Link from LCD Pinhole Confocal Microscope

Apologies for a somewhat geeky post.

First to explain confocal microscopy. In a conventional microscope, only a narrow plane is in focus, but light from out-of-focus areas (closer to and further from the objective lens) also contributes to the image. The blurred foreground and background images contribute noise to the image, especially if (as is usually the case in biology) you're looking at a semi-transparent object like a cell.

The confocal microscope is a brilliant solution to this problem. The light coming from the objective is focussed through a point (imagine the light rays forming an hour- glass shape, or rather the shape of a diabolo - two cones point-to-point). At this point in space, you put an opaque screen with a pinhole in it, like putting a tight collar around the waist of the hourglass. The result (easier to understand if you sketch out the light rays on paper) is that light coming from behind or in front of the focal plane is blocked (because those rays do not focus to the same point, and hence hit the screen rather than going through the pinhole).

The result is that you can image a perfect "optical section" through the cell or whatever. You can also take multiple optical sections, to reconstruct a true 3-D image.

The downside, though, is that the pinhole blocks all of the image except for a focal point (ie, light rays coming from points on either side are also blocked by the screen, as well as those from in front or behind). Therefore, confocal microscopes use a raster-scan technique, collecting the image point-by-point.

The raster scanning can be done in various ways. Typically, a pair of mirrors (X and Y) is used to shift the focal point in a raster pattern. On good systems, the scanning can be done at better-than-TV rates, so you can collect maybe a hundred full images per second (important for looking at dynamic processes). Another solution is to use a "Nipkow disc", which is a spinning disc with multiple pinholes; each pinhole moves across the image-plane in turn, to give the same effect as a single pinhole scanning the image.

(Note - to avoid confusion - this isn't a "pinhole lens"; the microscope objective is a normal one, and collects much more light than a pinhole would; that light is then funnelled through the pinhole to block the out-of-focus parts of it.)

Anyway, as far as I know, the raster-scanning of the pinhole is always done mechanically.

Now, it is possible to make LCD screens with very small pixel sizes, and I'm guessing that it would be possible to make a small LCD screen with pixel sizes similar to the pinhole size in a confocal microscope.

So, why not replace the pinhole (and the mechanics needed to scan the point-image) with a small, dense LCD array. The array would be set to "black", except for a single clear pixel, and the position of this clear pixel could be scanned to produce the confocal image, without any mechanics.

(Note: the pixel-size of the LCD array isn't really very important; you can funnel the light in such a way that a 100µm "pinhole" correponds to a diffraction-limited spot on the image, for instance. But small pixels would be convenient.)

I know this would play havoc with polarised light but, in most biological applications, polarization doesn't enter into things (it does in a few case, though, for which this method would not work.)

Incidentally, Googling did turn up a "virtual pinhole" confocal using LCDs. However, as far as I can tell, it's a light-field camera which collects data allowing a 3D image (or slice thereof) to be reconstucted in silico, rather than using the LCD as a moving pinhole.