Ungureanu Bogdan Silviu, Gheonea Dan Ionut, Florescu Dan Nicolae, Iordache Sevastita, Cazacu Sergiu Marian, Iovanescu Vlad Florin, Rogoveanu Ion, Turcu-Stiolica Adina
Department of Gastroenterology, University of Medicine and Pharmacy of Craiova, Craiova, Romania.
Department of Pharmacoeconomics, University of Medicine and Pharmacy of Craiova, Craiova, Romania.
Front Med (Lausanne). 2023 Feb 17;10:1134835. doi: 10.3389/fmed.2023.1134835. eCollection 2023.
Non-endoscopic risk scores, Glasgow Blatchford (GBS) and admission Rockall (Rock), are limited by poor specificity. The aim of this study was to develop an Artificial Neural Network (ANN) for the non-endoscopic triage of nonvariceal upper gastrointestinal bleeding (NVUGIB), with mortality as a primary outcome.
Four machine learning algorithms, namely, Linear Discriminant Analysis (LDA), Quadratic Discriminant Analysis (QDA), logistic regression (LR), K-Nearest Neighbor (K-NN), were performed with GBS, Rock, Beylor Bleeding score (BBS), AIM65, and T-score.
A total of 1,096 NVUGIB hospitalized in the Gastroenterology Department of the County Clinical Emergency Hospital of Craiova, Romania, randomly divided into training and testing groups, were included retrospectively in our study. The machine learning models were more accurate at identifying patients who met the endpoint of mortality than any of the existing risk scores. AIM65 was the most important score in the detection of whether a NVUGIB would die or not, whereas BBS had no influence on this. Also, the greater AIM65 and GBS, and the lower Rock and T-score, the higher mortality will be.
The best accuracy was obtained by the hyperparameter-tuned K-NN classifier (98%), giving the highest precision and recall on the training and testing datasets among all developed models, showing that machine learning can accurately predict mortality in patients with NVUGIB.
非内镜风险评分，如格拉斯哥布拉奇福德评分（GBS）和入院时罗卡尔评分（Rock），存在特异性较差的局限性。本研究的目的是开发一种用于非静脉曲张性上消化道出血（NVUGIB）非内镜分诊的人工神经网络（ANN），将死亡率作为主要结局。
使用GBS、Rock、贝洛出血评分（BBS）、AIM65和T评分，对四种机器学习算法，即线性判别分析（LDA）、二次判别分析（QDA）、逻辑回归（LR）、K近邻算法（K-NN）进行了分析。
本研究回顾性纳入了罗马尼亚克拉约瓦县临床急诊医院胃肠病科收治的1096例NVUGIB患者，这些患者被随机分为训练组和测试组。机器学习模型在识别达到死亡终点的患者方面比任何现有的风险评分都更准确。AIM65是检测NVUGIB患者是否会死亡的最重要评分，而BBS对此没有影响。此外，AIM65和GBS越高，Rock和T评分越低，死亡率就越高。
通过超参数调整的K-NN分类器获得了最佳准确率（98%），在所有开发的模型中，其在训练和测试数据集上的精度和召回率最高，表明机器学习可以准确预测NVUGIB患者的死亡率。
