Publications

When can you trust a computer’s outline of a tumour, and how does a busy clinician check it quickly? Everything below is one of three attempts at an answer.

T2 Fast and Reliable Dosimetric Contour QA, PhD thesis, University of Bern, 2025.

C conference · J journal · W workshop · T thesis, numbered oldest to newest.


Robust Segmentation Models

Which design choices survive a scan from a different hospital, and which quietly break. Watch them break.

  • J5 The impact of U-Net architecture choices and skip connections on the robustness of segmentation across texture variations
    A. Kamath, J. Willmann, N. Andratschke, M. Reyes · Computers in Biology and Medicine, 2025
    paper · code · project page
  • J4 DreamOn: a data augmentation strategy to narrow the robustness gap between expert radiologists and deep learning classifiers
    L. Lerch, L. S. Huber, A. Kamath, A. Pöllinger, A. Pahud de Mortanges, V. C. Obmann, F. Dammann, W. Senn, M. Reyes · Frontiers in Radiology, 2024
    paper
  • C7 Do we really need that skip-connection? Understanding its interplay with task complexity
    A. Kamath, J. Willmann, N. Andratschke, M. Reyes · MICCAI, 2023 · early accept, 14% · 🏆 Uni Bern BME Club Travel Award
    paper · code · talk · project page
  • W2 How do 3D image segmentation networks behave across the context versus foreground ratio trade-off?
    A. Kamath, Y. Suter, S. You, M. Mueller, J. Willmann, N. Andratschke, M. Reyes · Medical Imaging Meets NeurIPS Workshop, 2022
    paper · code · project page

Personalized Contour Review and Correction

Two experts outline the same tumour differently and both can be right. Telling genuine errors apart from reasonable disagreement.

  • J7 Predicting the impact of target volume contouring variations on the organ at risk dose: results of a qualitative survey
    J. Willmann, A. Kamath, R. Poel, E. Riggenbach, L. Mose, J. Bertholet, S. Muller, D. Schmidhalter, N. Andratschke, E. Ermiş, et al. · Radiotherapy and Oncology, 2025
    paper · project page
  • J6 Efficient review of automatic contouring of OARs in the brain: a dual-layer quality assurance approach combining geometric and dosimetric validation
    R. Poel, A. Kamath, E. Ermiş, E. Rüfenacht, N. Andratschke, P. Manser, D. M. Aebersold, M. Reyes · Radiotherapy and Oncology, 2025
    paper
  • W3 AutoDoseRank: automated dosimetry-informed segmentation ranking for radiotherapy
    Z. Mercado, A. Kamath, R. Poel, J. Willmann, E. Ermiş, E. Riggenbach, L. Mose, N. Andratschke, M. Reyes · MICCAI Workshop on Cancer Prevention through Early Detection, 2024
    paper · code · project page
  • C8 Comparing the performance of radiation oncologists versus a deep learning dose predictor to estimate dosimetric impact of segmentation variations
    A. Kamath, Z. Mercado, R. Poel, J. Willmann, E. Ermiş, E. Riggenbach, N. Andratschke, M. Reyes · MIDL, 2024 · oral, 18% acceptance
    paper · code · talk · project page
  • C6 Dose guidance for radiotherapy-oriented deep learning segmentation
    E. Rüfenacht, A. Kamath, et al. · MICCAI, 2023
    paper
  • C5 ASTRA: atomic surface transformations for radiotherapy quality assurance
    A. Kamath, R. Poel, J. Willmann, E. Ermiş, N. Andratschke, M. Reyes · IEEE EMBC, 2023 · 🏆 2nd Best Student Paper Award
    paper · code · talk · project page

Fast and Sensitive Dose Prediction

Predicting the radiation dose an outline would produce in seconds rather than hours, without losing the sensitivity that makes it worth having.

  • J2 Deep-learning-based dose predictor for glioblastoma: assessing the sensitivity and robustness for dose awareness in contouring
    R. Poel, A. Kamath, J. Willmann, N. Andratschke, E. Ermiş, D. M. Aebersold, P. Manser, M. Reyes · Cancers, 2023
    paper · project page
  • C4 How sensitive are deep learning based radiotherapy dose prediction models to variability in organs at risk segmentation?
    A. Kamath, R. Poel, J. Willmann, N. Andratschke, M. Reyes · IEEE ISBI, 2023
    paper · code · talk · project page
  • C3 Evaluating a deep learning based 3D dose prediction model for quality assurance of organ at risk contours
    A. Kamath, R. Poel, J. Willmann, N. Andratschke, M. Reyes · 55th Swiss Society for Radiobiology and Medical Physics, 2022
    paper

Other Work

Outside the three axes above. The older entries are from a previous life in diffusion MRI.

  • J3 Orchestrating explainable artificial intelligence for multimodal and longitudinal data in medical imaging
    A. Pahud de Mortanges, H. Luo, S. Z. Shu, A. Kamath, Y. Suter, M. Shelan, A. Pöllinger, M. Reyes · npj Digital Medicine, 2024
    paper
  • J1 PyRaDiSe: a Python package for DICOM-RT-based auto-segmentation pipeline construction and DICOM-RT data conversion
    E. Rüfenacht, A. Kamath, Y. Suter, R. Poel, E. Ermiş, S. Scheib, M. Reyes · Computer Methods and Programs in Biomedicine, 2023 · over 50 stars on GitHub
    paper · code
  • C2 Optimal acquisition protocol for white matter fiber orientation mapping using generalized CSA-ODF reconstruction
    A. Kamath, I. Aganj, J. Xu, E. Yacoub, K. Ugurbil, G. Sapiro, C. Lenglet · 21st Annual Meeting and Exhibition of the ISMRM, 2013
    paper
  • W1 Generalized constant solid angle ODF and optimal acquisition protocol for fiber orientation mapping
    A. Kamath, I. Aganj, J. Xu, E. Yacoub, K. Ugurbil, G. Sapiro, C. Lenglet · MICCAI Workshop on Computational Diffusion MRI, 2012
    paper
  • T1 A generalized CSA-ODF model for fiber orientation mapping
    A. Kamath · Master's thesis, University of Minnesota, 2012
    record
  • C1 A novel device to monitor mobilization of fingers during treatment for stiffness of tendons
    J. Gonda, A. Kamath, V. Prasad, J. Kamath · 5th International Conference on Industrial and Information Systems, 2010
    paper