Imorphics extracts actionable, quantitative information from 3D clinical images efficiently, creating unique opportunities for deep learning from big data analytics and artificial intelligence.
While it is possible to look for patterns in large quantities of numerical or textual information, images must first be processed to extract and quantify the salient content. Doing this manually for large data sets is an impossible task. Imorphics statistical modelling technology enables fully automated, rapid analysis of images to:
- accurately extract anatomical regions necessary for deep learning algorithms; and
- provide real metadata descriptions of shape – something that cannot be elicited from manual processing.
This additional data can be used to compare shapes between patients and time points, delivering efficient input into deep learning algorithms not seen before.
Imorphics work with big data in osteoarthritis has already revealed surprising findings. By enabling the analysis of 10,000 knee images from the Osteoarthritis Initiative (OAI), Imorphics have been able to show that during the progression of osteoarthritis, knee joints undergo bone shape changes in a very similar way. The result of this finding is that Imorphics have developed a method of using bone shape changes as a very sensitive imaging biomarker of osteoarthritis progression.
Read more in Imorphics OA whitepaper or get in touch to talk about how we can help you leverage your big data.