| 何文.BP神经网络、随机森林预测结直肠癌根治术后并发症的效能对比[J].安徽医药,待发表. |
| BP神经网络、随机森林预测结直肠癌根治术后并发症的效能对比 |
|
| 投稿时间：2024-04-17 录用日期：2024-05-11 |
| DOI： |
| 中文关键词: 结直肠癌 根治术 术后并发症 BP神经网络 随机森林 |
| 英文关键词: |
| 基金项目: |
|
| 摘要点击次数: 461 |
| 全文下载次数: 0 |
| 中文摘要: |
| 目的 BP神经网络、随机森林预测结直肠癌根治术后并发症的效能对比分析。方法 选取2023年1月-2024年1月本院收治的128例行根治术治疗的结直肠癌患者为研究对象。根据受试者术后是否发生并发症分为并发症组（90例）和无并发症组（38例）。收集患者一般资料、病理学特征、围术期指标。采用t检验或c2检验筛选出可能影响结直肠癌根治术患者术后发生并发症的因素。将受试者按照7:3的比例分为训练集和测试集，训练集用于构建BP神经网络和随机森林模型，测试集用于评估上述两种模型的预测效能。结果 128例患者中，38例患者术后发生并发症，并发症发生率为29.69%（38/128）。并发症组年龄、手术时间水平高于无并发症组（P<0.05），贫血、aCCI评分＞4分、PNI＜42.15占比高于无并发症组（P<0.05）。ROC曲线分析结果显示，BP神经网络和随机森林模型预测测试集结直肠癌根治术患者术后并发症的AUC分别为0.816（95CI%：0.765~0.863）、0.921（95CI%：0.893~0.982），灵敏度分别为85.70%、94.50%，特异度分别为80.30%、90.50%，正确率分别为79.00%、88.40%，召回率分别为76.40%、90.20%，精确率分别为79.70%、90.20%。Delong检验结果显示，随机森林预测测试集结直肠癌根治术患者术后并发症的AUC大于BP神经网络，差异有统计学意义（D=-3.021，P=0.004）。结论 BP神经网络、随机森林模型对结直肠癌根治术患者术后并发症的预测效能较好，其中随机森林模型对该病患者术后发生并发症的预测效能更优，值得在临床推广应用。 |
| 英文摘要: |
| Objective Comparative analysis of efficacy of BP neural network and random forest in predicting complications after radical resection of colorectal cancer. Methods From January 2023 to January 2024, 128 patients with colorectal cancer who underwent radical surgery in our hospital were selected as the study objects. The subjects were divided into complication group (90 cases) and no complication group (38 cases) according to whether complications occurred after surgery. General data, pathological features and perioperative indexes were collected. t test or c2 test were used to screen out the factors that may affect the postoperative complications of patients with radical resection of colorectal cancer. The subjects were divided into a training set and a test set according to a ratio of 7:3. The training set was used to build the BP neural network and the random forest model, and the test set was used to evaluate the prediction efficiency of the two models. Results Among the 128 patients, 38 had postoperative complications, with a complication rate of 29.69% (38/128). The level of age and operation time in the complication group was higher than that in the uncomplication group (P<0.05), the ratio of anemia, aCCI score > 4 and PNI < 42.15 was higher than that in the uncomplication group (P<0.05). ROC curve analysis results showed that the AUC of postoperative complications in patients with radical resection of rectal cancer grouped by BP neural network and random forest model prediction test were 0.816 (95CI% : 0.765~0.863) and 0.921 (95CI% : 0.921), respectively. 0.893~0.982), sensitivity 85.70%, 94.50%, specificity 80.30%, 90.50%, accuracy 79.00%, 88.40%, recall rate 76.40%, 90.20%, accuracy 79.70%, 90.20%, respectively. The results of Delong test showed that the AUC of postoperative complications of patients undergoing radical resection of rectal cancer collected by random forest prediction test was greater than that of BP neural network, with statistical significance (D=-3.021, P=0.004). Conclusion BP neural network and random forest model have better predictive efficacy for postoperative complications of patients with radical resection of colorectal cancer, and random forest model has better predictive efficacy for postoperative complications of patients with this disease, which is worthy of clinical application. |
|
查看/发表评论 下载PDF阅读器 |
| 关闭 |
|
|
|
