A drawback to large-scale multicentre studies is the time required for the
centralized evaluation of diagnostic images. We evaluated the feasibility o
f digital transfer of echocardiographic images to a central laboratory for
rapid and accurate interpretation. Ten patients undergoing trans-oesophagea
l echocardiographic scanning at three sites had representative single image
s and multiframe loops stored digitally. The images were analysed in the or
dinary way. AII images were then transferred via the Internet to a central
laboratory and reanalysed by a different observer. The file sizes were 1.5-
72 MByte and the transfer rates achieved were 0.6-4.8 Mbit/min. Quantitativ
e measurements were similar between most on-site and central laboratory mea
surements (all P>0.25), although measurements differed for left atrial widt
h and pulmonary venous systolic velocities (both P<0.05). Digital transfer
of echocardiographic images and data to a central laboratory may be useful
for multicentre trials.