Image inversion interferometers have the potential to significantly enhance the lateral resolution and detection efficiency of fluorescence scanning microscopes. Constructive interference between a non-inverted and an inverted image of a point source leads to a strong signal for on-axis sources, while sources away from the optical axis yield a weak signal, resulting in an improved resolution.

Background – confocal microscopy

The confocal fluorescence microscope is the workhorse of the biomedical sciences. By combining laser scanning illumination and detection through a pinhole, they eliminate unwanted out-of-focus contributions in the acquired images. Additionally, the use of a small pinhole also leads to a slight improvement of the lateral resolution. Unfortunately, these benfits come at the price of reduced light efficiency, as not only out-of-focus light is blocked.

The idea – image inversion interferometers

An image inversion interferometer in the detection pathway of a confocal microscope can improve the resolution of such system without the need for a pinhole.

After descanning, the image inversion interferometer splits the light in the detection path into two components. These are rotated laterally by 180° relative to each other before interferometrically being recombined and detected in the constructive and destructive channel using integrating detectors (e.g. PMTs), as shown schematically in the figure below.

 UZ-interferometer Principle-Setup

For the wavefront eminating from a single fluorophore the result can be visualied using three cases:

  1. If the fluorophore is located on the optical axis (i.e., the image inversion axis) of the microscope, the non-inverted amplitude point spread function (APSF) and the inverted APSF will be identical. They can therefore interfere perfectly; all of the signal will end up at the constructive detector.
    Here, the 1st (left) image shows the combined amplitude distribution in the constructive pinhole plane, whereas the 4th image shows the corresponding distribution in the destructive channel. The 2nd (3rd) image show the intensity distribution in these planes before integration. Note that these assume full-field illumination, the amount of light emitted by the fluorophore does not change with position.
  2. If the fluorophore is located far from the optical axis, the non-inverted and inverted APSF will have no spatial overlap and therefore will not interfere. Half of the light will end up at each of the detectors.
  3. For a fluorophore located at intermediate distances, the two wavefronts will have some (but not perfect) spatial overlap. The will interfere, but not perfectly, leading to signal in both detectors.
Skip to toolbar