Microscopic sample rotating under imaging device symbolizing Controlled Rotation Tomography

Revolutionizing Imaging: How Controlled Rotation Tomography Could Change Medicine

"A novel approach to tomography promises clearer images with less radiation, impacting everything from materials science to medical diagnostics."


In the relentless pursuit of higher resolution imaging, scientists continually face a significant hurdle: the trade-off between image clarity and sample integrity. Traditional tomographic reconstruction, a cornerstone of medical and materials science, depends on capturing numerous projections from different angles. However, this necessity poses a challenge, especially when examining beam-sensitive organic or hybrid materials. The incident radiation from electron probes can inflict sample damage, thereby distorting the very data researchers aim to collect. This limitation has spurred the exploration of innovative methods to minimize radiation exposure while maximizing image quality.

Enter Inpainting Assisted Controlled Rotation Tomography, or CORT, a novel sampling strategy designed to redefine the boundaries of what's achievable in imaging. CORT addresses the inherent limitations of conventional techniques by cleverly manipulating the image acquisition process. The essence of CORT lies in its ability to increase the number of projection images obtained for a given exposure time. It achieves this by sparsely imaging the object in real space, tailored explicitly for scanning probe transmission microscopy. This approach recognizes and exploits the principle that adjacent pixels in a single projection image contain overlapping information, thus making it possible to restore images effectively, even from under-sampled data.

The concept of inpainting, which involves algorithms that fill in missing or damaged portions of an image, plays a pivotal role in CORT. By applying inpainting techniques to near-randomly under-sampled electron projection images, researchers can accurately estimate the true projection, thereby enhancing image quality. This method effectively maximizes the utilization of electron exposure, ensuring more detailed and precise imaging. As CORT continues to evolve, its potential applications span various fields, from enhancing medical diagnostics to refining materials science research, promising a future where high-resolution imaging is both more accessible and less destructive.

The Mechanics of CORT: A New Approach to Imaging

Microscopic sample rotating under imaging device symbolizing Controlled Rotation Tomography

Conventional tomography typically involves a staccato-like process where the sample is held stationary at various angles to capture a complete image at each stop. CORT, however, introduces a dynamic shift: the sample rotates continuously as the electron beam scans across defined pixels. This simultaneous movement of both the sample and the scanning probe at varying speeds marks a significant departure from traditional methods. The result is multiple images taken in a single continuous sweep, all preprogrammed and precisely controlled, hence the name Controlled Rotation Tomography.

The innovation of CORT extends beyond mere mechanics; it redefines how image data is processed. Although the microscope still produces individual images, each frame comprises pixels captured at different orientations. These pixels are then sorted based on their orientation, effectively forming a sparsely-sampled projection image. Given that the scanning speed of the electron beam is typically much faster than the rotation of the sample, each pixel integrates data from a narrow range of angles, providing a comprehensive view of the specimen.
To effectively utilize CORT data, the following steps are crucial:
  • Data Sorting: Pixels of the same orientation are grouped together to form sparsely sampled images.
  • Inpainting: Algorithms like beta process factor analysis and 3D wavelet inpainting fill in missing data in the sinogram.
  • Tomographic Reconstruction: The fully recovered sinograms are then used for conventional tomography reconstruction.
The workflow is strategically designed to leverage the strengths of both continuous motion and computational enhancement. By applying inpainting algorithms to the sinogram—a representation of the data collected—researchers can fill in missing pixels, particularly useful when the mask (or pixels that have been visited) forms a check-board pattern due to CORT sampling. This ensures a more evenly distributed and complete dataset for subsequent reconstruction. The ultimate goal is to obtain fully recovered sinograms that can be used for conventional tomography reconstruction, thus achieving high-resolution imaging with minimal sample damage.

Future Implications and Applications

The development and refinement of CORT hold significant promise for the future of imaging across various scientific and medical disciplines. By minimizing radiation exposure and enhancing image resolution, CORT opens new avenues for studying sensitive materials and biological samples that were previously challenging to examine. As computational power continues to grow and inpainting algorithms become more sophisticated, the potential of CORT to transform imaging is only set to increase. The ability to achieve high-quality tomographic reconstructions from sparsely sampled data not only reduces the risk of sample damage but also accelerates the imaging process, making it a valuable tool for research and diagnostics.

Newsletter Subscribe

Subscribe to get the latest articles and insights directly in your inbox.