内科

目的探讨普通白光内镜（C-WLE）、窄带成像放大内镜（ME-NBI）、小探头超声内镜（EUS）单独及联合应用对胃癌的定性诊断价值及对胃癌浸润深度的评估准确性。方法前瞻性选取2022年5月至2024年5月因胃部症状或胃癌高危因素就诊于信阳市人民医院的116例患者，所有患者均行C-WLE、ME-NBI、小探头EUS检查。以术后组织病理学检查结果为金标准，评估C-WLE、ME-NBI、小探头EUS单独及联合检查与组织病理学检查结果的一致性，并比较不同检查方法的定性诊断效能；分析联合检查对浸润深度的评估准确性。结果 116例患者中病理确诊胃癌20例（17.24%）；C-WLE、ME-NBI、小探头EUS单独及联合检查与术后组织病理学检查的一致性均较好（Kappa=0.635～0.723，均P<0.05），其中联合检查Kappa值最低（0.635）；联合检查定性诊断灵敏度最高（85.00%），优于三种单一检查（均P<0.05），特异度、准确度组间差异无统计学意义（均P>0.05）。术后组织病理学检查共发现127处病灶，联合检查评估浸润深度的总体准确率为86.61%（110/127），允许相邻等级误差的加权准确率为96.7%，加权Kappa=0.629；误判以高估为主（16例），仅1例低估。结论三种内镜单独及联合应用对胃癌定性诊断均有较高价值，联合检查虽未提高诊断一致性，但灵敏度最优，且在浸润深度评估方面优势突出，尤其对深层病变低估率低，可为治疗决策提供可靠依据。

Objective To investigate the value of application of conventional white light endoscopy (C-WLE), magnifying endoscopy with narrow band imaging (ME-NBI), and mini-probe endoscopic ultrasound (EUS), alone or in combination, in qualitative diagnosis of gastric cancer and the accuracy in invasion depth assessment of gastric cancer. Methods A total of 116 patients with gastric symptoms or high-risk factors for gastric cancer admitted to Xinyang People's Hospital from May 2022 to May 2024 were prospectively enrolled. All patients underwent C-WLE, ME-NBI, and mini-probe EUS examinations. With postoperative histopathological results as the gold standard, the consistency of C-WLE, ME-NBI, and mini-probe EUS, used alone or in combination, with histopathological examination results was evaluated, the qualitative diagnostic performance of different examination modalities was compared, and the accuracy of combined examination in assessing tumor invasion depth was analyzed. Results Among the 116 patients, 20 cases (17.24%) were pathologically confirmed as gastric cancer. C-WLE, ME-NBI, and mini-probe EUS, used alone or in combination, showed good consistency with postoperative histopathology (Kappa=0.635-0.723, all P<0.05), with the combined examination yielding the lowest Kappa value (0.635); the combined examination achieved the highest qualitative diagnostic sensitivity (85.00%), which was superior to that of any single examination (all P<0.05), and no statistically significant intergroup difference was observed in specificity and accuracy (all P>0.05). A total of 127 lesions were identified by postoperative histopathology. The overall accuracy of the combined examination for invasion depth assessment was 86.61% (110/127). The weighted accuracy allowing adjacent-grade error reached 96.7%, with a weighted Kappa of 0.629. Misjudgments were predominantly overestimation (16 cases), and only 1 case was underestimated. Conclusion Individual and combined use of the 3 endoscopic modalities possess high value for qualitative diagnosis of gastric cancer. Although the combined examination does not improve diagnostic consistency, it achieves optimal sensitivity and shows prominent advantages in invasion depth assessment, especially with a low underestimation rate for deep-seated lesions, which can provide reliable evidence for therapeutic decision-making.