fairfolio
Saved Projects

Using Object Detection Machine Learning to Count and Categorize C. Elegans

AJAS · 2020 Robotics and Intelligent Machines (inferred)

Celine Tan, Matthew Adner · and

Overview

C. elegans worms are a very powerful model organism for conducting research: they are easy to cultivate, have consistent life cycles, and their genome has been fully mapped. As a result, they are used by labs around the world to conduct research. Counting C. elegans is tedious and time consuming, especially for those who are new to using this model organism. To combat this problem, we used Apple’s ML-Kit to train an object detection model to identify worms and their developmental stages. Object detection is a type of computer vision that is used to detect the location and number of certain objects or classes. Our model uses a deep learning algorithm called a Region-based Convolutional Neural Network (R-CNN) for this. To create training data for the model, we manually annotated images, marking regions of an image containing worms, with MakeML. The algorithm was then trained and tested with CreateML. The annotated pictures were taken at various magnifications through an Olympus SZ4045 stereozoom microscope with an Olympus PEN Lite E-PL7 camera. The algorithm was trained, using these annotated pictures, to box and classify any worms it sees in a picture. We compared the results from pictures annotated by the algorithm to the results of a mediocre human counter (a novice C. elegans researcher) using an excellent human counter (an experienced C. elegans researcher) as the accepted value. The algorithm’s requirements include high contrast images, well-focused images, and images that are not overcrowded. If an image meets these requirements, the algorithm can outperform the moderately-experienced human counter by 3% on accuracy of the total number of worms and 8% on the correct identification of the worms’ developmental stages.

Competition history

  • AJAS 2020 Category not listed

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

Source: AAAS Annual Meeting (Confex) / American Junior Academy of Science

Save projects to your library

Sign in with Google to keep track of projects you find interesting, organized into folders. An account also raises your daily allowance for “Has this been done?”, and lets you create a key for the MCP server with a much higher limit than anonymous use. Browsing stays public.

Continue with Google