Objective To construct a machine learning prediction model for poorly differentiated invasive non-mucinous lung adenocarcinoma (INMA) based on the quantitative and qualitative features of low-dose CT (LDCT) and the expression of NRF2 and Ki-67 in tissues, and to verify its efficacy. Methods The data of 196 INMA patients from January 2023 to January 2025 in the 73rd Army Group Hospital of PLA Army were retrospectively collected as the training set. Another 85 INMA patients’ data from February 2025 to July 2025 were selected as the test set at a ratio of 7∶3. The training set was divided into high/moderately differentiated (n=146) and poorly differentiated (n=50) groups according to the differentiation degree of the lesions. Univariate analysis, LASSO regression, and multivariate logistic regression were conducted on the variables, and the independent predictive factors were selected to establish machine learning prediction models of logistic regression (LR), random forest (RF), extreme gradient boosting (XGBoost), and support vector machine (SVM). The predictive value of the models constructed by the four methods was compared using the receiver operating characteristic (ROC) curve. Results The independent predictors selected by univariate, LASSO regression and multivariate Logistic included mean nodule diameter, nodule volume, mean nodule density, NRF2 expression, and Ki-67 index. The areas under the curve (AUC) of the LR, SVM, RF, and XGBoost machine learning models in the training set were 0.958, 0.952, 0.977, and 0.987, respectively. The XGBoost model demonstrated the highest accuracy 94.39%, sensitivity 96.00%, and AUC, outperforming other models. Conclusions The XGBoost model based on the quantitative and qualitative features of low-dose CT and the expression of NRF2 and Ki-67 in tissues can accurately predict the risk of developing poorly differentiated INMA, providing evidence-based basis for early identification and targeted intervention in high-risk patients.
Key words
low-dose CT /
nuclear factor erythroid 2-related factor 2 /
Ki-67 /
machine learning /
predictive model
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
References
[1] 叶元滋, 张思远, 夏万里, 等. ⅠA-ⅢB期肺浸润性非黏液腺癌淋巴结转移临床病理特征及预后研究[J]. 安徽医科大学学报, 2025, 60(5): 834-841.
[2] 马夏林, 贾若楠, 韩 凯,等. 肺腺癌常见驱动基因变异与临床病理分型的相关性[J]. 中华病理学杂志, 2024, 53(6): 578-584.
[3] 阿里甫·依马木, 文 智, 周 永, 等. 高危人群低剂量CT与肿瘤标志物联合检查早期肺癌的效能研究[J]. 中国CT和MRI杂志, 2024, 22(2): 58-60.
[4] 赵立众, 张 晗, 印凌华. 低剂量CT筛查技术对于健康体检中肺癌高危人群的应用研究[J]. 中华保健医学杂志, 2022, 24(4): 263-265.
[5] 孙定周, 张 鑫, 朱洁晨, 等. Nrf2调控PI3K/Akt信号通路促进肺腺癌进展的机制研究[J]. 浙江医学, 2023, 45(22): 2385-2391.
[6] 黎川华, 吴 红, 覃 秋, 等. 肺癌18F-FDG PET/CT代谢参数与病理类型、分化程度及Ki-67指数的相关性[J]. 分子影像学杂志, 2025, 48(6): 726-730.
[7] 张 地, 吴 毅, 徐 瑜, 等. 基于机器学习的多模态数据建立Ⅰ期非小细胞肺癌患者术后复发预测模型[J]. 陆军军医大学学报, 2025, 47(14): 1602-1611.
[8] Kendall M G. Rank correlation methods[M]. London: Charles Griffin, 1948: 108.
[9] 中华医学会肿瘤学分会, 中华医学会杂志社. 中华医学会肿瘤学分会肺癌临床诊疗指南(2021版)[J]. 中华医学杂志, 2021, 101(23): 1725-1757.
[10] Travis W D, Brambilla E, Nicholson A G, et al. The 2015 World Health Organization classification of lung tumors: impact of genetic, clinical and radiologic advances since the 2004 classification[J]. J Thorac Oncol,2015,10(9):1243-1260.
[11] 李 媛, 谢惠康, 武春燕. WHO胸部肿瘤分类(第5版)中肺肿瘤部分解读[J]. 中国癌症杂志, 2021, 31(7): 574-580.
[12] 杨 溪, 刘杰克, 李 勇, 等. 人工、半自动及自动方法测量肺实性结节直径和体积的一致性研究[J]. 中华放射学杂志, 2022, 56(1): 43-49.
[13] Chen M J, Lin P L, Wang L, et al. Cytoplasmic, but not nuclear Nrf2 expression, is associated with inferior survival and relapse rate and response to platinum-based chemotherapy in non-small cell lung cancer[J]. Thorac Cancer,2020, 11(7): 1904-1910.
[14] 韦道福, 徐 坤, 马咏涛. 低剂量螺旋CT(LDCT)联合血清肺癌自身抗体P53在肺癌诊断中的应用价值[J]. 影像技术, 2025, 37(4): 67-71.
[15] Zhu Y, Yan C, Tang W, et al. Correlation between imaging features of pure ground-glass opacities and pathological subtypes of lung minimally invasive adenocarcinoma and precursor lesions[J]. Sci Rep,2025, 15(1): 7572.
[16] 李 琰, 郭洋洋, 周 青, 等. 低剂量CT联合血清CKAP4、ESM-1对肺结节良恶性的鉴别价值[J]. 临床肺科杂志, 2025, 30(6): 826-831.
[17] 江小红, 库 雄, 魏从兵. 基于Nrf2/GPX4介导的铁死亡探讨异荭草素对非小细胞肺癌细胞增殖和凋亡的影响[J]. 湖南中医药大学学报, 2025, 45(3): 445-452.
[18] 余凯音, 张馨月, 陈 宇, 等. KEAP1/NRF2在各组织学亚型肺癌中的变异特征比较及预后意义分析[J]. 循证医学, 2025, 25(3): 174-180.
[19] 陈 玺, 曾晓颖, 陈佳艳, 等. 错配修复蛋白联合血清肿瘤标志物与Ki-67增殖指数对结直肠癌预后的临床价值[J]. 中国癌症杂志, 2022, 32(3): 243-250.
[20] Li Z, Li F, Pan C, et al. Tumor cell proliferation (Ki-67) expression and its prognostic significance in histological subtypes of lung adenocarcinoma[J]. Lung Cancer, 2021, 154: 69-75.
[21] 周 怡, 张 贝, 黄科峰, 等. 高分辨CT联合Ki-67表达预测低分化浸润性非黏液肺腺癌的临床应用[J]. 武警医学, 2024, 35(9): 777-780.
[22] 张高山, 蒋青青, 陈 敏, 等. Nrf2与p53/p21信号通路在肿瘤中作用的研究进展[J]. 实用药物与临床, 2023, 26(6): 549-555.
[23] 张 洋, 赵化荣, 刘 攀, 等. 多因素Logistic回归模型和XGBoost模型对舌癌患者放疗期间发生口腔感染的预测价值[J]. 中国现代医学杂志, 2023, 33(19): 66-73.
[24] 陈 凯, 吴天晨, 宋德胤, 等. 基于XGBoost算法探究混合磨玻璃结节型肺腺癌浸润程度的关键影像学特征[J]. 实用心脑肺血管病杂志, 2025, 33(1): 23-28.
[25] 王 璐, 刘杰克, 徐富阳, 等. 基于低剂量CT定量-定性特征预测低分化浸润性非黏液肺腺癌[J]. 临床放射学杂志, 2023, 42(12): 1887-1894.
[26] 庄 宇,唐 翎,姚本波,等. 基于CT影像组学联合机器学习预测Ⅰ期肺腺癌气腔播散[J]. 实用放射学杂志,2024,40(9):1434-1438.
PDF(1649 KB)
