The aim of this clinical study was to evaluate the use of high-resolution m
agnetic resonance imaging (MRI) in the diagnosis of facial haemiatrophy, A
total of 14 patients with clinically suspected facial haemiatrophy were inv
estigated using high-resolution MRI. The T1- (500/25) and T2- (2200/50) wei
ghted images were analysed visually and numerically. The results of the aff
ected skin portions were compared with the contralateral skin and correlate
d with the clinical results. The subcutis could not be delineated by high-r
esolution MRI in 9 patients with facial haemiatrophy. The dermis was not di
scernible in 6 cases and was "smooth" in a further 6 patients. The signal-t
o-noise ratio of affected skin portions and contralateral skin or clinical
severity did not correlate. The higher the clinical severity, the more pron
ounced was the magnetic resonance ratio between dermis and subcutis thickne
ss. Thus high-resolution MRI is a useful method for objective description o
f pathological changes in clinically suspected facial haemiatrophy.