Introduction

The invention relates to a computer-implemented method for determining the ventricular volume of fetuses from diffusion-weighted magnetic resonance images, defined as DWI (Diffusion Weighted Imaging) . The invention concerns an NMR diagnosis system and an NMR diagnostic method that uses the method of the invention.

Technical features

VM Detector is a software that performs post-processing analysis of DWI images and can calculate the volume of the ventricles of the fetal brain. VM Detector, on the other hand, using diffusion-weighted images (DWI) that increase the contrast between brain tissue and water in the ventricles, compared to a T2-weighted image and an open-source machine learning algorithm for the classification of data groups, provides extremely precise segmentations of the images, separating the pixels relating to the ventricular areas from the rest of the brain in a clear and accurate way. VM Detector may in the future be exported from the MATLAB @ environment in which it was developed, to allow an independent usufruct from the further purchase of MATLAB @. The better agreement of volumes with outcomes with respect to diameters makes VM detector an extremely advantageous tool both for fetal diagnosis and for neonatal prognosis and treatment.

Possible Applications

• Improvement of prenatal diagnosis;
• Development of Ventricular volume growth curves as a function of GA.

Advantages

• Efficient;
• Precise and Quick;
• Enhancement of prenatal diagnosis;
• Ease of use of the program.