Auto-referencing in digital holographic microscopy reconstruction

1. A computer-implemented method for analyzing digital holographic microscopy (DHM) data for hematology applications, the method comprising:

receiving a plurality of DHM images acquired using a digital holographic microscopy system, each DHM image comprising depictions of one or more cell objects and background;

generating a reference image based on the plurality of DHM images comprising only contributions from the background of the plurality of DHM images;

using the reference image to reconstruct a fringe pattern in the DHM image into an optical depth map,

wherein the reference image is generated by:

applying one or more filtering algorithms to dynamically estimate the reference image,

wherein applying the one or more filtering algorithms comprises:

computing intensity statistics for each pixel region across the plurality of DHM images;

selecting stable distribution values corresponding to the background from the intensity statistics.

2. The method of claim 1 , wherein the plurality of DHM images are received in succession and the one or more filtering algorithms are applied to each DHM image by: updating a previously generated reference image by applying a filter to the DHM image and a predetermined number of previously received DHM images to yield a new reference image.

3. The method of claim 2 , wherein the filter comprises a moving average filter.

4. The method of claim 2 , wherein the filter comprises a Gaussian filter.

5. The method of claim 2 , wherein the filter comprises a median filter.

6. The method of claim 2 , wherein the predetermined number of previously received DHM images is selected by a user.

7. The method of claim 1 , further comprising:

receiving an object-free original reference image,

wherein the reference image is generated by modifying the object-free original reference image to reflect changes in imaging conditions occurring in the plurality of DHM images.

8. The method of claim 1 , wherein generating of the reference image is triggered by switching from one patient blood sample to another with a microfluidic or spotting control unit.

9. The method of claim 1 , wherein the reference image is generated at user-specified time intervals and reconstruction of the fringe pattern in the DHM image into the optical depth map is performed using a most recently computed reference image.

10. The method of claim 1 , further comprising:

generating a real-time visualization of one or more fringe-free reconstructed phase images using the plurality of DHM images and the fringe pattern.

11. The method of claim 1 , further comprising:

generating a 3-D visualization of a cell surface using the plurality of DHM images and the fringe pattern.

12. A system for analyzing digital holographic microscopy (DHM) data for hematology applications, the system comprising:

a DHM system configured to acquire a plurality of DHM images, each DHM image comprising depictions of one or more cell objects and background;

one or more processors;

a non-transitory, computer-readable storage medium in operable communication with the processors, wherein the computer-readable storage medium comprises one or more programming instructions that, when executed, cause the processors to:

generate a reference image comprising only contributions from the background of the plurality of DHM images; and

use the reference image to reconstruct a fringe pattern in the DHM image into an optical depth map,

wherein the reference image is generated by:

applying one or more filtering algorithms to dynamically estimate the reference image,

wherein applying the one or more filtering algorithms comprises:

computing intensity statistics for each pixel region across the plurality of DHM images; and

selecting stable distribution values corresponding to the background from the intensity statistics.

13. The system of claim 12 , wherein the one or more processors comprise a plurality of graphical processing units.

14. The system of claim 12 , wherein the one or more processors further comprise a plurality of central processing units.