Eosinophilic solid and cystic renal cell carcinoma (ESC-RCC) is rare and often misdiagnosed as clear cell renal cell carcinoma (ccRCC). Therefore, a CT-based scoring system was developed to improve differential diagnosis. Retrospectively, 25 ESC-RCC and 176 ccRCC cases, were collected. The two groups were matched on a 1:2 basis using the propensity-score-matching (PSM) method, with matching factors including sex and age. Finally, 25 ESC-RCC and 50 ccRCC cases were included and randomly divided into a training cohort (52 cases) and a validation cohort (23 cases). Logistic regression identified significant factors, constructed the primary model, and assigned weights for the scoring model. Diagnostic performance was compared using receiver operating characteristic curves, dividing points into three intervals. Multifactorial logistic regression identified three independent factors: intra-tumour necrosis (3 points), degree of corticomedullary phase (CMP) enhancement (3 points), and pseudocapsule (2 points). The primary model’s area under the curve (AUC) value was 0.954 (95% confidence interval [CI]: 0.857–0.993, P < 0.001), with 85.7% sensitivity and 94.1% specificity. The scoring model’s AUC value for the training cohort was 0.950 (95% CI: 0.852–0.991, P < 0.001), with 77.1% sensitivity and 100% specificity at a cut-off of 4 points. The validation cohort’s AUC was 0.942 (95% CI: 0.759–0.997, P < 0.001). The scoring system intervals were: ≥0 to < 2 points, ≥ 2 to ≤ 3 points, and > 3 to ≤ 8 points. Higher scores correlated with increased ccRCC incidence and decreased ESC-RCC incidence.The limitation of this study is the small sample size. A CT-based scoring system effectively differentiates ESC-RCC from ccRCC.
BackgroundEndoscopic treatments for non-ampullary superficial duodenal lesions (NASDLs) are yet to be standardized. Endoscopic submucosal dissection (ESD) for NASDLs demands advanced techniques and a long procedure time to prevent perforation and bleeding. Precutting endoscopic mucosal resection (EMR) is a technical modification of ESD that overcomes the limitations of ESD. This study aimed to compare the efficacy and safety of precutting EMR versus ESD for NASDLs.MethodsWe conducted a retrospective analysis of patients with NASDLs treated with either precutting EMR or ESD from January 2015 to March 2023.ResultsA total of 90 patients with NASDLs were analyzed, with 44 patients in the precutting EMR group and 46 patients in the ESD group. The endoscopic procedure achieved satisfactory outcomes in both groups, with en block resection rate of 100.0 %. The R0 resection rates in the precutting EMR and ESD groups were 95.5 % and 93.5 %, respectively. No delayed perforation occurred postoperatively in either group. There were no significant differences between the two groups in age, gender, lesion location, layer of lesion origin, macroscopic type, and lesion size. The procedure time was significantly shorter in the precutting EMR group than in the ESD group (22.9 ± 7.1 min vs 36.0 ± 10.6 min, p<0.001). The intraoperative perforation rate was significantly lower in the precutting EMR group compared to ESD group (4.5% vs 19.6 %, p = 0.030).ConclusionsPrecutting EMR is comparable to ESD for NASDLs, demonstrating a lower intraoperative perforation rate and shorter procedure time compared to ESD.
ABSTRACTBackgroundEndoscopic submucosal dissection (ESD) is a widely utilized treatment for early esophageal cancer. However, the rising incidence of postoperative esophageal stricture poses a significant challenge, adversely affecting patients' quality of life and treatment outcomes. Developing precise predictive models is urgently required to enhance treatment outcomes.Materials and MethodsThis study retrospectively analyzed clinical data from 124 patients with early esophageal cancer who underwent ESD at Ningbo Medical Center Lihuili Hospital. Patients were followed up to assess esophageal stricture incidence. Binary logistic regression analysis was used to identify factors associated with post‐ESD esophageal stricture. A novel nomogram prediction model based on Systemic Immune‐inflammation Index (SII) was constructed and evaluated using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA).ResultsROC curve analysis showed that the optimal value of SII for predicting esophageal stricture was 312.67. Both univariate and multivariate analyses identified lesion infiltration depth (< M2 vs. ≥ M2, p = 0.002), lesion longitudinal length (< 4 cm vs. ≥ 4 cm, p = 0.008), circumferential resection range (< 0.5, 0.5–0.75, ≥ 0.75, p = 0.014), and SII (< 312.67 vs. ≥ 312.67, p = 0.040) as independent risk factors for post‐ESD esophageal stricture. A novel nomogram prediction model incorporating these four risk factors was developed. Validation using ROC curve analysis demonstrated satisfactory model performance, while calibration curves indicated good agreement between model‐predicted risk and observed outcomes.ConclusionWe successfully constructed a novel nomogram prediction model based on SII, which can accurately and intuitively predict the occurrence of esophageal stricture after ESD, providing guidance for clinicians and improving treatment outcomes.
Introduction: Liver transplant (LT) recipients were at a high risk of infection during the coronavirus disease 2019 (COVID-19) pandemic. Our purpose was to compare the clinical characteristics of severe and non-severe groups of LT recipients with COVID-19, and to analyze their risk factors for severe disease. Methodology: 79 LT recipients with COVID-19 were divided into a non-severe group (n = 60) and a severe group (n = 19), and differences in clinical characteristics, laboratory tests, and chest computed tomography (CT) performance were analyzed. Logistic regression was used to identify risk factors with severe COVID-19. Receiver operating characteristic (ROC) curves were plotted and the area under curve (AUC) values were calculated to assess the predictive value for severe COVID-19. Results: Age was statistically different (p < 0.001) between the two groups. The difference in neutrophil-to-lymphocyte ratio (NLR), serum creatinine (Scr), D-dimer, urea, C-reactive protein (CRP), lactate dehydrogenase (LDH), and the number of lung segments involved in inflammation between the two groups were statistically significant (p < 0.05). The results revealed that age (OR = 1.255, 95% CI 1.079-1.460), NLR (OR = 1.172, 95% CI 1.019-1.348), and Scr (OR = 1.041, 95% CI 1.016-1.066) were independent risk factors for severe COVID-19. The ROC results showed that high values for age, NLR and Scr predicted severe COVID-19, with AUC values of 0.775, 0.841 and 0.820, respectively, and 0.925 for the three factors combined. Conclusions: Advanced age, and elevated NLR and Scr are independent risk factors for severe COVID-19 in LT recipients.
This study aimed to develop a tumor radiomics quality and quantity model (RQQM) based on preoperative enhanced CT to predict early recurrence after radical surgery for colorectal liver metastases (CRLM). A retrospective analysis was conducted on 282 cases from 3 centers. Clinical risk factors were examined using univariate and multivariate logistic regression (LR) to construct the clinical model. Radiomics features were extracted using the least absolute shrinkage and selection operator (LASSO) for dimensionality reduction. The LR learning algorithm was employed to construct the radiomics model, RQQM (radiomics-TBS), combined model (radiomics-clinical), clinical risk score (CRS) model and tumor burden score (TBS) model. Inter-model comparisons were made using area under the curve (AUC), decision curve analysis (DCA) and calibration curve. Log-rank tests assessed differences in disease-free survival (DFS) and overall survival (OS). Clinical features screening identified CRS, KRAS/NRAS/BRAF and liver lobe distribution as risk factors. Radiomics model, RQQM, combined model demonstrated higher AUC values compared to CRS and TBS model in training, internal and external validation cohorts (Delong-test P < 0.05). RQQM outperformed the radiomics model, but was slightly inferior to the combined model. Survival curves revealed statistically significant differences in 1-year DFS and 3-year OS for the RQQM (P < 0.001). RQQM integrates both “quality” (radiomics) and “quantity” (TBS). The radiomics model is superior to the TBS model and has a greater impact on patient prognosis. In the absence of clinical data, RQQM, relying solely on imaging data, shows an advantage in predicting early recurrence after radical surgery for CRLM.
OBJECTIVE:Eosinophilic solid and cystic renal cell carcinoma (ESC-RCC) is rare and difficult to diagnose. Therefore, we aim to investigate the imaging and pathologic features of ESC-RCC. METHODS:Fifteen cases of ESC-RCC with pathologically confirmed diagnoses were retrospectively collected: CT was performed in 15 cases and MRI in 9 cases. RESULTS:In these patients (6 males and 9 females) (age: mean, 53.3 ± 14.7 years; range, 27-72 years), all tumors were unilateral, renal, and solitary with no clinical symptoms and were classified into-type 1: cystic-solid component, with equal cystic and solid components, was the most common (8/15, 53.3%); type 2: predominantly cystic with a small solid component (4/15, 26.7%); and type 3: predominantly solid (3/15, 20%). The solid component showed equal/slightly higher density on the CT-plain-scan, equal/slightly high signal on the T1-weighted image (T1WI), and low signal on the T2-weighted image (T2WI). Ten cases showed progressive enhancement, while 5 showed a fast-wash-in and fast-wash-out enhancement. One patient experienced hemorrhage, while the others showed no signs of hemorrhage, necrosis, fat, or calcification. Pathologically, the tumor showed cystic solidity, with eosinophilic cytoplasm and granular basophilic-colored spots with focal or diffuse expression of CK20. Ten patients had componential nephrectomy and 5 had radical nephrectomy. No recurrence or metastasis was noted in any case at the follow-up (8-49 months). CONCLUSION:This study describes the imaging and pathologic features of a rare type of renal cancer and proposes 3 imaging types to enhance physicians' diagnosis of this disease and guide clinical diagnosis and treatment.
Ningbo Medical Centre Lihuili Hospital, China; Xiangshan County First People's Hopsital Medical Health Group, China; Ningbo Yinzhou Anorectal Hospital Co.,Ltd, China; Ningbo Municipal Hospital of TCM, China; Yuyao Hospital of T.C.M, China; Ninghai County Hospital of Traditional Chinese Medicine, China.
BACKGROUND:Nonalcoholic fatty liver disease (NAFLD) is a significant etiological factor in liver-related diseases, which can lead to severe consequences such as steatohepatitis, cirrhosis and death. Cdh1 is considered as a crucial protein involved in cell cycle regulation. The purpose of this study is to explore the biological role of Cdh1 in NAFLD. MATERIALS AND METHODS:NAFLD cell model was established, and L02 cells and AML12 cells were infected by shRNA lentivirus with Cdh1 knockdown in vitro, and the effect of Cdh1 deletion on cell lipid deposition was evaluated. The effects of Cdh1 deletion on Akt phosphorylation and PPAR/PGC-1α signaling pathway in L02 cells were examined. In addition, the NAFLD mouse model was constructed, and the conditional knockout mice of Cdh1 were selected to verify the results. RESULTS:In vitro experiments showed that the Cdh1 deletion enhanced cell lipid deposition. In vivo experiments showed that conditional knockdown of Cdh1 aggravated fatty degeneration and damage of liver in mice. Cdh1 deletion promotes Akt phosphorylation and inhibits PPAR/PGC-1α signaling pathway in L02 cells. Conditional knockout of Cdh1 down-regulates PPAR/PGC-1α signaling pathway in NAFLD mouse model. CONCLUSION:The deletion of Cdh1 may promote Akt phosphorylation by up-regulating Skp2 and inhibit the PPAR/PGC-1α signaling pathway. Cdh1 serves a protective function in the occurrence and progression of NAFLD.
Background: To evaluate the value of the spectral CT parameters in predicting the risk of esophageal variceal bleeding in cirrhosis with portal hypertension and to provide a reference for clinical diagnosis and treatment. Methods: Seventy-eight patients were divided into an esophageal variceal bleeding group and a non- esophageal variceal bleeding group. A comparison of variables including age, gender, platelet count, Child-Pugh classification, and spectral parameters between the 2 groups was done. Baseline model and spectral model were constructed with conventional parameters and conventional parameters coupled with spectral parameters, respectively. The 2 models were analyzed by the Receiver Operating Characteristic (ROC) curve. Results: The baseline model was established based on 4 conventional parameters and evaluated by ROC curve analysis. The spectral model was constructed based on the variables in the baseline combined with normalized iodine density in the liver parenchyma for the arterial phase, normalized iodine density in the liver parenchyma for the portal phase, normalized iodine density in the splenic parenchyma for the portal phase, diameter of the main portal vein, diameter of the splenic vein, and normalized iodine density of the left gastric vein. Normalized iodine density of the left gastric vein, normalized iodine density in the liver parenchyma for the portal phase, and Child-Pugh classification were the influencing factors of esophageal variceal bleeding in cirrhosis patients. The Area Under Curve (AUC) for the baseline and spectral models were compared (0.664 vs. 0.860) and the difference was found to be statistically significant (P <.001). Conclusions: The use of spectral CT parameters in consort with the conventional parameters can improve the diagnostic effectiveness of esophageal variceal bleeding in cirrhosis cases and screen for high-risk esophageal variceal bleeding patients. It may also provide an objective basis for the clinical prevention and treatment of esophageal variceal bleeding.
Background Gastric varices (GV) with spontaneous portosystemic shunts (SPSS) pose considerable risks and challenges for administering endoscopic cyanoacrylate (CYA) injection. This study aimed to evaluate the efficacy and safety of EUS-guided coil embolization in combination with CYA injection compared to conventional endoscopic CYA injection for managing GV with SPSS. Methods This retrospective analysis included patients with SPSS treated with either EUS-guided coil embolization in combination with CYA injection or conventional CYA injection for gastric variceal bleeding at Ningbo Medical Center Lihuili Hospital (Zhejiang, China) between January 2018 and March 2023. Patient demographics, procedural details, and follow-up results were reviewed. Results The study evaluated 57 patients: 21 in the combined treatment group undergoing EUS-guided coil embolization in combination with CYA injection and 36 in the conventional group receiving conventional endoscopic CYA injection. Both cohorts achieved a 100% technical success rate. The mean volume of CYA used was significantly lower in the combined group (1.64 ± 0.67 mL) than in the conventional group (2.38 ± 0.72 mL; P < 0.001). Early GV rebleeding rates did not differ significantly between the groups; in contrast, the combined treatment group exhibited a considerably lower incidence of late GV rebleeding than the conventional group (4.8% vs 27.8%, P = 0.041). Conclusions EUS-guided coil embolization in combination with CYA injection demonstrated superiority over conventional endoscopic CYA injection in reducing late GV rebleeding in treating GV with SPSS.
OBJECTIVE:Liver cancer (LC) is a frequently occurring lethal malignancy worldwide, yet the molecular mechanisms of carcinogenesis and their development remain uncharacterized. In this study, bioinformatics methods were used to find candidate hub genes for prognosis assessment and clinical treatment of LC.METHODS:Differential analysis was carried out based on the evidence of gene expression profiling in LC on The Cancer Genome Atlas (TCGA). The differentially expressed genes (DEGs) were constructed into co-expression networks and divided into modules by virtue of weighted gene co-expression network analysis (WGCNA). Based on the Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG), the module genes were subjected to functional enrichment analysis. The LC microarray (GSE105130) in the Gene Expression Omnibus was selected to verify the hub genes' expression profiles. The validity of the hub genes was verified via survival analysis, as well as expression correlation with the clinicopathological features. Thereafter, gene set variation analysis (GSVA) and single-sample gene set enrichment analysis (GSEA) were applied to investigate the possible biological functions of the hub genes.RESULTS:In total, 3780 DEGs and 17 co-expression modules were obtained. The blue module had the strongest correlation with the tumour stage and the module genes were principally enriched in tumour-associated GO terms, as well as pathways such as Ras protein signal transduction, ERK1/2 cascade, Ras signal pathway, and ECM-receptor interaction. RASAL1, which is highly expressed in LC, was identified as a hub gene for LC progression. Its high expression suggested unfavorable patient prognosis and was correlated with T stage, gender and tumour stage. Further analysis identified that the overexpression of RASAL1 was substantially enriched in cancer-associated gene sets.CONCLUSION:RASAL1 is a hub gene that influences LC progression, constituting a novel biomarker and molecular target in the future diagnosis and therapy of LC.
目的 比较预切开内镜下黏膜切除术(Pre-cut-EMR)与内镜黏膜下剥离术(ESD)治疗十二指肠非壶腹部病变的临床疗效及安全性.方法 回顾性分析2015年6月-2021年6月宁波市医疗中心李惠利医院64例确诊为十二指肠非壶腹部病变的患者的临床资料.其中,30例行Pre-cut-EMR,34例行ESD.比较两组患者病灶特征、手术时间、并发症发生率、整块切除率、R0切除率和术后组织病理学结果.结果 两组患者病灶宏观形态、迟发性出血、整块切除率、R0切除率和术后病理等比较,差异均无统计学意义(P>0.05).所有患者均成功完成内镜下切除,整块切除率为100.0%,总体R0切除率为96.9%,术后均未发生迟发性穿孔.Pre-cut-EMR组手术时间为(24.6±6.5)min,明显短于ESD组的(37.5±9.5)min,术中穿孔率为3.3%(1/30),明显低于ESD组的23.5%(8/34),差异均有统计学意义(P<0.05).Pre-cut-EMR组病灶直径为1.40(1.30,1.70)cm,小于ESD组的1.45(1.30,2.00)cm,但差异无统计学意义(P>0.05);Pre-cut-EMR组病灶直径≥2.0 cm的比例为10.0%(3/30),低于ESD组的32.4%(11/34),差异有统计学意义(P<0.05).结论 Pre-cut-EMR和ESD治疗十二指肠非壶腹部病变的疗效相当,但Pre-cut-EMR手术时间更短,术中穿孔率更低,治疗直径<2.0 cm的病灶是安全有效的.
目的 探讨能谱CT参数在评估肝硬化门脉高压患者的血流动力学变化及食管静脉曲张(EV)严重程度中的价值.方法 选取肝硬化门脉高压患者70例,在能谱碘图上测量肝动脉期和门脉期肝各叶碘浓度,同层面动脉期腹主动脉和门脉期门脉主干、脾实质、胃左静脉碘浓度.以胃镜为"金标准"将EV分为轻、中和重度,比较3组EV严重程度与每个能谱参数的差异及相关性.结果 3组肝动脉期标准化碘浓度(NIDLAP)、胃左静脉标准化碘浓度(LGVI)差异均有统计学意义(均P<0.05),且与EV程度呈正相关(r=0.729、0.822,均P<0.05).3组肝门脉期标准化碘浓度(NIDLVP)、脾实质门脉期标准化碘浓度(NIDSVP)差异均有统计学意义(均P<0.05),轻度、中度组间差异无统计学意义(P>0.05).NIDLVP与EV程度呈负相关(r=—0.477,P<0.05),NIDSVP与EV程度呈正相关(r=0.378,P<0.05).结论 能谱CT参数可客观反映肝脾血流动力学改变,评估肝硬化门脉高压患者EV的严重程度,其中LGVI与EV严重程度相关性最强,可作为EV严重程度的非侵入性随访观察指标.
Studies over the past decades have implicated lncRNAs in promoting the development, migration and invasion of gastric cancer (GC). However, the role and mechanism of lncRNA MBNL1-AS1 in GC promotion are poorly understood. In this research, qRT-PCR showed that MBNL1-AS1 was down-regulated in GC tissues and cells. Cell experiments and the animal study demonstrated that MBNL1-AS1 knockdown accelerated GC cell proliferation, migration, and invasion, thus restraining cell apoptosis. Meanwhile, overexpression of MBNL1-AS1 repressed GC cell promotion. Bioinformatics analysis confirmed that MBNL1-AS1 binds to miR-424-5p via negative modulation. Rescue experiments showed that decreased miR-424-5p level inhibited GC cell promotion by silencing MBNL1-AS1. Furthermore, Smad7 was identified as a target of miR-424-5p that could reverse the promotion of GC cell growth mediated by miR-424-5p. Western blot results proved that MBNL1-AS1 affected TGF-β/SMAD pathways by regulating the miR-424-5p/Smad7 axis. Collectively, MBNL1-AS1 restrained GC growth via the miR-424-5p/Smad7 axis and thus could be a promising target for GC therapy. These findings illustrate that lncRNA MBNL1-AS1, as a tumor suppressor gene, participates in GC progression by regulating miR-424-5p/Smad7 axis, thus activating TGF-β/EMT pathways. The evidence may provide a potential marker for GC patients.
目的 探讨内镜下黏膜剥离术(ESD)在非乳头十二指肠黏膜病变中的应用效果.方法 收集行ESD的十二指肠黏膜病变患者32例.评价ESD剥离效果,记录操作时间、术中及术后并发症及处理情况,分析剥离标本的病理学结果.结果 本组32例患者均一次性完整切除病变,剥离标本行全瘤病理检查,其中低级别上皮内瘤变6例,高级别上皮内瘤变8例,黏膜内腺癌5例,绒毛管状及管状腺瘤9例,淋巴瘤2例,Brunner腺瘤2例.手术操作时间45~112min,平均(60.7±25.4)min,切除标本长径1.0~3.5 cm.ESD术中发生小穿孔3例,均经内镜下软组织夹成功闭合,未发生气腹、腹膜炎、胆瘘、胰瘘等并发症.1例术中穿孔创口较大,内镜缝合未成功行外科腹腔镜修补,2例患者术后第3、4d出现大出血,经再次内镜止血夹治疗后出血停止.患者术后5~10d出院,术后3、6及12月内镜检查观察创面情况,随访3~30个月,3例患者因各种原因失访,其余患者随访病变无残留及复发.结论 ESD是一种治疗非乳头十二指肠黏膜病变安全、有效、简便的治疗方法,大部分病变都能完整剥离,由于十二指肠解剖结构的特殊性,更要警惕出血和穿孔等并发症的发生,必要时仍需要外科手术治疗.
本文报道1例50岁女性左肾嗜酸性实性和囊性肾细胞癌的影像表现。病灶呈类圆形囊实性,边界清晰,CT平扫实性成分呈等密度,增强肾皮质期及实质期持续渐进性强化,排泄期强化减退,强化程度均稍低于同期肾皮质,囊性部分无强化。MRI平扫实性部分T 1WI及T 2WI均呈稍低信号,DWI为高信号,增强表现类似CT强化。
目的 探讨血清钙离子(Ca2+)水平与炎症性肠病(inflammatory bowel disease,IBD)疾病活动性的关系.方法 选取2007年1月至2018年12月宁波市医疗中心李惠利医院消化内科收治的IBD患者共120例,其中克罗恩病(CD)患者60例和溃疡性结肠炎(UC)患者60例,并选取同期在本院体检中心行常规体检的健康人群60例作为正常对照组,采用克罗恩病活动指数(CDAI)评分和简单临床结肠炎活动指数(SCCAI)评分分别用于评估CD和UC患者的疾病活动性,Spearman相关分析血清Ca2+水平与SCCAI、CDAI评分、PLT及CRP的相关性.结果 CD及UC组血清Ca2+及PLT水平比较差异均无统计学意义(均P>0.05),而在CD与正常组及UC与正常组之间的差异均有统计学意义(均P<0.05).UC患者血清Ca2+水平与SCCAI评分呈负相关(rs=-0.27,P<0.05),CD患者血清Ca2+水平与CDAI评分不相关(rs=-0.21,P>0.05).进一步分析CD患者的血清Ca2+水平与PLT及CRP的相关性,两者均呈负相关(rs=-0.32、-0.31,均P<0.05).结论 血清Ca2+水平在IBD患者的疾病诊断和疾病活动性评估中具有重要的临床意义.
目的 比较放大内镜(ME)、放大内镜联合窄带成像技术(ME-NBI)、放大内镜联合靛胭脂染色(ME-IDC)及放大内镜联合靛胭脂染色与窄带成像技术(ME-IDC-NBI)对早期胃癌的诊断价值.方法 因癌前病变或早期胃癌行内镜下黏膜剥离术治疗患者185例,先后行ME、ME-NBI、ME-IDC及ME-IDC-NBI检查,以术后病理结果为“金标准”,统计分析各项检查与病理诊断的一致性,并比较其对早期胃癌诊断的准确性.结果 185例患者术后病理结果显示,癌性病变107例,非癌性病变78例.ME、ME-NBI及ME-IDC与病理诊断一致性一般(k=0.62、0.74、0.72,均P<0.01),ME-IDC-NBI与病理诊断一致性好(k=0.86,P< 0.01).MEI-NBI、ME-IDC及ME-IDC-NBI诊断早期胃癌准确率均高于ME(均P<0.05);ME-IDC-NBI准确率均高于ME-NBI与ME-IDC(均P< 0.05).结论 ME联合NBI或IDC均能提高对早期胃癌的诊断准确性,与病理结果有较好的一致性,而ME与NBI及IDC相结合后,对早期胃癌的诊断则更具优势.
目的 比较人工智能与放射科医师判断孤立性肺结节(SPN)良恶性的效果.方法 回顾性分析2018年9月至2019年5月在宁波市医疗中心李惠利医院行胸部高分辨率CT扫描的原发性SPN患者150例,分别进行人工智能肺结节检测系统检测和副主任及以上放射科医师3人判读,以术后病理诊断为金标准,比较二者对SPN良恶性结果判断的准确率.结果 人工智能的敏感度(92.6%)、特异度(82.1%)、阳性预测值(95.8%)、准确率(90.7%)略低于放射科医师的判断(96.7%、89.3%、97.5%、95.3%).结论 人工智能具有较高的敏感度、特异度及准确率,在一定程度上可以帮助放射科医师辅助诊断SPN良恶性.