High-resolution tomographic images are generated by reconstructing a 3D image of an object from several 2D electron microscope images captured from different angles around the object. Aligning the 2D images—a prerequisite to 3D image reconstruction—is a tricky, time consuming task that must be completed by trained specialists.

Leti’s Guidar software automates this critical step. The software leverages a robust, patented denoising, iterative alignment, and 3D reconstruction algorithm that delivers a precise 3D image reconstruction with no action required on the part of the user. “The robust algorithm utilizes graphics processor units to artificially eliminate drift and instabilities created when the sample is rotated,” said a Leti researcher. “The iterative process compares the projections measured to numerical simulations of perfect projections to eliminate artefacts before automatically reconstructing the 3D image.”

Automating the method results in faster, easier 3D image reconstruction and delivers higher resolution (currently around a nanometer) than the traditional method. The software will pave the way for broader use of the technique, which is currently limited mainly to research in fields like biology, microelectronics, materials, energy, and nanoparticles.