Project leader Mark Schnitzer says that the device weighs just 1.1 grams, and thus can be worn by a mouse without significantly impairing its movement.
He has revealed that his team has already used the device to study the circulation of blood through the one-cell-wide capillaries in the brain of active mice.
The researcher says that the microscope is attached to the head of a mouse under anaesthetic, while a marker dye is injected into the brain to label blood plasma, but leave blood cells unaffected.
According to him, the device uses light delivered by a mercury arc lamp through a bundle of optical fibres, which causes the dyed blood plasma to fluoresce, showing up individual blood cells as dark spots.
The image is sent back up the fibre-optic bundle to a camera that records the image, he adds.
Schnitzer says that nearly 100 images can be taken every second, something that makes it possible for the researchers to watch high-speed video of individual blood cells flowing in the brain.
Once the mouse wakes up from the anaesthetic, according to him, it is possible to watch the movement of cells as it behaves normally.
The researchers have revealed that combining the technique with a dye that makes the activity of brain cells visible, they could see how Purkinje neurons, involved in controlling movement, become more active when a mouse is moving than when resting.
"The advance here is we are able to look at cells in [moving] animals and we can do this in mice - the mammalian species of choice from the perspective of having advanced genetic techniques. So we can look at mouse disease models and see what the cells are doing at the same time as we monitor what the mouse is doing," New Scientist magazine quoted Schnitzer as saying.
Carl Petersen at the Swiss Federal Institute of Technology in Lausanne, Switzerland, thinks that "it is a good advance".
He, however, says that the approach cannot look at all kinds of brain activity. According to him, the brain scatters light extensively, meaning only cells relatively near to the microscope, labelled with dye, can be imaged.
"This was not a big problem for the current study where they look at very brightly labelled structures with very high contrast. However, there are very few structures in the brain that are organised in this way," he says.
Schnitzer, however, disagrees that the technique can only be used to study high contrast structures, and insists that the new microscope detects changes in fluorescence as small as 0.5%.
A research article describing the tiny microscope has been published in the journal Nature Methods.