Aiforia Create: Velocity Software Release

The image analysis software features and tools with this release speed up image analysis by reducing the amount of training data to train AI models.
Written by Aiforia

Designed for a variety of medical specialties, Velocity enables the production of high-quality results with cloud technology and deep learning AI

The following Velocity tools and features are available – for research use only – to existing and future customers on the Aiforia Platform:

  • Annotation Assistant: Aiforia’s active learning tool significantly reduces AI model creation time by critically reviewing images and searching for areas that can provide good training data. By automatically providing training areas and suggested annotations to the user, this tool drastically speeds up AI model creation.
  • Image Match: Enables a completely new methodology for image analysis by combining tissue registration and overlay with our AI model analysis and advanced spatial metrics. We have fully automated the technically difficult execution of pairing images of serial sections, with different staining, as well as combining AI models of distinct function e.g. cell phenotyping based on IHC with tumor grading based on HE. Results from any analysis, such as object identification or segmentation, can thus be layered and viewed in parallel or superimposed, allowing medical experts to perform spatial measurements across multiple samples.

Additional Platform Improvements:

  • Object Classification, such as detection and quantification of immune cells, has traditionally been limited by large training set requirements. The Velocity release includes AI and analysis engineering solutions that drastically reduce the required size of the training set. Together with Annotation Assistant, these new improvements push Object Classification to a new level of efficiency and ease.
  • Integrated AI model validation tools give the end user an easy interface to define validation sets and to invite colleagues or consultants to provide scoring or diagnosis according to intended use criteria. The validation tool then provides a human versus AI comparison at the pixel level, a necessary requirement for robust AI model evaluation.

The company also recently announced its new product for clinical diagnostics in Europe only:

  • Aiforia Clinical: Aiforia Clinical is an intuitive CE-IVD marked platform that helps researchers and healthcare professionals improve and advance clinical diagnostics by facilitating faster and more informative decision making. Aiforia Clinical can be integrated into any hospital IT infrastructure and has been optimized for the clinical pathology workflow to improve productivity in visual diagnostic tasks when analyzing images from tissue samples.

In EU and EEA countries Aiforia® Clinical is CE marked under EU’s In Vitro Diagnostic Device Regulation 2017/746 for in vitro diagnostic use. In the USA, Japan, and other countries Aiforia® Clinical is for research use only and is not for use in diagnostic procedures.