NU 2015-143

Inventors

Vinayak Dravid*

Benjamin Myers

Karl Hujsak

Short Description

A strategy to capture highly accurate microscopy images in a time- and cost-efficient manner.

Abstract

Northwestern researchers have developed a new technology, using a combination of hardware and software, to construct microscopic images more efficiently based on incomplete signal data.  They propose a method that can reconstruct and “de-noise” images without being plagued by typical time- and dosage-dependent signal damage.  In some microscopy cases, it is not feasible or practical to capture or image all the data from a one-dimensional signal, two-dimensional image, or three-dimensional structure. However, there is a need to use an incomplete data set to produce a complete representation while maintaining a high a degree of accuracy and minimizing the need for extensive modification of the detection/sampling equipment.  Northwestern researchers have developed this new strategy which is comprised of the several components: 1) a signal source, 2) a machine-learning algorithm to statistically fill in missing information, and 3) a custom-written graphical user interface (GUI)-equipped computer program. 

Applications

• Image reconstruction in various types of microscopy:
• Fluorescent microscopy (confocal, 2-photon)
• Electron Microscopy
• Atomic force Microscopy

Advantages

• Economical
• Time-saving signal acquisition
• No need for scanning hardware
• Easy translatable use for image reconstruction in other (non-EM) types of microscopy

IP Status

A US patent application has been filed.