Softwareunterstützte semiautomatische Tumorvolumetrie von Lungentumoren

Die Quantifizierung der Tumorlast eines Patienten spielt sowohl in Studien, als auch im klinischen Alltag, eine große Rolle beim Therapiemonitoring. Als Goldstandard gelten die Messkriterien nach RECIST mit der Bestimmung des längsten Tumordurchmessers. In der vorliegenden Studie wurde ein Vergleich...

Full description

Saved in:
Bibliographic Details
Main Author: Haag, Katharina
Contributors: Heverhagen, Johannes (Prof. Dr. Dr.) (Thesis advisor)
Format: Doctoral Thesis
Language:German
Published: Philipps-Universität Marburg 2011
Subjects:
Online Access:PDF Full Text
Tags: Add Tag
No Tags, Be the first to tag this record!

The quantification of a patient‘s tumor burden plays an important role in clinical studies, as well as in the clinical routine for tumor response assessment. Current gold standards are the RECIST criteria based on the measurement of the maximum axial diameter of a tumor lesion. This study compares volumetric assessment of a tumor with manual measurement following the RECIST guidelines. CT-scans of patients with lung nodules were used for comparison of the different measurement methods, evaluation of the reproducibility of the two measurement methods through two readers, as well as arranging the tumor lesions into different tumor response categories to evaluate the differences between the measurement methods. A time analysis was conducted for practicability purposes in the clinical practice. • Comparison of manually measured diameter and volumetric data of a tumor lesion led to different classification in the RECIST tumor response categories. In 32% of the studied cases it could have caused divergent therapy decisions depending on the measurement method. • The investigation of reproducibility of both methods by two different readers showed a correspondence of 84.6% for manual diameter measurement and 92.3% for volumetry, respectively. • Assessment of agreement using different technique and method but the same reader revealed mean values of the differences between manual and semi-automatic measurement of 1.35 mm, and 1.99 mm respectively, with a high significance (p < 0.001). By means of Bland-Altman-Plots it could be demonstrated that manually collected data was generally measured smaller than the semi-automatic. • The volumetric software was tested for practicability in the daily routine by conducting a time analysis. With 0.84 s in the fastest and 23.92 s in the longest case and a mean value of 3.32 s it demonstrated to be a quick method. Time needed for a CT-Evaluation can be reduced. In summary, this study shows that by using a semi-automatic volumetric software lung lesions can be measured in all three dimensions, less time consuming and with better reproducibility than the current gold standard of manual one dimensional measurement by RECIST.