Negative stain transmission electron microscopy (nsTEM) can provide an informative and unique insight into the characterization of viral gene delivery platforms. Traditionally, the analysis of such nsTEM images required an experienced microscopist to identify and categorize the particles of interest. A neural network pre-trained on thousands of nsTEM images on a broad range of expert segmented viral morphologies is used as an adaptive backbone for segmentation and classification of new nsTEM material, specifically geared towards the gene and cell therapy field. Using only a handful of image examples, a user can guide the pre-trained network and instantly obtain automatic and deterministic segmentation and classification of the new images. Leveraging a solid backbone model, the network can adapt to both heterogeneous and pleomorphic new particles. This versatility enables investigatory analyses in a fast and continuous workflow. This is particularly useful when analyzing for example lentiviral vectors, where a pre-set definition guideline for detection is challenging. This morphological characterization of gene therapy delivery platforms can support and speed up the drug development process.

Harnessing expertise for creating smart automated image analysis: Training an AI for segmentation and classification of pleomorphic particles in nsTEM for gene therapy

Macromolecular Chemistry: The Second Century

[Author Benefits](https://scimeetings.acs.org/?utm_source=pubs_content_marketing&utm_medium=website&utm_campaign=0320_MCF_NPI_Launch_Spring_Homepage&ref=pubs_content_marketing)

[How to use SciMeetings](https://storage.googleapis.com/dl.morressier.com/acs/SciMeetings-ACSSpring2020-HowToGuide_M3020-Final.pdf)

[Frequently Asked Questions](https://storage.googleapis.com/dl.morressier.com/acs/SciMtng_ACSNMSp2020_FAQ_M3120_Final.pdf)

ACS Spring 2020 National Meeting & Expo

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.