European journal of surgical oncology : the journal of the European Society of Surgical Oncology and the British Association of Surgical Oncology · 2026
Artificial intelligence-enhanced three-dimensional radiological analysis for predicting spread through air spaces in early-stage lung adenocarcinoma.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Amemiya Ryosuke, Kudo Yujin, Matsubayashi Jun, Ichinose Akimichi, Park Jinho, Tanei Saki, Nagase Wakako, Takada Ikki, Ito Maki, Furumoto Hideyuki, Hagiwara Masaru, Kakihana Masatoshi, Ohira Tatsuo, Nagao Toshitaka, Masumoto Jun, Ikeda Norihiko
Аннотация
OBJECTIVES: Spread through air space (STAS) is a pathological feature that correlates with poor prognosis, especially in patients with lung adenocarcinoma undergoing sublobar resection. The microscopic nature of STAS complicates direct detection using preoperative radiological imaging. We aimed to develop an artificial intelligence (AI)-assisted three-dimensional (3D) computed tomography (CT) imaging model to preoperatively predict STAS in early-stage lung adenocarcinoma.
METHODS: This retrospective study included 364 patients with stage 0-IA3 lung adenocarcinoma who underwent complete surgical resection between 2018 and 2021. Preoperative CT images were analyzed using 3D image analysis with AI. This process involved automatic quantification of pulmonary nodule volume and assessment of 22 distinct imaging characteristics of the nodules. Pathological analysis of STAS was conducted on excised specimens, and its relationship with imaging features was explored.
RESULTS: STAS was identified in 107 patients. Compared with a non-STAS group (n = 257), patients with STAS had a significantly higher median standard uptake value on positron emission tomography/CT (4.5 vs. 2.0, p < 0.01). AI-driven 3D analysis showed that STAS-positive tumors exhibited a significantly greater solid component volume and percentage (1.37 cm3 and 68.6%, respectively) compared with STAS-negative tumors (0.38 cm3 and 26.0%, respectively; p < 0.01). Utilizing 22 imaging features derived from AI-driven 3D analysis, multivariate analysis identified lobulated edge (p = 0.002) or pleural hypertrophy (p = 0.022), and the absence of ground-glass opacity (p = 0.029) as independent predictive factors for STAS.
CONCLUSION: AI-driven 3D imaging can enhance preoperative STAS prediction for the risk stratification of lung adenocarcinoma, potentially optimizing surgical decisions and treatment strategies.
METHODS: This retrospective study included 364 patients with stage 0-IA3 lung adenocarcinoma who underwent complete surgical resection between 2018 and 2021. Preoperative CT images were analyzed using 3D image analysis with AI. This process involved automatic quantification of pulmonary nodule volume and assessment of 22 distinct imaging characteristics of the nodules. Pathological analysis of STAS was conducted on excised specimens, and its relationship with imaging features was explored.
RESULTS: STAS was identified in 107 patients. Compared with a non-STAS group (n = 257), patients with STAS had a significantly higher median standard uptake value on positron emission tomography/CT (4.5 vs. 2.0, p < 0.01). AI-driven 3D analysis showed that STAS-positive tumors exhibited a significantly greater solid component volume and percentage (1.37 cm3 and 68.6%, respectively) compared with STAS-negative tumors (0.38 cm3 and 26.0%, respectively; p < 0.01). Utilizing 22 imaging features derived from AI-driven 3D analysis, multivariate analysis identified lobulated edge (p = 0.002) or pleural hypertrophy (p = 0.022), and the absence of ground-glass opacity (p = 0.029) as independent predictive factors for STAS.
CONCLUSION: AI-driven 3D imaging can enhance preoperative STAS prediction for the risk stratification of lung adenocarcinoma, potentially optimizing surgical decisions and treatment strategies.