🏥 MIP-AIM? MIP-IM? Whatever. UNet Zoo 🦄

(The only thing reliably updated in this repo)

Welcome to my glorious dumpster fire of PhD code! 🔥🗑️ This repository contains the chaotic scripts, models, and experiments that powered my PhD research—focused on semantic segmentation (because playing "connect the pixels" counts as science).

⚠️ Warning: This repo is ZERO-ORGANIZED (yes, in bold again for emotional impact). It's a raw, barely-commented archive of years of work. I might organize it between the heat death of the universe and my next LinkedIn post. Until then, bring a compass.

Here you'll find:

• Questionably-thawed models 🧊
• Half-digested experiments 🤢
• That one script that might be important ✨

🔍 What’s Inside? (Besides Mild Regret?)

� Model Zoo (Mostly Unleashed)

• unet: Classic U-Net (the "Hello World" of segmentation, but with extra steps).

• inception_unet: U-Net after three espresso shots – now with Inception modules (we go deeper).

• mipaim_unet: Medical Image Parcellation + Attention + Inception – because academia rewards complexity.

📦 Other Artifacts

• 📊 Scripts labeled final_try (they lied).

• 🧨 Jupyter notebooks that may or may not explode.

• 📝 Comments like # magic number – don't touch (absolute lies).

🚀 How to Navigate This Chaos?

1. Assume nothing works – Start from this mindset to avoid disappointment.

2. Worship Ctrl + F – Your only ally in this labyrinth.

3. Sacrifice a coffee ☕ to the debugger gods before running anything.

📜 How to Cite This Work

If you (against all odds) find something useful, here’s how to credit me without guilt:

• Cabeza-Ruiz, R., Velázquez-Pérez, L., Linares-Barranco, A., & Pérez-Rodríguez, R. (2022). Convolutional Neural Networks for Segmenting Cerebellar Fissures from Magnetic Resonance Imaging. Sensors, 22(4), 1345.

• Cabeza-Ruiz, R., Velázquez-Pérez, L., González-Dalmau, E., Linares-Barranco, A., & Pérez-Rodríguez, R. (2025). Deep Learning-Based Assessment of Brainstem Volume Changes in Spinocerebellar Ataxia Type 2 (SCA2): A Study on Patients and Preclinical Subjects. Sensors, 25(19), 6009.

📄 Related Papers (For Extra Credit)

• Cabeza-Ruiz, R., Velázquez-Pérez, L., Pérez-Rodríguez, R. (2021). Convolutional Neural Networks as Support Tools for Spinocerebellar Ataxia Detection from Magnetic Resonances. In: Hernández Heredia, Y., Milián Núñez, V., Ruiz Shulcloper, J. (eds) Progress in Artificial Intelligence and Pattern Recognition. IWAIPR 2021. Lecture Notes in Computer Science(), vol 13055. Springer, Cham.

• Cabeza-Ruiz, R., Velázquez-Pérez, L., Pérez-Rodríguez, R. et al. ConvNets for automatic detection of polyglutamine SCAs from brain MRIs: state of the art applications. Med Biol Eng Comput 61, 1–24 (2023).

• Cabeza-Ruiz, R., Velázquez-Pérez, L., Linares-Barranco, A., & Pérez-Rodríguez, R. (2024, October). Automated Thalamic Nuclei Segmentation from Brain T1-W MRI Using Convolutional Neural Networks. In Workshop in R&D+ i & International Workshop on STEM of EPS (pp. 422-431). Cham: Springer Nature Switzerland.

🙏 Inspiration Credits

• Han, S., Carass, A., He, Y., & Prince, J. L. (2020). Automatic cerebellum anatomical parcellation using U-Net with locally constrained optimization. Neuroimage, 218, 116819.

• Kamnitsas, K., Ferrante, E., Parisot, S., Ledig, C., Nori, A. V., Criminisi, A., ... & Glocker, B. (2016). DeepMedic for brain tumor segmentation. In Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries: Second International Workshop, BrainLes 2016, with the Challenges on BRATS, ISLES and mTOP 2016, Held in Conjunction with MICCAI 2016, Athens, Greece, October 17, 2016, Revised Selected Papers 2 (pp. 138-149). Springer International Publishing.

• Woo, S., Park, J., Lee, J. Y., & Kweon, I. S. (2018). Cbam: Convolutional block attention module. In Proceedings of the European conference on computer vision (ECCV) (pp. 3-19).

• Nishimaki, K., Onda, K., Ikuta, K., Chotiyanonta, J., Uchida, Y., Mori, S., ... & Australian Imaging Biomarkers and Lifestyle Flagship Study of Ageing. (2024). OpenMAP‐T1: A Rapid Deep‐Learning Approach to Parcellate 280 Anatomical Regions to Cover the Whole Brain (Vol. 45, No. 16, p. e70063). Hoboken, USA: John Wiley & Sons, Inc.

🍻 Acknowledgments

• Reviewers who pretended my code was readable – You’re the real heroes. 🦸

• DeepSeek – For writing this hilarious README when I was too tired to be funny.

• GitHub’s "Blame" feature – For confirming that yes, I wrote that mess.

• You, brave archaeologist – For digging through this digital midden.

🌟 Good luck! (You’ll need it.)