Great Invesalius! Some suggestions and questions

[mauigna06]

You have developed a great software. Here are some suggestions:

^It'd be great if the masks could be imported|exported between projects that share the same DICOM data

^ It'd be great if the masks would be named in a coded way so that you can remember what you did to achieve them... Here is my naming protocol:

 [MaskId*]-[PartMicroname]-[MethodNº1_par*1_..._parn]-...-[MethodNºn_par1_..._parn]-[TimeItTookMinutes]

 For example for a calcaneus bone segmentation from a CT Scan I chose C1 as Id, Cal as microname, the segmentation was made with Multiple Region Growing in 2D with Threshold with (225,2000), then I did Multiple Region Growing in 2D with confidence parameters 1.0, 0.5 and sinus WWWL, then I made Multiple Fill Holes and after that Multiple Manual Edition and it took 40 minutes. So the name of the layer should be: C1-Cal-MRG2_Thr_225,2000-MRG2_Con_1.0,5,sinus-MFH-MMa-40

 *This is for when you do boolean operations, crop or select part operations so the source masks can be identified properly
 *par means parameter

This is my table to name the methods, you can copy-past it in an Excel

Nombre Metodos Nombre codigo Threshold Thr WaterSheed Ws Region Growing Rg Manual Ma Select Part CAPA SP_CAPA Remove Part CAPA RP_CAPA Fill Holes Automatically Multi2D FH2 Fill Holes Automatically 3D FH3 Crop CAPA Cr_CAPA

Also I make an initial project where I only load the dicom files and delete any mask. I save it with a naming code and then all projects related to that dicom files share the name prefix of the initial one. Something alike to what I explained with the mask could be done for surfaces.

^It'd be good to have a marker on the surface where the plains intersect. ^It'd be good if you could make your software so that it could read files from slicer and vice versa so the advantages of both of them could be taken into account. ^It'd be good if surface smoothing could be made after binary surface generation. So one is able to see the modifications made to the surface

Finally I have some questions: ^With what kind of materials do you print bones for surgical planning so that they can be not only looked at but cut with a saw? What kind of printing method (FDM, SLA, SLS) do you suggest and why? What kind of printers in the market do you see affordable and reliable to a hospital that wants to use 3D Printing? ^Have you ever used some assistive hardware to speed the segmentation such as a digitalizing tablet or 3d mouse? If yes, were they useful? ^How long does it take you to make segmentations?

Thank you for your development. Mauro

[tfmoraes]

Hi @mauigna06,

Thanks for the suggestions!

^ It'd be great if the masks could be imported|exported between projects that share the same DICOM data

It's a good suggestion. I'll add it to the to do list.

^ It'd be great if the masks would be named in a coded way so that you can remember what you did to achieve them... Here is my naming protocol

It's a interesting idea. But what if the user give a mask name manually? Also, I think it'll generate a very long name that will complicate the user usability. I don't know...

^It'd be good to have a marker on the surface where the plains intersect.

If I understood you correctly, that's what the Cross tool do. See red sphere in the image bellow:

^It'd be good if you could make your software so that it could read files from slicer and vice versa so the advantages of both of them could be taken into account.

We need to see and understand the file format used by Slicer. You are the person to suggest it. Do you know if Slicer is much used?

^It'd be good if surface smoothing could be made after binary surface generation. So one is able to see the modifications made to the surface

That's something we want to do: give user the possibility of applying smooth and decimation to a generated surface.

^With what kind of materials do you print bones for surgical planning so that they can be not only looked at but cut with a saw? What kind of printing method (FDM, SLA, SLS) do you suggest and why? What kind of printers in the market do you see affordable and reliable to a hospital that wants to use 3D Printing?

We use Polyamide in a SLS printer. We also used a 3D printer from ZCorp with gypsum material (gesso in Portuguese). I don't know what printer to suggest you, but I think an SLS printer is good.

^Have you ever used some assistive hardware to speed the segmentation such as a digitalizing tablet or 3d mouse? If yes, were they useful?

Yes, a Wacon tablet. I don't know if it really helps because the person which tested InVesalius with the wacom tablet prefers to use the mouse.

^How long does it take you to make segmentations?

It depends on the images. Some times a simple threshold is all you need, so it take just a few minutes. If watershed or region growing work perfectly, it'll also take just a few minutes. The problem is if you need to a manual segmentation. It may take 10 minute to some hours.

[mauigna06]

We need to see and understand the file format used by Slicer. You are the person to suggest it. Do you know if Slicer is much used?

I don't have statistics so I don't know. I like your interface more, but they have some other algorithms for segmentation that I'd like you to have. Although thinking it well, if you'd be considering to add just another algorithm for segmentation I would suggest the ITKSnap snakes algorithm. It is a good algorithm but I prefer your user interface.

We use Polyamide in a SLS printer. We also used a 3D printer from ZCorp with gypsum material (gesso in Portuguese). I don't know what printer to suggest you, but I think an SLS printer is good.

So you never used an SLA or FDM printer to make some medical models?

[tfmoraes]

I don't have statistics so I don't know. I like your interface more, but they have some other algorithms for segmentation that I'd like you to have. Although thinking it well, if you'd be considering to add just another algorithm for segmentation I would suggest the ITKSnap snakes algorithm. It is a good algorithm but I prefer your user interface.

ITKSnap is a great software, that snakes algorithm is very good. We want to implement this algorithm in InVesalius or something similar. Do you use the ITKSnap to segment what type the images and regions?

So you never used an SLA or FDM printer to make some medical models?

We don't have a SLA printer here. We've already used FDM printer to print some medical models.

[mauigna06]

ITKSnap is a great software, that snakes algorithm is very good. We want to implement this algorithm in InVesalius or something similar. Do you use the ITKSnap to segment what type the images and regions?

I mainly segment bones and I have just used ITKSnap a few times but I think It has great potential to segment long bones (tibia, peroneus, humerus, femur) fast and well.

We've already used FDM printer to print some medical models.

1) Could you be more specific about what kind of models you have printed (bones, vessels, airways, tumors, etc)?

2) What was the main application of the printed model?

[tfmoraes]

1. Could you be more specific about what kind of models you have printed (bones, vessels, airways, tumors, etc)?

Mainly bones, tumors and prosthesis. Not only to human applications, but also to veterinary, archeology and engineering applications.

2. What was the main application of the printed model?

Mainly to Surgical planning. But we also print molds to prosthesis, since the material used in the printers are not bio-compatible. We have 3D printer which prints with titanium, which is bio-compatible. But the prosthesis printed by this printer are only used experimentally.

[mauigna06]

Mainly to Surgical planning. But we also print molds to prosthesis, since the material used in the printers are not bio-compatible. We have 3D printer which prints with titanium, which is bio-compatible. But the prosthesis printed by this printer are only used experimentally.

Thank you for your answer. Just a few more questions:

1 What is the material you use to make the molds?

2 What is the material you use to put in the molds to make prosthesis? How is the method of production called so we are able to research more about it?

[tfmoraes]

1 What is the material you use to make the molds?

Polyamide (Nylon).

2 What is the material you use to put in the molds to make prosthesis? How is the method of production called so we are able to research more about it?

The material is chosen by physician. They use bone cement or silicone. We use a tool of Magics called Tooling to make the molds.