So decided to see how fast I could go from a patient CT to a 3D printed model with high fidelity. Well we went from CT acquisition to printed in 8 hours! The quality is amazing, with extremely high detail. This is a CT of the tibia/fibula with a tiny bit of ankle thrown in. The workflow was a bit complex in that PACS systems (radiology image systems) do 3D reconstruction for visual use (who cares if the surface is manifold, etc) and so a significant repair in Emendo repaired thousands of intersecting triangles, etc. So step 1 in Osirix was a 3D surface reconstruction using bone windows, then export as polymesh. Then into meshmixer to do a plane cut to remove the bits of foot (they aren't well imaged in this CT) and make the model have a flat bottom. Then into Emendo, and finally S3D. Printed in PLA/PHA with 0.15mm layers at 200C with a ratio of 0.9; a few supports were used on the malleolus to support the tip. The surface is even smoother than it appears here. Note the hole you see is anatomic (that's where the blood vessels come through) and this is printed actual size (it's a $1m 3D scanner after all). Images were from a 64-detector GE Helical CT-scanner. The marrow cavity in the middle of the bone is fully visualized
Nice work, Henry. How were those 8 hours divided for the different steps, acquisition and processing vs. pure build time? The talus looks a bit atypically shaped. Fracture or artifact? Somewhen in the nineties, we made a mandible model (fracture, to be operated the next morning) in six and half hours, if I remember correctly from data in (the hospital was more or less next door) to model out, milled polyurethane hard foam. About one hour of data processing, the rest chopping off the block what did not look like a broken mandible... Cheers, Ulli
That scanner does not look like an x-ray CT (computed tomography) scanner. You can't look into a patient's body with that.
