RESETTING OF 3-DIMENSIONAL MEDICAL IMAGES - APPLICATION TO MRI AND CT

Citation
E. Cuchet et al., RESETTING OF 3-DIMENSIONAL MEDICAL IMAGES - APPLICATION TO MRI AND CT, Journal of neuroradiology, 21(3), 1994, pp. 170-175
Citations number
6
Categorie Soggetti
Radiology,Nuclear Medicine & Medical Imaging",Neurosciences
Journal title
ISSN journal
01509861
Volume
21
Issue
3
Year of publication
1994
Pages
170 - 175
Database
ISI
SICI code
0150-9861(1994)21:3<170:RO3MI->2.0.ZU;2-Z
Abstract
Several kinds of images, each giving a different information, are now available to radiologists. The MRI images have excellent contrast reso lution and enable soft tissues to be differentiated, but they do not d istinguish structures with low water content, notably air and bone, wh ereas these are easily recognized by CT. The aim of this study is to p resent a simple, entirely radiologist-supervised method to examine the radiological data of any patient, obtained from several kinds of imag es. MRI is performed using a GEMS Signa, 1.5 Tesla, 4.9 version magnet . Acquisitions are T1- or T2-weighted spin-echo or gradient sequences, with a 256 or 512 matrix, on axial sections, with of without contrast injection. CT is performed using a GEMS Hi Speed scanner. Acquisition s are obtained on a 512 matrix and with a ''Soft'' or ''Bone'' filter, without contrast injection. The two series of sections are transmitte d, through an Etherne network, to a Sun console where the two correspo nding volumes are reconstructed on a GEMS Voxtool by means of a 3-dime nsional soft ware for image treatment. At least 3 couples define the r otation and translation required for one of the two volumes to reset i t in the guide mark of the other. The soft ware then looks for the bes t transformation, in terms of least square between the two 3-dimension al volumes. The calculation demands only a few seconds. One of the two objects is then recalculated in the guide mark of the other. The curs or positioned by the user on any point of the object is linked to a se cond cursor which will automatically position itself on the correspond ing point of the other object. The accuracy obtained (about one millim etre) is specified by the soft ware which indicates how to improve res etting. In addition to its teaching value, this superimposition image can help in the diagnosis and can be used for surgical stimulation bec ause it is possible to mix the images. This mixing gives access to a n ew type of imaging, since the images spared can be reconstructed in vo lume, and treated in all planes, as a CT or MRI examination. The term ''Anatomical Reconstruction Images'' may be suggested for this new typ e of examination. Beside intermodal comparison, one may also imagine t hat the soft ware can be used to follow up the patient over time (repe at MRI) or to make comparisons between several objects, although the e lastic resetting method is more often appropriate in the second case.