Your browser doesn't support javascript.
loading
Texture Feature Analysis in Follow-up of Pulmonary Ground Glass Nodule / 中国医学影像学杂志
Chinese Journal of Medical Imaging ; (12): 441-446,451, 2017.
Article in Chinese | WPRIM | ID: wpr-614929
ABSTRACT
Purpose At present,morphological observation and CT value measurement were mainly used to evaluate ground-glass nodule (GGN),and there was no effective image feature-quantization evaluation method.Therefore,in this study,a follow-up quantization analysis was conducted on GGN within 2 years using texture feature analysis method to confirm reasonable GGN follow-up time.Materials and Methods Baseline and highresolution CT images of 100 GGN follow-up patients were retrospectively analyzed.They were assigned into three groups,3 months follow-up (group A),6 to 12 months follow-up (group B) and 2 years follow-up (group C).For each group,using firstly founded GGN image as baseline,GGN texture features (including energy,contrast,autocorrelation,inverse difference moment and entropy) were analyzed.Results There were 1 case of narrowed nodules in group A,1 case of increased nodules and 1 case of narrowed nodules in group B,and 4 cases of increased nodules in group C,2 of which showed density differences.There was no significant change in shape,density and size of the remaining nodules.There were no significant differences in texture features (energy,contrast,autocorrelation,deficit,entropy) among group A,group B and group C (P>0.05).Conclusion Texture feature analysis can quantitatively evaluate the change of GGN attribute characteristics,and as a GGN follow-up quantitative tool,it can guide patients to choose reasonable follow-up mode.

Full text: Available Index: WPRIM (Western Pacific) Type of study: Prognostic study Language: Chinese Journal: Chinese Journal of Medical Imaging Year: 2017 Type: Article

Similar

MEDLINE

...
LILACS

LIS

Full text: Available Index: WPRIM (Western Pacific) Type of study: Prognostic study Language: Chinese Journal: Chinese Journal of Medical Imaging Year: 2017 Type: Article