Connect with us


New open source tool for machine learning researchers

Avatar photo



The software “empowers” developers at research institutes, hospitals and life science organisations to build their own medical imaging AI models.

Project InnerEye at Microsoft Research Cambridge has released an open-source software which will make it easier for researchers to train and deploy machine learning models for medical imaging.

The project has been running for over a decade, with a focus on developing machine learning methods for automating and analysing 3D medical images.

One of the key applications of the research is in assisting clinicians with image preparation and planning tasks for radiotherapy cancer treatment.

Currently these tasks can take over an hour, depending on the type of cancer. The research team says their work shows that machine learning can reduce this to just a few minutes.

Project InnerEye has worked in collaboration with the University of Cambridge and Cambridge University Hospitals NHS Foundation.

Dr Raj Jena, group leader in machine learning and radiomics in radiotherapy at the University of Cambridge, says: “The strongest testament to the success of the technology comes in the level of engagement with InnerEye from my busy clinical colleagues.

“For over 15 years, the promise of automated segmentation of images for radiotherapy planning has remained unfulfilled. With the InnerEye ML model we have trained on our data, we now observe consistent segmentation performance to a standard that matches our stringent clinical requirements for accuracy.”

The open source software, known as InnerEye Deep Learning Toolkit, allows for developers to build 3D and 2D medical imaging classification, segmentation or sequential models.

Medical imaging classification is an important problem for doctors as it plays a key role in disease classification. Segmentation identifies the pixels of lesions from CT or MRI images, providing important information about shape and volume.

The toolkit uses a number or templates for common scenarios, including radiotherapy segmentation, radiology segmentation, and ophthalmology classification.

Traditionally building machine learning models at this scale involve building and maintaining clusters of GPUs; a time-consuming process that most researchers would rather avoid. InnerEye says the toolkit will make it easier for researchers to distribute training across large numbers of GPUs.

Javier Alvarez-Valle, principal research manager for project InnerEye says: “We have released the InnerEye Deep Learning Toolkit as open-source software on GitHub to make this ML library and technical components available to as many people and organizations as possible.

“The toolkit can be used by researchers to build and refine their own models and apply them in many ways, including applications yet to be thought of.”

The toolkit can be used by healthcare providers and companies to develop their own machine learning products and services using Microsoft Azure, a cloud computing service created by Microsoft, and extensions such as Azure Stack Hub, which provides a way to run apps.

Dr. Jena says: “With the potential to refine and take ownership of the models ourselves through technology like Azure Stack Hub, we see a way that we can integrate machine learning technologies into our treatment pathway as a long-term solution that can grow and evolve over time.”

“We are already supporting several research teams in using the InnerEye Deep Learning Toolkit to build their own ML models.

“We invite anyone interested in medical imaging AI to join and contribute to the open-source project. We welcome all contributions, no matter how small, whether using the toolkit, filing issues and bugs, or writing and extending the toolkit in new directions.

“We look forward to the research and healthcare technology community building on this foundation to ultimately benefit patients around the world.”

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending stories