Fully automated relocalization of samples between different microscopy systems
Abstract
The inventive method allows samples to be examined using different microscopy systems, thereby combining the advantages of high-throughput of low-resolution methods with low-throughput, high-resolution microscopy. The method can thus help solve the problem of retrieving the same structures of a sample in different microscopy systems down to the sub-µm range.
Background
From widefield microscopy with low resolution at high throughput to super-resolution microscopy at very low throughput – there is a wide range of systems available on the market. Despite the fact that all types of microscopy systems have become increasingly automated in recent years, the transfer of samples between the various systems proves to be difficult or can only be performed manually.
Problem
In basic biological and medical research as well as in the pharmaceutical industry, there is a great need to combine the advantages of different microscopy systems.
Solution
In a research project funded by the Baden-Württemberg Foundation, scientists at Heidelberg University have developed a process and a software solution for the fully automated and platform-independent relocalization of sample substrates. This inventive method allows samples to be examined using different microscopy systems, thereby combining the advantages of high-throughput of low-resolution methods with low-throughput, high-resolution microscopy. Multiscale imaging, the essential feature of the invention, allows relevant matches to be identified in a sample using high-throughput microscopy. These matches will then be analyzed using high-resolution microscopy with near-molecular resolution. The method according to the invention can thus help solve the problem of retrieving the same structures of a sample in different microscopy systems down to the sub-µm range.
This is achieved by a two-step process: In the first step, a high-precision reference matrix of the sample is created by using reference points. This matrix provides absolute coordinates on the sample substrate, which allows the examined structures to be retrieved, independent of the unavoidable twisting of the sample substrates in the image. Accuracy in the range of 5 µm is already achieved in this step. In the second step, fine adjustment is carried out using pattern recognition methods supported by image processing. The target area is identified by the shape and arrangement of the objects or markers. Correction parameters are then returned to the controller and used to approach the target area. The resolution is only limited by the pixel size of the microscope used.
Figure: Exemplary approach: Sample preparation (1.) and imaging using the first system (2.). Identification and marking of interesting areas (3.). Transferring the sample (4.) and referencing it (5.) on another microscope. Automatic relocalization and imaging of the previously identified areas using the second system (6.).
Advantages
- Fully automated localization or relocalization
- Independent of platforms and reference objects
- Use of any reference points on any sample substrate
- Fine adjustment supported by image processing
- Broadening the range of applications of existing microscopy systems
Fields of application
The method presented here offers the first reliable solution that allows users to automatically retrieve objects when switching between systems.