Scholarlink

Background <p>When three-dimensional computer graphics (3DCG) images are used, artificial intelligence (AI) engines can semiautomatically estimate the pancreatic thickness, pancreatic stump shape (PSS), surface area of the resected pancreatic stump, and remnant pancreatic volume (RPV). Here, we report the potential applications of the PSS and RPV and investigate postoperative pancreatic fistula (POPF) risk factors.</p> Patients and Methods <p>We included 109 patients with pancreatic cancer in this study. First, we employed the roundness value to objectively evaluate the PSS. The correlations between the roundness value and the pancreatic thickness and surface area of the resected pancreatic stump were examined. In terms of RPV evaluation, we employed the RPV ratio, which was automatically calculated by the software by dividing the volume of the remnant pancreas by the total volume of the pancreas. Second, we performed a multivariable analysis to identify POPF risk factors, including the roundness value and RPV ratio.</p> Results <p>The roundness value was significantly correlated with the pancreatic thickness and surface area of the resected pancreatic stump. Multivariable logistic regression analysis also revealed that a roundness value ≤&#xa0;11.3, an RPV ratio &gt;&#xa0;0.65, the absence of pathological fibrosis, and a pancreatic duct diameter ≤&#xa0;2.0&#xa0;mm were independent risk factors for POPF.</p> Conclusions <p>The preoperative roundness value and RPV ratio measured from 3DCG images were able to predict POPF occurrence.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of Postoperative Pancreatic Fistula Occurrence by the Pancreatic Stump Shape and Remnant Pancreatic Volume Measured Using AI-Guided Three-Dimensional Simulation

  • 期刊论文

  • 发表日期：

    2026年2月21日

查看全文
  • Ryoichi Miyamoto,
  • Masahiro Shiihara,
  • Mitsuru Watanabe,
  • Jiro Shimazaki,
  • Mitsugi Shimoda,
  • Shuji Suzuki

摘要

Background

When three-dimensional computer graphics (3DCG) images are used, artificial intelligence (AI) engines can semiautomatically estimate the pancreatic thickness, pancreatic stump shape (PSS), surface area of the resected pancreatic stump, and remnant pancreatic volume (RPV). Here, we report the potential applications of the PSS and RPV and investigate postoperative pancreatic fistula (POPF) risk factors.

Patients and Methods

We included 109 patients with pancreatic cancer in this study. First, we employed the roundness value to objectively evaluate the PSS. The correlations between the roundness value and the pancreatic thickness and surface area of the resected pancreatic stump were examined. In terms of RPV evaluation, we employed the RPV ratio, which was automatically calculated by the software by dividing the volume of the remnant pancreas by the total volume of the pancreas. Second, we performed a multivariable analysis to identify POPF risk factors, including the roundness value and RPV ratio.

Results

The roundness value was significantly correlated with the pancreatic thickness and surface area of the resected pancreatic stump. Multivariable logistic regression analysis also revealed that a roundness value ≤ 11.3, an RPV ratio > 0.65, the absence of pathological fibrosis, and a pancreatic duct diameter ≤ 2.0 mm were independent risk factors for POPF.

Conclusions

The preoperative roundness value and RPV ratio measured from 3DCG images were able to predict POPF occurrence.