# Evaluating the Quality of Serial EM Sections with Deep Learning

## TL;DR

This paper introduces a deep learning model to assess the quality of serial electron microscopy images, helping to improve imaging efficiency and data quality.

## Contribution

A modified ResNet-50 model, called QEN, is developed to reliably predict user-assigned image quality scores for ssSEM.

## Key findings

- QEN reliably predicts user-generated quality scores for ssSEM images.
- Running QEN in parallel with image acquisition helps identify and flag problematic images quickly.
- The code and training dataset for QEN are publicly shared for use and further development.

## Abstract

Automated image acquisition can significantly improve the throughput of serial section scanning electron microscopy (ssSEM). However, image quality can vary from image to image depending on autofocusing and beam stigmation. Automatically evaluating the quality of images is, therefore, important for efficiently generating high-quality serial section scanning electron microscopy (ssSEM) datasets. We tested several convolutional neural networks for their ability to reproduce user-generated evaluations of ssSEM image quality. We found that a modification of ResNet-50 that we term quality evaluation Network (QEN) reliably predicts user-generated quality scores. Running QEN in parallel to ssSEM image acquisition therefore allows users to quickly identify imaging problems and flag images for retaking. We have publicly shared the Python code for evaluating images with QEN, the code for training QEN, and the training dataset.

Graphical Abstract