RSNA 2014 

Abstract Archives of the RSNA, 2014


SST07-06

Diffusion Imaging of Gynaecological Neoplasms as Aids to Characterization

Scientific Papers

Presented on December 5, 2014
Presented as part of SST07: Genitourinary (New Technology for Imaging the GU Tract)

Participants

Emily Lemoniati MBBS, Presenter: Nothing to Disclose
Tina Mistry MBBS, FRCR, Abstract Co-Author: Nothing to Disclose
Vivek Malhotra MD, Abstract Co-Author: Nothing to Disclose
ANDY COADY MBBS, Abstract Co-Author: Nothing to Disclose
Malcolm Padwick MD, Abstract Co-Author: Nothing to Disclose
Anwar Roshanali Padhani MD, Abstract Co-Author: Advisory Board, Acuitas Medical Ltd Advisory Board, Siemens AG Speakers Bureau, Siemens AG Researcher, Siemens AG Speakers Bureau, Johnson & Johnson

PURPOSE

To describe and assess quantitative parameters derived from diffusion weighted (DWI), diffusion tensor (DTI) and diffusion kurtosis (DKI) imaging for characterization of gynaecological masses.                                                  

METHOD AND MATERIALS

Retrospective evaluation of 76 scans in women with gynaecological masses was performed. Multi-directional (b-values: 0 & 800 s/mm2) and 3 scan-trace (b-values: 100, 1100, 1600 s/mm2) DW sequences were evaluated. Regions of interest (ROI) measurements of muscle-normalized b800 signal intensity (nb800SI), apparent diffusion coefficient (ADC0-800), fractional anisotropy (FA0-800) and kurtosis coefficient (Kapp100-1600) of masses, urine and gluteal muscle were correlated with histology and compared using descriptive statistics & non-parametric distribution statistics (Kruskal-Wallis Test). 

RESULTS

10 benign and 20 malignant endometrial lesions, 8 cervical cancers, 17 benign and 9 malignant ovarian lesions, 9 fibroids, 3 sarcomas were compared with 75 gluteus medius muscle and 73 urine measurements. Significant differences (p=0.003) in nb800SI values were identified between endometrial lesions (benign: mean 3.4 (±1.6) and malignant: mean 6.0 (±1.0)). Malignant endometrial and cervical lesions had lower ADC values than benign endometrial lesions (p<0.001). No ADC difference was seen between benign and malignant ovarian masses. Malignant lesions had lower FA values (endometrial cancer: mean 0.18 (±0.06), cervical cancer: mean 0.17 (±0.05)) in comparison to benign myometrial lesions (mean 0.22 (±0.15)) and muscle (mean 0.36 (±0.06)). Benign lesions had greater FA values in comparison to muscle (7 lesions) and maligant lesions (12 lesions). Malignant cervical and endometrial lesions had higher Kapp values than benign endometrial, myometrial & ovarian lesions (p=0.002). Significant differences in Kapp values for malignant versus benign endometrial lesions were noted (2.1 (± 0.04) & 1.2 (± 0.04); p=0.002).  

CONCLUSION

Quantitative diffusion parameters have some discriminatory value for characterizing gynaecologic masses despite overlapping values related to necrosis and cystic change. The organised stroma of benign lesions is noted on DTI. DKI can reflect the high cellular complexity of malignant masses.

CLINICAL RELEVANCE/APPLICATION

Quantitative pelvic diffusion MRI using DWI, DTI and DKI is technically feasible. They should be evaluated for their ability to assist in the characterization of gynaecological masses.

Cite This Abstract

Lemoniati, E, Mistry, T, Malhotra, V, COADY, A, Padwick, M, Padhani, A, Diffusion Imaging of Gynaecological Neoplasms as Aids to Characterization.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14014853.html