Hello! I am Mrudul.

[Mr-MVP]

Hello Everyone! I am Mrudul Pawar, a second-year student at Veermata Jijabai Technological Institute(VJTI) pursuing a degree in Computer Engineering.

I've been part of a project named 3D Reconstruction from a Single RGB Image, during which I had extensive exposure to ML to build and train a model to convert a single image of an object into a 3D mesh using various approaches and algorithms.

Also, I've created a Medicinal Plant Recognition system that can recognize over 100 plant species as a project for my participation in the Smart India Hackathon. I'm enthusiastic about contributing to GSoC to challenge and enhance my AI/ML skills and contribute to the open-source community!

I've gone through the repository and would love to get tasks related to projects.

• Mrudul Pawar

[paulojamorim]

Hi @Mr-MVP

Thank you for your interest in contributing to InVesalius!

Today InVesalius has AI-based segmentation for brain (MRI) and mandible (CT).

It would be interesting to have an AI-based tool for other parties. MedSAM seems like an interesting technique to be integrated into InVesalius.

This could be a good starting point to present your skills.

[Mr-MVP]

Thanks a lot for the opportunity! Were the MRI and CT integrated by a separate repo if so may I get their links to get a better idea of how to integrate the tool.

[paulojamorim]

They are in the main repository. The part that does this is this.

You will need to draw a bounding box, InVesalius already has this for the mask cropping tool. You can use it.

[Mr-MVP]

I've been working on it . Just needed to clarify that should I integrate MedSAM alone or MedLSAM completely ?

[paulojamorim]

Hi @Mr-MVP

Very good!

It would be nice to integrate MedLSAM (Localize and segment). You can use the pre-trained models available in the MedLSAM repository. The idea is for the user to draw the rectangle around the organ they want to segment and obtain the segmented mask.

Edit: Regarding the rectangle, actually, I meant it to be a box (3D), hence the MedLSAM.