Anatomical Models for Surgical Pre-Planning

[beltrans]

Hi everyone,

We have come across a new project that wants to develop 3D printed anatomical models based on DICOM images in order to pre-plan surgical procedures. I'm assuming this is a medical device Class I under Rule 1.

I have seen that it is actually a Class 2 in the FDA, so now I'm not so sure of the classification under the MDR 2017/745, as it means that there are some risks that I'm not considering and could upgrade Class I to other Classes.

Any insights into this?

 

[indubioush]

Can you tell me why you think this would be a medical device? I don't see how it fits the definition of medical device.

 

[mihzago]

If the device generates models based on patient specific images for planning of surgical procedures, then it could quite possibly be a Class IIa under MDR, maybe even IIb for high-risk procedures or situations where surgeons heavily rely on the images/models. If the models are not patient specific but still used for planning a specific procedure then maybe you could argue for Class I. I imagine there's software involved so this would be an active device.

If the models are generic in nature and/or used for general education or training, then probably not a medical device.

The risk, that I can see, is that if the models do not accurately represent the patient's anatomical characteristics, then it could lead to errors during surgery, which depending on type of surgery could be minor or easily avoidable or quite severe.

 

[beltrans]

The anatomical models would be patient specific and could lead to high-risk procedures or situations where surgeons heavily rely on the models, which would make much more sense as a class IIb. Nonetheless I'm still wondering what would happen in the 3D Printed Model, it would also be classified as IIb or as an accessory Class I to the software that creates the models?

 

[indubioush]

Thanks for the info.
Regarding the 3d printed model, I think you need to consult with your notified body. I would be very curious what they say. Please report back so we can all learn. This guidance doc may help as well: https://health.ec.europa.eu/system/files/2021-10/mdcg_2021-24_en_0.pdf

 

[mihzago]

Checking in with your NB as indubioush suggested is probably best idea, but my guess would be the model itself is just an output of the system that creates that model, so it's part of the device.

 

[KShaw]

This is such an interesting question!

It might boil down to the specifics of your intended purpose and intended users. E.g. is it just a simple anatomical structure or is it colour coded to show information on tumour resection margins? Will the software be used by the surgeons themselves, or are DICOM sent to a third-party company that then provides the anatomical model?

From one perspective, the software and printing processes can be collectively considered to be fabrication processes i.e. the software is not considered a medical device. At the end of the day, it is the anatomical model which is being used for pre-planning. If the anatomical model is incorrect, you can blame it on the fabrication process and this would be a big part of your risk management and validation.

From another perspective, the software is doing all the computations and the accuracy of the final model will certainly depend on the image processing. I think if it is being used by the surgeons themselves in-house, then perhaps you could also consider the software as preplanning software, but then again, depends on the specifics of intended purpose.

In these kind of situations, it is certainly a good idea to get a NB opinion if possible however, make sure they provide a good rationale for their reply.