A PET/CT-based 3D deep learning model for predicting spread through air spaces in stage I lung adenocarcinoma.

IF 2.8 3区医学 Q2 ONCOLOGY

Clinical & Translational Oncology Pub Date : 2025-02-24 DOI:10.1007/s12094-025-03870-9

Cheng Zheng, Yujie Cai, Jiangfeng Miao, BingShu Zheng, Yan Gao, Chen Shen, ShanLei Bao, ZhongHua Tan, ChunFeng Sun

{"title":"A PET/CT-based 3D deep learning model for predicting spread through air spaces in stage I lung adenocarcinoma.","authors":"Cheng Zheng, Yujie Cai, Jiangfeng Miao, BingShu Zheng, Yan Gao, Chen Shen, ShanLei Bao, ZhongHua Tan, ChunFeng Sun","doi":"10.1007/s12094-025-03870-9","DOIUrl":null,"url":null,"abstract":"Purpose: This study evaluates a three-dimensional (3D) deep learning (DL) model based on fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) for predicting the preoperative status of spread through air spaces (STAS) in patients with clinical stage I lung adenocarcinoma (LUAD).Methods: A retrospective analysis of 162 patients with stage I LUAD was conducted, splitting data into training and test sets (4:1). Six 3D DL models were developed, and the top-performing PET and CT models (ResNet50) were fused for optimal prediction. The model's clinical utility was assessed through a two-stage reader study.Results: The fused PET/CT model achieved an area under the curve (AUC) of 0.956 (95% CI 0.9230-0.9881) in the training set and 0.889 (95% CI 0.7624-1.0000) in the test set. Compared to three physicians, the model demonstrated superior sensitivity and specificity. After the artificial intelligence (AI) assistance's participation, the diagnostic accuracy of the physicians improved during their subsequent reading session.Conclusion: Our DL model demonstrates potential as a resource to aid physicians in predicting STAS status and preoperative treatment planning for stage I LUAD, though prospective validation is required.","PeriodicalId":50685,"journal":{"name":"Clinical & Translational Oncology","volume":" ","pages":""},"PeriodicalIF":2.8000,"publicationDate":"2025-02-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Clinical & Translational Oncology","FirstCategoryId":"3","ListUrlMain":"https://doi.org/10.1007/s12094-025-03870-9","RegionNum":3,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ONCOLOGY","Score":null,"Total":0}

引用次数: 0

Abstract

Purpose: This study evaluates a three-dimensional (3D) deep learning (DL) model based on fluorine-18 fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) for predicting the preoperative status of spread through air spaces (STAS) in patients with clinical stage I lung adenocarcinoma (LUAD).

Methods: A retrospective analysis of 162 patients with stage I LUAD was conducted, splitting data into training and test sets (4:1). Six 3D DL models were developed, and the top-performing PET and CT models (ResNet50) were fused for optimal prediction. The model's clinical utility was assessed through a two-stage reader study.

Results: The fused PET/CT model achieved an area under the curve (AUC) of 0.956 (95% CI 0.9230-0.9881) in the training set and 0.889 (95% CI 0.7624-1.0000) in the test set. Compared to three physicians, the model demonstrated superior sensitivity and specificity. After the artificial intelligence (AI) assistance's participation, the diagnostic accuracy of the physicians improved during their subsequent reading session.

Conclusion: Our DL model demonstrates potential as a resource to aid physicians in predicting STAS status and preoperative treatment planning for stage I LUAD, though prospective validation is required.

查看原文本刊更多论文

基于PET/ ct的三维深度学习模型用于预测I期肺腺癌通过空气空间的扩散。

目的：本研究评估基于氟-18氟脱氧葡萄糖正电子发射断层扫描/计算机断层扫描（18F-FDG PET/CT）的三维（3D）深度学习（DL）模型预测临床I期肺腺癌（LUAD）患者术前通过空气间隙扩散（STAS）状态。方法：对162例I期LUAD患者进行回顾性分析，将数据分为训练集和测试集（4:1）。建立了6个3D深度DL模型，并融合了表现最好的PET和CT模型（ResNet50）以进行最佳预测。该模型的临床效用通过两个阶段的读者研究进行评估。结果：PET/CT融合模型在训练集的曲线下面积（AUC）为0.956 (95% CI 0.9230 ~ 0.9881)，在测试集的AUC为0.889 （95% CI 0.7624 ~ 1.0000）。与三位医生相比，该模型表现出更高的敏感性和特异性。在人工智能（AI）辅助的参与下，医生在随后的阅读过程中诊断的准确性得到了提高。结论：尽管需要前瞻性验证，但我们的DL模型显示了作为一种资源来帮助医生预测STAS状态和I期LUAD的术前治疗计划的潜力。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Clinical & Translational Oncology 医学-肿瘤学

CiteScore

6.20

自引率

2.90%

发文量

240

审稿时长

1 months

期刊介绍： Clinical and Translational Oncology is an international journal devoted to fostering interaction between experimental and clinical oncology. It covers all aspects of research on cancer, from the more basic discoveries dealing with both cell and molecular biology of tumour cells, to the most advanced clinical assays of conventional and new drugs. In addition, the journal has a strong commitment to facilitating the transfer of knowledge from the basic laboratory to the clinical practice, with the publication of educational series devoted to closing the gap between molecular and clinical oncologists. Molecular biology of tumours, identification of new targets for cancer therapy, and new technologies for research and treatment of cancer are the major themes covered by the educational series. Full research articles on a broad spectrum of subjects, including the molecular and cellular bases of disease, aetiology, pathophysiology, pathology, epidemiology, clinical features, and the diagnosis, prognosis and treatment of cancer, will be considered for publication.