Department of General Surgery, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai 200025, China.
Shanghai Minimally Invasive Surgery Center, Shanghai 200025, China.
World J Gastroenterol. 2023 Jan 21;29(3):536-548. doi: 10.3748/wjg.v29.i3.536.
Multiple linear stapler firings during double stapling technique (DST) after laparoscopic low anterior resection (LAR) are associated with an increased risk of anastomotic leakage (AL). However, it is difficult to predict preoperatively the need for multiple linear stapler cartridges during DST anastomosis.
To develop a deep learning model to predict multiple firings during DST anastomosis based on pelvic magnetic resonance imaging (MRI).
We collected 9476 MR images from 328 mid-low rectal cancer patients undergoing LAR with DST anastomosis, which were randomly divided into a training set ( = 260) and testing set ( = 68). Binary logistic regression was adopted to create a clinical model using six factors. The sequence of fast spin-echo T2-weighted MRI of the entire pelvis was segmented and analyzed. Pure-image and clinical-image integrated deep learning models were constructed using the mask region-based convolutional neural network segmentation tool and three-dimensional convolutional networks. Sensitivity, specificity, accuracy, positive predictive value (PPV), and area under the receiver operating characteristic curve (AUC) was calculated for each model.
The prevalence of ≥ 3 linear stapler cartridges was 17.7% (58/328). The prevalence of AL was statistically significantly higher in patients with ≥ 3 cartridges compared to those with ≤ 2 cartridges (25.0% 11.8%, = 0.018). Preoperative carcinoembryonic antigen level > 5 ng/mL (OR = 2.11, 95%CI 1.08-4.12, = 0.028) and tumor size ≥ 5 cm (OR = 3.57, 95%CI 1.61-7.89, = 0.002) were recognized as independent risk factors for use of ≥ 3 linear stapler cartridges. Diagnostic performance was better with the integrated model (accuracy = 94.1%, PPV = 87.5%, and AUC = 0.88) compared with the clinical model (accuracy = 86.7%, PPV = 38.9%, and AUC = 0.72) and the image model (accuracy = 91.2%, PPV = 83.3%, and AUC = 0.81).
MRI-based deep learning model can predict the use of ≥ 3 linear stapler cartridges during DST anastomosis in laparoscopic LAR surgery. This model might help determine the best anastomosis strategy by avoiding DST when there is a high probability of the need for ≥ 3 linear stapler cartridges.
腹腔镜低位前切除术（LAR）后双吻合技术（DST）中多次线性吻合器击发与吻合口漏（AL）风险增加相关。然而，术前很难预测 DST 吻合过程中需要多次线性吻合器管。
开发一种基于盆腔磁共振成像（MRI）预测 DST 吻合中多次击发的深度学习模型。
我们收集了 328 例接受 LAR 合并 DST 吻合的中低位直肠癌患者的 9476 例 MRI 图像，随机分为训练集（n=260）和测试集（n=68）。采用二元逻辑回归分析，利用 6 个因素创建临床模型。对整个骨盆快速自旋回波 T2 加权 MRI 序列进行分割和分析。使用掩模区域卷积神经网络分割工具和三维卷积网络构建纯图像和临床图像集成深度学习模型。计算每个模型的敏感性、特异性、准确性、阳性预测值（PPV）和受试者工作特征曲线下面积（AUC）。
≥3 个线性吻合器管的发生率为 17.7%（58/328）。≥3 个吻合器管的患者 AL 发生率明显高于≤2 个吻合器管的患者（25.0% vs. 11.8%，P=0.018）。术前癌胚抗原水平>5ng/ml（OR=2.11，95%CI 1.08-4.12，P=0.028）和肿瘤大小≥5cm（OR=3.57，95%CI 1.61-7.89，P=0.002）被认为是使用≥3 个线性吻合器管的独立危险因素。与临床模型（准确性=86.7%，PPV=38.9%，AUC=0.72）和图像模型（准确性=91.2%，PPV=83.3%，AUC=0.81）相比，基于 MRI 的深度学习模型在预测 DST 吻合术中使用≥3 个线性吻合器管的诊断性能更好（准确性=94.1%，PPV=87.5%，AUC=0.88）。
基于 MRI 的深度学习模型可以预测腹腔镜 LAR 手术中 DST 吻合术中使用≥3 个线性吻合器管的情况。该模型通过避免 DST 吻合，有助于在高概率需要≥3 个线性吻合器管时确定最佳吻合策略。
