mean CT value,shape,lobed,burr and microvascular signs between the pre-invasion group and the invasion group．The maximum mean diameter of nodules was an independent predictor of IAC．

目的 探讨子宫瘢痕的超声弹性成像结合厚度分析对剖宫产后再妊娠产妇子宫破裂的预测应用。方法 选择2020年1月—2021年12月在中山市中医院分娩的剖宫产术后再次妊娠经阴道分娩（VBAC）产妇作为研究对象。根据纳入和排除标准,共纳入子宫破裂的VBAC产妇32例、非子宫破裂的VBAC产妇90例。通过住院病历信息系统查询研究对象的基本信息及其在妊娠晚期（≥37周）用B超对研究对象行子宫瘢痕厚度和弹性的测量结果,采用受试者工作特征曲线（ROC）曲线分析子宫瘢痕厚度和弹性SI值对子宫破裂的预测作用。结果 子宫破裂组中年龄>35岁、妊娠>2次、与上次剖宫产间隔<2年、新生儿体质量≥3 kg、单层缝合者的比例高于非子宫破裂组（P<0.05）。122例产妇子宫瘢痕厚度的均值为（3.42±0.49）mm,SI的均值为（2.57±0.45）。ROC曲线分析结果显示：子宫瘢痕厚度单独预测子宫破裂的曲线下面积（AUC）为0.805（95%CI：0.730~0.880,P<0.05）,cut off值为3.05 mm,灵敏度为0.726,特异度为0.910,约登指数为0.636;子宫瘢痕SI单独预测子宫破裂的AUC为0.730（95%CI：0.635~0.824,P<0.05）,cut off值为2.11,灵敏度为0.767,特异度为0.781,约登指数为0.548;子宫瘢痕厚度联合预测子宫破裂的AUC为0.874（95%CI：0.812~0.937,P<0.01）,灵敏度为0.875,特异度为0.811,约登指数为0.686。子宫瘢痕厚度结合子宫瘢痕SI值预测子宫破裂的AUC高于单独使用子宫瘢痕厚度（Z=7.611,P=0.041）和子宫瘢痕SI值（Z=25.864,P=0.025）。结论 子宫瘢痕的超声弹性成像SI值联合子宫厚度可有效提高超声对于VBAC产妇子宫破裂的预测效能,具有一定的应用意义。

Objective To study the application of ultrasound elasticity imaging combined thickness analysis of uterine scar in predicting uterine rupture in women pregnant after cesarean section．Methods Pregnant women with vaginal birth after cesarean（VBAC）from January 2020 to December 2021 in Zhongshan Hospital of Traditional Chinese Medicine were selected as the research subjects．A total of 32 VBAC parturients with uterine rupture and 90 VBAC parturients without uterine rupture were included according to the inclusion and exclusion criteria．The basic information of the subjects was queried through the medical record information system of the hospital．In the third trimester（≥37 weeks）,the thickness and elasticity of uterine scar were measured by ultrasound,and the predictive effect of uterine scar thickness and elastic SI value on uterine rupture was analyzed by ROC curve．Results Chi-square test showed that the incidence of uterine rupture was higher in patients with age>35 years,pregnancy>2 times,interval from last cesarean section<2 years,newborn weight≥3kg,and the proportion of uterine rupture in single suture was higher than that in double suture（P<0.05）．The mean uterine scar thickness of 122 subjects was（3.42±0.49）mm,and the mean SI was（2.57±0.45）．The area under curve（AUC）of uterine scar thickness alone for predicting uterine rupture was 0.805（95%CI：0.730-0.880,P<0.05）,the cut off value was 3.05 mm,the sensitivity was 0.726,the specificity was 0.910,and the Youden coefficient was 0.636 by ROC curve analysis．The AUC of uterine scar SI alone for predicting uterine rupture was 0.730（95%CI：0.635-0.824,P<0.05）,the cut off value was 2.11,the sensitivity was 0.767,the specificity was 0.781,and the Youden coefficient was 0.548 by ROC curve analysis．The AUC of uterine scar thickness combination for predicting uterine rupture was 0.874（95%CI：0.812-0.937,P<0.01）,the sensitivity was 0.875,the specificity was 0.811,and the Youden coefficient was 0.686 by ROC curve analysis．The AUC predicted by uterine scar thickness combined with uterine scar SI value was higher than that predicted by uterine scar thickness alone（Z=7.611,P=0.041）and uterine scar SI value（Z=25.864,P=0.025）．Conclusions Elastic SI value of ultrasound imaging of uterine scar combined with uterine thickness can effectively improve the prediction efficiency of ultrasound for VBAC maternal uterine rupture,which has certain application significance,but further demonstration is still needed．