Establishment of CT diagnosis of small adenocarcinoma of the lung based on quantitative image analysis

2019 Fiscal Year Final Research Report

• Project/Area Number: 17K10352
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Research Field: Radiation science
• Research Institution: Niigata University
• Principal Investigator: Ishikawa Hiroyuki 新潟大学, 医歯学総合病院, 講師 (90377151)
• Project Period (FY): 2017-04-01 – 2020-03-31
• Keywords: 肺腺癌 / 薄層CT / テクスチャ解析

Outline of Final Research Achievements

Adenocarcinoma in situ and minimally invasive adenocarcinoma were accurately differentiated from invasive adenocarcinoma of the lung using quantitative CT image analysis, which appeared as ground-glass opacity at high-resolution CT.
Reproducibility of quantitative analysis was maintained regardless of experience of CT image interpretation. Quantitative image analysis and automatic feature extraction using artificial intelligence can stratify the postoperative prognosis and may have equal or greater prognostic accuracy than conventional features evaluated by experienced thoracic radiologists.

Free Research Field

医歯薬学

Academic Significance and Societal Importance of the Research Achievements

肺腺癌の治療法の選択や予後の予測において、上皮内腺癌～微少浸潤性腺癌と浸潤性腺癌の鑑別は重要である。この鑑別は薄層CTにより非侵襲的に行われることが期待されているが、視覚的評価のみでは限界があった。今回、定量的画像解析を用いることで良好な鑑別能が示され、新たな可能性を示すことができた。また、定量的画像解析による予後予測能も視覚的評価より優れている可能性が示された。