Olivier Lucidarme MD, Presenter: Nothing to Disclose

Jean-Paul Akakpo MD, Abstract Co-Author: Nothing to Disclose

Betty Lauratet MD, Abstract Co-Author: Nothing to Disclose

S. Granberg MD, Abstract Co-Author: Nothing to Disclose

M. Sideri MD, Abstract Co-Author: Nothing to Disclose

R. Mashiach MD, Abstract Co-Author: Nothing to Disclose

Harry Bleiberg MD, Abstract Co-Author: Research Consultant, Advanced Medical System, Brussels, Belgium

Dror Nir PhD, Abstract Co-Author: Consultant, Advanced Medical Diagnostics, NV, Brussels, Begium

Phikippe Autier MD, Abstract Co-Author: Nothing to Disclose

Jean-Pierre LeFranc MD, Abstract Co-Author: Nothing to Disclose

Philippe Andre Grenier MD, Abstract Co-Author: Nothing to Disclose

et al, Abstract Co-Author: Nothing to Disclose

Ovarian HistoScanningTM (OVHS) is an innovative software technology quantifying statistical features of backscattered ultrasounds. Mathematical algorithms initially calibrated through animal experiments and retrospective and prospective human studies quantify changes induced in the backscattered waves by cancerous tissues. Our aim was to prospectively assess the contribution of OVHS for distinguishing benign from malignant ovarian masses.

A total of 269 ovaries secondarily removed from 228 women operated for pelvic mass in 9 different European institutions were prospectively studied. 3D transvaginal ultrasound was performed before surgery and voxel data was analyzed using OVHS. Results were compared to histology and, to diagnosis made by radiologist based on ultrasound image only.

Histology reported 176 non cancerous ovaries [normal (n=87), benign tumors (n=80), past endometriosis (n=9)] and 93 cancers [borderline cancers (n=21), adenocarcinomas (n=53), carcinomatosis (n=12), metastases (n=4) and other cancers (n=3)]. 91/93 cancers were correctly identified by OVHS given a sensitivity of 98% while the sensitivity of the radiologist was 75% (P<0.0001. Reasons for false negatives of OVHS were: lack of relevant voxel data (n=1) and pathological volume bellow US resolution (Cystadenofibroma with foci of low malignant potential; Borderline) (n=1). The overall specificity of OVHS was 77%. In a subpopulation (n=173) for which the received gain level was maintained below a threshold of -4 dB the OVHS specificity increased to 92%, for a radiologist specificity of 82 % (p=0.0065). Identical study using 3D-TVS performed with controlled gain level is ongoing with results expected by July 2006.

OVHS is highly sensitive for the diagnosis of malignant ovarian masses while having a good specificity.

OVHS could be used as a computer aided diagnostic tool for the non invasive characterization of pelvic masses using 3D transvaginal ultrasound examination.

Lucidarme, O, Akakpo, J, Lauratet, B, Granberg, S, Sideri, M, Mashiach, R, Bleiberg, H, Nir, D, Autier, P, LeFranc, J, Grenier, P, et al, , HistoScanning: A New Computer-aided Diagnostic Tool for Distinguishing Benign from Malignant Ovarian Masses: A Multicentric Study Report.  Radiological Society of North America 2006 Scientific Assembly and Annual Meeting, November 26 - December 1, 2006 ,Chicago IL. http://archive.rsna.org/2006/4429810.html