Today, Permutable.ai is announcing the first of its ethically based initiatives using artificial intelligence.
Permutable will be working with the Royal Marsden Hospital Foundation in using its client data to further the development of automated CT image classification using deep learning neural networks. These networks may subsequently be used to develop AI tools to perform specific measurements on other routine images.
By using its own proprietary AI engine R2, Permutable will be able to leverage its research in fast-learning systems known as one or few-shot learning where an AI is able to classify or learn immediately. Permutable’s platform utilizes concepts such as data augmentation and model transfer learning within a human feedback loop system.
While deep learning systems have provided breakthroughs in several tasks in the medical domain, they are still limited by the problem of dependency on the availability of training data. To deal with this limitation, Permutable will be providing on-going research in the area of few-shot learning. Few-shot learning algorithms aim to overcome the data dependency by exploiting the information available from a very small amount of data. In medical scannings such as MRI and CT, due to the naturally rare occurrence of particular diseases, there is often a limitation on the available patient data in hospitals, therefore advancements in the application of few-shot learning algorithms can prove to be a significant advancement. Permutable will apply processes such as pre-processing, augmentation, and a human-feedback loop that it currently uses in detecting time-series patterns in financial data-sets.
“I am pleased to announce the start of Permutable’s collaboration with the Royal Marsden Foundation Trust, this is a testament of our efforts to take our deep learning technology beyond financial markets. The project will be led by Permutable’s lead engineer Alex Medvedev and machine learning architect adviser Jon Machtynger.”
– Wilson Chan, CEO Permutable Technologies
For further information, contact Talya Stone, press officer on firstname.lastname@example.org