Leica Biosystems Presentation
Title: “Antibodies to Artificial Intelligence” sponsored by Leica Biosystems
- To understand the impact that pre-analytics can have on Digital Pathology
- To understand the impact of image quality, objectives, image compression and file format has on downstream analysis
- To understand where machine learning has come from and where Artificial Intelligence (AI) is leading too. Applications and considerations for development.
Today both Industry and Academia are heavily investing in the area of Artificial Intelligence (AI). The applications are largely aiming to enhance pathologist workflow efficiencies and potentially enhance the accuracy of human reviews within histological evaluations. Machine learning has already proven that it is possible to train computers to try and emulate the performance of humans. However, with advances in technology and computing power there is an understanding that AI will provide the tools to be able to move past emulations and toward enhancement on what is currently possible with the human eye. AI is currently deployed in other industries; however, histology is a highly complex science, with much more intricacies than other domains.
During this talk we aim to enlighten the audience on the impact that pre-analytical variables and image digitization will have on deep learning classification. Tissue Fixation, tissue processing, H & E staining, single and multiplex antibodies are all factors which must be considered prior to designing a AI classifier. In addition, the attributes of digital scanners must be factored into the study design as they will also greatly impact the performance of the classifier.
During this talk we will discuss the options available for training datasets, and how to design and approach an AI research project. The speaker will recap where machine learning is being used today, how quantitative algorithms are utilized currently and hypothesise where AI will be used in the future of Digital Pathology.
Who should attend?
- Pathologists who are interesting in understanding where AI may be used in future applications
- Researchers who are considering developing AI tools
- IT departments who are considering adopting Digital Pathology today