Ask the Expert

What I'll do is I'll upload a small set of DICOMs and what will happen is as soon as the DICOMs actually hit our servers, we can run a number of checks on the actual images that are being uploaded to the system. So we can check for things like slice spacing, 

• Is the imaging data correct in terms of whether we're expecting the CT, is it a CT?

• Are there any slices missing or are there any significant artifacts within that data set? 

All of that can be done before the case actually hits the first patient set of workflow. So what I'll do is I'll upload this to the system and as soon as that data actually hits the Axial3D INSIGHT system or if this is a configured workflow in Axial3D  AXCEL, we can actually kick off a number of different workflows from that. So the data can be scrubbed of any patient identifiable information. And we can also run a series of machine learning algorithms on that. 

So what I'll do now is I'll actually show you some of the results of the machine learning algorithms. From that CT scan that's just been uploaded, I can show you the web based viewer on our backend. This web -based viewer has zero footprint. It can be available in any internet -enabled device where medical device reps or biomedical engineers are actually processing the cases, and can ensure that the data that's been uploaded, along with the automated checks, is correct for that specific case. 

Then I can show you also what the automated outputs look like from our algorithm. So as I said, as soon as that data hits our servers, typically within a small number of minutes, one or two minutes, the data scrubs and then a machine learning algorithm is actually applied to that data set. And one of the important things to note is these specific machine learning algorithms that I am running and showing here are for our Axial3D INSIGHT platform. We have done a number of projects in the past where we can create bespoke algorithms for individual requirements.