I have observed with some patients that the initial plain film image of a lipohaemarthrosis of the knee can demonstrate an indistinct fat-blood interface. One of my colleagues suggested that this could be due to the mixing of blood and fat associated with movement of the patient onto the X-ray table (or movement of the patient on the gurney/trolley/barouche/stretcher to the X-ray department). If this theory was to hold weight, the fat/fluid interface should become sharper if the image is repeated while the patient is stationary (see below)
	This is the same patient in the same position several minutes later. The fat/blood interface is now sharply defined (arrowed).
adapted from http://thumbs.dreamstime.com/thumb_297/1218067161tWnq3P.jpg	I have no idea whether the suprapatellar pouch can act as a snow globe in the presence of lipohaemarthrosis, but it makes for a good theory!
http://upload.wikimedia.org/wikipedia/commons/a/ab/Snow_Globe_girl.jpg	If you don't know what a snow globe is, it's the toy (often a souvenir) that you shake to produce an effect like falling snow.

The snow globe effect could result in an equivocal finding of lipohaemarthrosis of the knee joint. Given the potential importance of this finding, it should be considered whether the lateral horizontal ray knee projection should be repeated.