Fourier Ptychographic Microscopy

Add One

Two of the key parameters that describe the performance of the optical microscope are its resolution and its field-of-view. In fact, these two parameters are coupled: switching to a higher magnification objective will improve the resolution, but also tend to reduce the field-of-view. This trade-off is encapsulated in the idea of the space-bandwidth product, which is (conceptually at least) a measure of how many useful pixels of information an imaging system can transmit. Typical microscopes and microscope objectives are limited to around 10 Megapixels; if we make these pixels smaller by increasing the resolution then the area covered must be reduced as well. So if we want to image large areas at high resolution, as we might want to do in histology for example, then we have to mechanically scan the slide underneath the microscope and stitch multiple images together.

A team from Caltech have suggested a way of increasing the effective space-bandwidth product of a microscope without mechanical scanning1. Rather than acquiring multiple high resolution, small field-of-view images and combining them to synthesise a large field of view, they acquire multiple low resolution, large field-of-view images using a low magnification objective. For each image, the sample is illuminated from a different angle using an array of LEDs. So each image contains a portion of the spatial frequency information that would be collected by a high magnification objective, but for a large field-of-view.

To reconstruct a high resolution image, it’s then necessary to combine the various pieces of spatial frequency information. This is similar to the idea of the synthetic aperture microscope, but with a crucial difference. Normally, in order to recover a high resolution image it would be necessary to combine the images coherently, taking into account both phase and amplitude. But the Caltech team used incoherent light, with intensity-only detection. They were able to show that, using a phase recovery algorithm, they can reconstruct the high resolution, large field-of-view image without needing to make phase measurements. This makes their system low-cost and simple to retrofit to any existing microscope: simply add an array of LEDs.

While the image acquisition procedure is likely to be quicker than the conventional scanning approach, the disadvantage at the moment is the long-time taken to reconstruct the image, almost 30 minutes. However, the algorithm can be parallelised, suggesting that a significant speed-up might be possible using a GPU or other dedicated hardware. And there are other intriguing possibilities; the authors have subsequently shown that the phase recovery is reasonably accurate and quantitative, meaning that phase-resolved imaging is possible2 Once you have phase information, it’s also possible to correct any images that are out-of-focus, which again could help accelerate throughput. The technique has achieved some media attention, including a feature in optics and electronics news,


  1. Zheng, Guoan, Roarke Horstmeyer, and Changhuei Yang. “Wide-field, high-resolution Fourier ptychographic microscopy.” Nature Photonics 7, no. 9 (2013): 739-745.(PDF)
  2. Ou, Xiaoze, Roarke Horstmeyer, Changhuei Yang, and Guoan Zheng. “Quantitative phase imaging via Fourier ptychographic microscopy.” Optics letters 38, no. 22 (2013): 4845-4848.

Leave a Comment

Your email address will not be published. Required fields are marked *


You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>