The High Content Screening Station (HCS) scanR enables fully automated image acquisition and data analysis. Version 3.3 improves the capabilities of deep learning technology to reliably separate objects in biological samples using instance segmentation, the ability to detect and delineate distinct objects of interest in an image.
Precise segmentation of objects: raw data (left), standard threshold (middle), segmentation of TrueAI instances (right). Instance segmentation reliably separates hard-to-distinguish objects that are very close to each other, such as cells or nuclei in colonies or tissues. Image Credit: Olympus Life Sciences Solutions
Reliable image segmentation
Using a self-learning microscopy approach, the scanR system AI automatically analyzes the data in a test-based workflow. Deep learning technology can detect cells, nuclei, and subcellular objects, and extract features from a list of over 100 object parameters. Version 3.3 dramatically improves deep learning object segmentation capabilities to more accurately segment hard-to-distinguish objects, such as cells or nuclei that are very close to each other, such as in cell colonies or tissues.
In addition to the tools for developing neural network models for specific applications, version 3.3 of scanR comes with pre-trained neural network models for nuclei and cells. These can be used in a wide range of standard applications, including the ability to distinguish between confluent cells and dense nuclei, thus eliminating the time required to train the neural network.
Easier calibration and collaboration
Version 3.3 of the scanR software also includes a Well Plate Calibration Wizard that makes it quick and easy to calibrate a new well plate for the system. In addition, a new license level allows collaborators to open, examine and reassign scanR analysis files for easier sharing of results.
Olympus Solutions for Life Sciences