GSOC 2026 Introduction

[siddharthsiva]

Dear Paulo and InVesalius Mentors,

My name is Siddharth, and I’m a Computer Science student at UC San Diego. I am writing to express my interest in contributing to InVesalius for GSoC 2026.

I have a strong background in high-performance computing and medical-adjacent software. During my recent internship at a bioinformatics startup, I architected a platform that reduced genomic analysis time by 80% using OpenCV pipelines and Python REST APIs. Additionally, I lead the architecture for a C++20 columnar database that utilizes SIMD and CUDA to achieve extremely high data throughput.

I am particularly interested in the intersection of 3D reconstruction and performance optimization. I have already begun setting up the InVesalius development environment and am preparing the recorded demo and code sample as required by the guidelines.

I'm currently looking through the project ideas to see where my experience in hardware-level optimization (SIMD/CUDA) or computer vision might be most useful. I'll be sharing my draft proposal folder via Google Drive shortly.

I look forward to discussing how I can contribute to the project this summer!

Best regards, Siddharth

[paulojamorim]

Hi @siddharthsiva

Thank you for your interest.

We have the following topics where your experience could be useful:

• Optimization of InVesalius for micro-CT and synchrotron micro-CT

These imaging modalities generate very large datasets. As an example, we have a dataset of approximately 29 GB, with dimensions of 2711 × 3424 × 3200 voxels. Even on a workstation equipped with a Xeon processor, 128 GB of RAM, and an NVIDIA A4000 GPU, InVesalius can operate only at half of the original resolution. Even so, the process of generating the triangular mesh, correcting normals, and applying smoothing (smooth) takes approximately 30 minutes.

Previously, this processing required several hours, but there was a significant improvement after the optimizations in the context-aware smoothing algorithm and the Rust implementation carried out by @tfmoraes.

Slice visualization and volume rendering (raycasting) are also relatively slow in these cases.

Other ideas include:

• Extract surface using Dual contouring See

• Cinematic Volume Rendering (raycasting)
  The volume rendering in InVesalius performs well, but there is still room for improvement in terms of realism. See this paper

[himanshumahajan138]

Dear Paulo and InVesalius Mentors,

My name is Himanshu Mahajan, and I am currently working as a Data Scientist where I regularly build data-driven systems using Python. I have over three years of practical experience with Python and backend development, and I actively enjoy working on open-source projects and exploring real-world problem solving through code. I am writing to express my interest in contributing to InVesalius as part of Google Summer of Code (GSoC) 2026.

I am particularly interested in the issue “Saving timestamp for each created marker.” Adding automatic timestamp storage when markers are created would improve data tracking and enable better synchronization with external systems such as EEG, EMG, or TMS. With my experience in Python and handling structured data systems, I believe I can implement this feature while ensuring compatibility with existing marker files.

I have started exploring the InVesalius codebase to understand the current marker creation and export workflow and would be excited to contribute this feature with proper documentation.

Thank you for your time and consideration.

Best regards,
Himanshu Mahajan

[adityakrmishra]

Dear @paulojamorim, Thiago, and the InVesalius Mentors,

My name is Aditya Kumar Mishra, and I am writing to express my strong interest in contributing to InVesalius for GSoC 2026.

I am a 16-year-old Class 12 student. While I am young and just meeting the official age requirement to participate in GSoC, I am highly capable, deeply passionate about software development, and eager to gain exposure by contributing to a meaningful open-source project like yours.

I already have experience navigating open-source workflows, having successfully contributed to the GoFr framework (with merged pull requests), and I have a strong foundation in technical problem-solving, having recently won a state-level first prize in an ISEA technical paper writing competition.

I have successfully set up the InVesalius development environment locally on Windows, navigating the Rust/Cargo dependencies and the Python 3.12 virtual environment setup. I am currently exploring the codebase and GUI. As a warm-up, I am preparing a quick pull request to resolve some Python 3.12 SyntaxWarning (invalid escape sequence) logs I noticed during my local build.

Moving forward, I am carefully reviewing the GSoC idea list and am particularly drawn to projects involving Python data handling (such as the marker timestamp feature) and the Python/Rust integrations for the .3MF format export.

I am ready to put in the hard work and learn from this amazing community. You can find more about my background on my LinkedIn profile: [Insert your LinkedIn Link here]

I look forward to interacting with you all and making my first contributions soon!

Best regards,
Aditya Kumar Mishra