Digitally prepping a Pliocene gannet skull using machine learning

[Doctor Mud]

Hello everyone,

 

I have quite a few projects going on and I'm going to be a bit more active on here to share things.

Sorry I've been a bit quiet!

 

I see there are a few threads on here about CT scanning, 3D printing and segmentation, but I thought I would add this one. I've been doing it for a few years, but just bit the bullet and bought a machine custom built to do this.

 

I really think we should create a sub section of fossil preparation for CT scanning and segmentation? Resources are scattered around the internet and it would be great to document them all on here.

I have lots to learn and I'd like to "upload" this knowledge here as I aquire it. E.g. Hardware set up, software, things like digitization tips and tricks (stylus pen and tablet vs mouse). The same sort of thing we have for physical prep on here: set-up, tools and techniques.

 

Anyway. I just acquired a very powerful computer and I fired it up last night. I have a CT scan of a Pliocene gannet skull that was found last year that I've been sitting on. At least I think it is gannet.

I am using imageJ to do the segmentation. Segmentation is just the process of telling the computer what is bone, what is rock and what is air.

 

 

I don't have a photo of the concretion, but here is a 3D model above. It is about 10cm long. This is looking from above.

 

 

Here is one of 760 slices from the CT scan of the skull. A vertical slice with the top of the skull at the top of the image.  Pixel size is about 30 microns!

 

 

Here is the view after ONE round of training the computer. I selected some areas of bone, some of rock and some of air.

Then the computer thought about it, using 160 GB of ram (out of my total 192 GB) and the latest Intel chipset in a water-cooled CPU to classify every pixel as bone (red), rock (green) or air (purple). It does this for every of the 760 slices.

This is a first pass. You can go back and train the computer further and correct it. It gets better with each round.

 

 

Here is the first reconstruction of the skull. You can see there is still a bit of noise. I could get rid of that with a few more learning phases. A lot of loose pixels could be removed in rendering software such as Blender too.

 

Hope you enjoyed this. 

I'll keep you posted as I improve the model. And I'd like to 3D print it at the end! 

 

 

 

Edited March 29 by Doctor Mud

[Doctor Mud]

 

See if this works. This is just a threshold adjust that strips away the rock. Not as good as proper segmentation, but looks cool!

[jpc]

you,  my friend, are having way too much fun.  I say this as another fan of CT scanning of fossils.   

By this: "bought a machine custom built to do this." you are referring to the computer you bought, right.  Or did you buy a CT scanner for the living room?

[SPrice]

Reminds me of the first movie in the franchise - Jurassic Pak -  back in '93 where they're on site and the "new Tech" guys fire a 12 gauge shot shell in a seismic wave radar machine to get an underground scan of a velociraptor skeleton. 

 

 

Tech says, "  Two more years' development and we won't even have to dig anymore." 

 

Dr. Allan Grant nails it with his response... " What's the fun in that?" 

 

Me?...the new tech is pretty cool. 3D printing resin fossil reproductions is at an amazing level right now.

 

Once the above scan is as good as it gets then a 3D resin print of the gannet skull will be the bee's knees. 

 

 

[FossilDAWG]

I'm pretty sure one could never physically prep a specimen like that to get the level of detail you are getting with your CT scan.  I think in the future it will become standard practice to leave super delicate scientifically important specimens in the rock and rely on scans to visualize the fossil.  At least that would be completely repeatable if someone needed to reexamine the specimen.  Once they are physically prepped, whatever has been removed is gone for good.

 

Don

[Doctor Mud]

5 hours ago, jpc said:

you,  my friend, are having way too much fun.  I say this as another fan of CT scanning of fossils.   

By this: "bought a machine custom built to do this." you are referring to the computer you bought, right.  Or did you buy a CT scanner for the living room?

Let’s just say it’s addictive and I’ve had some late nights/early mornings!

 

I wish I could afford a CT scanner at home,  I meant a PC workstation with enough power and memory that means it can handle decent sized files without taking ages or crashing. 
before my laptop could handle 200MB in the machine learning phase. It’s really RAM hungry.

The file for the skull is 1.5 GB. I think I’ll be able to do files up to about 3 GB on this PC. 

[Doctor Mud]

5 hours ago, SPrice said:

Reminds me of the first movie in the franchise - Jurassic Pak -  back in '93 where they're on site and the "new Tech" guys fire a 12 gauge shot shell in a seismic wave radar machine to get an underground scan of a velociraptor skeleton. 

 

 

Tech says, "  Two more years' development and we won't even have to dig anymore." 

 

Dr. Allan Grant nails it with his response... " What's the fun in that?" 

 

Me?...the new tech is pretty cool. 3D printing resin fossil reproductions is at an amazing level right now.

 

Once the above scan is as good as it gets then a 3D resin print of the gannet skull will be the bee's knees. 

 

 

I remember that scene! At the time as a kid I felt it would take away the fun and mystery. Like technology shining a bright light onto those dark mysterious corners of our world. Like we should have to work hard to unlock some secrets. 
 

But like @FossilDAWG says this technology can help us go where we couldn’t before. These concretions aren’t acid soluble and scribing or air blasting would give is such detail. I suppose air scribes were new technology at one point.

 

im still like a kid at Christmas time whenever I get scan data like this 

[Doctor Mud]

4 hours ago, FossilDAWG said:

I'm pretty sure one could never physically prep a specimen like that to get the level of detail you are getting with your CT scan.  I think in the future it will become standard practice to leave super delicate scientifically important specimens in the rock and rely on scans to visualize the fossil.  At least that would be completely repeatable if someone needed to reexamine the specimen.  Once they are physically prepped, whatever has been removed is gone for good.

 

Don

Yes - these concretions aren’t acid soluble . So prepping out such thin bone is impossible in this hard rock. 
There is some interpretation involved with the CT scan processing and some people I’ve talked to have been reluctant to publish on scans alone. But when you look at the data we can get you can see bone histology so you would think you can tell when a feature is truly absent or has been eroded off. 
 

That’s a great point you make about reproducibility - one of the keystones of the scientific method. Sometimes manual tweaking is required with CT data processing. But as long as it’s recorded, the intact fossil is still there for someone else to repeat or improve upon it. 

[Sjfriend]

Very cool use of tech! Great way to expose those buried, fragile finds.