NePathTK

NePathTK (NephroPathology Toolkit) is a powerful open-source end-to-end pipeline integrated with QuPath, enabling accurate multistain and multicompartment segmentation for decision support in nephropathology and large-scale analysis of kidney biopsies.
NePathTK enables the automatic segmentation and quantification of relevant tissue structures from kidney biopsies, also in the presence of severely damaged histological traits as in the case of thrombotic microangiopathies (TMA). Furthermore, the tool can be used to recognize, at instance-level, the occurrence of TMA, hence working as an effective screening tool for this subtle and rare pathology.

---

💾 Pretrained Weights on Hugging Face

Pretrained NePathTK models for both segmentation and classification of TMA are publicly available on Hugging Face.

• TMA Segmentation Models: https://huggingface.co/NePathTK/tma-segmentation
• TMA Classification Models: https://huggingface.co/NePathTK/tma-classification

---

📖 References

If you use NePathTK in your research, please cite the following works:

🔹 NePathTK

@article{NePathTK,
  title   = {Multistain Multicompartment Automatic Segmentation in Renal Biopsies with Thrombotic Microangiopathies and other Vasculopathies},
  journal = {Computerized Medical Imaging and Graphics},
  pages   = {102658},
  year    = {2025},
  issn    = {0895-6111},
  doi     = {https://doi.org/10.1016/j.compmedimag.2025.102658},
  url     = {https://www.sciencedirect.com/science/article/pii/S0895611125001673},
  author  = {Nicola Altini and Michela Prunella and Surya V. Seshan and Savino Sciascia and Antonella Barreca and Alessandro Del Gobbo and Stefan Porubsky and Hien Van Nguyen and Claudia Delprete and Berardino Prencipe and Deján Dobi and Daan P.C. {van Doorn} and Sjoerd A.M.E.G. Timmermans and Pieter van Paassen and Vitoantonio Bevilacqua and Jan Ulrich Becker},
}

🔹 MESCnn

@article{MESCnn,
  title   = {Performance and Limitations of a Supervised Deep Learning Approach for the Histopathological Oxford Classification of Glomeruli with IgA Nephropathy},
  journal = {Computer Methods and Programs in Biomedicine},
  pages   = {107814},
  year    = {2023},
  issn    = {0169-2607},
  doi     = {https://doi.org/10.1016/j.cmpb.2023.107814},
  url     = {https://www.sciencedirect.com/science/article/pii/S0169260723004807},
  author  = {Nicola Altini and Michele Rossini and Sándor Turkevi-Nagy and Francesco Pesce and Paola Pontrelli and Berardino Prencipe and Francesco Berloco and Surya Seshan and Jean-Baptiste Gibier and Anibal Pedraza Dorado and Gloria Bueno and Licia Peruzzi and Mattia Rossi and Albino Eccher and Feifei Li and Adamantios Koumpis and Oya Beyan and Jonathan Barratt and Huy Quoc Vo and Chandra Mohan and Hien Van Nguyen and Pietro Antonio Cicalese and Angela Ernst and Loreto Gesualdo and Vitoantonio Bevilacqua and Jan Ulrich Becker}
}