Lightnews — Scholar-powered news

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · 9d

🎯 Authors propose a hybrid 3D CNN with cross-attention for glaucoma classification from OCT scans, designed to capture asymmetries across hemiretinas and integrate optic nerve head and macula features.
🔎 Free to read: doi.org/10.59275/j.m...

AI-CNet3D: An Anatomically-Informed Cross-Attention Network with Multi-Task Consistency Fine-tuning for 3D Glaucoma Classification

R Kenia, A Li, R Srivastava, K A Thakoor

doi.org

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · 9d

🚨 New publication alert:
📢 “AI-CNet3D: An Anatomically-Informed Cross-Attention Network with Multi-Task Consistency Fine-tuning for 3D Glaucoma Classification.”
🖊️ R Kenia, A Li, R Srivastava, K A Thakoor.
⬇️

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · 10d

🎯 Authors propose a domain-adaptive framework for brain vessel segmentation that uses image-to-image translation and disentanglement to handle varied imaging modalities without domain-specific design.
🔎 Free to read: doi.org/10.59275/j.m...

Multi-Domain Brain Vessel Segmentation Through Feature Disentanglement

F Galati, D Falcetta, R Cortese, F Prados, N Burgos, M A Zuluaga

doi.org

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · 10d

🚨 New publication alert:
📢 “Multi-Domain Brain Vessel Segmentation Through Feature Disentanglement.”
🖊️ F Galati, D Falcetta, R Cortese, F Prados, N Burgos, M A Zuluaga.
⬇️

Multi-Domain Brain Vessel Segmentation Through Feature Disentanglement

F Galati, D Falcetta, R Cortese, F Prados, N Burgos, M A Zuluaga

doi.org

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Sep 10

🎯 Authors examine sex bias in deep learning models for ECG classification and find that performance varies across conditions and model types. Even with balanced training data, disparities persist, emphasizing the need for fairness in clinical AI.
🔎 Free to read: doi.org/10.59275/j.m...

Investigating sex bias in ECG classification for Atrial Fibrillation, Sinus Rhythm and Myocardial Infarction

M Galanty, B van der Ster, A P Vlaar, C I Sánchez

doi.org

1 3

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Sep 10

🚨 New publication alert:
📢 “Investigating sex bias in ECG classification for Atrial Fibrillation, Sinus Rhythm and Myocardial Infarction.”
🖊️ M Galanty, B van der Ster, A P Vlaar, C I Sánchez.
⬇️

1 2

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Sep 5

Heartfelt thank you to @ja-schnabel.bsky.social for her contributions as Founding Executive Editor of MELBA! Julia played a central role in kickstarting the journal and shaping its mission and vision.
We look forward to working with Julia in her new role as an advisor to the MELBA journal.

1 2

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 28

🎯 Authors propose a distance map–based cell segmentation method that supports partially annotated objects, addressing limitations of fully supervised approaches. It enables effective transfer learning while maintaining segmentation quality.
🔎 Free to read: doi.org/10.59275/j.m...

Sketchpose: Learning to Segment Cells with Partial Annotations

C Cazorla, N Munier, R Morin, P Weiss

doi.org

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 28

🚨 New publication alert:
📢 “Sketchpose: Learning to Segment Cells with Partial Annotations.”
🖊️ @ccazorla31.bsky.social, N Munier, R Morin, P Weiss.
⬇️

1 1 1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 27

🎯 Authors propose a lightweight debiasing method that fine-tunes model parameters based on their contributions to bias and prediction. With minimal data and training, it improves fairness and generalization without compromising accuracy.
🔎 Free to read: doi.org/10.59275/j.m...

SWiFT: Soft-Mask Weight Fine-tuning for Bias Mitigation

J Yan, F Chen, Y Xue, Y Du, K Vilouras, S A Tsaftaris, S McDonagh

doi.org

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 27

🚨 New publication alert:
📢 “SWiFT: Soft-Mask Weight Fine-tuning for Bias Mitigation.”
🖊️ J Yan, F Chen, Y Xue, Y Du, K Vilouras, S A Tsaftaris, S McDonagh.
⬇️

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 26

🎯 Authors propose Neural CRF (NCRF), an end-to-end model for prostate MRI segmentation using learnable deep feature-based potentials. NCRF improves consistency and outperforms traditional CRFs in zonal segmentation accuracy.
🔎 Free to read: doi.org/10.59275/j.m...

A Neural Conditional Random Field Model Using Deep Features and Learnable Functions for End-to-End MRI Prostate Zonal Segmentation

A L Y Hung, K Zhao, K Pang, H Zheng, X Du, Q Miao, D Terzopoulos, K Sung

doi.org

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 26

🚨 New publication alert:
📢 “A Neural Conditional Random Field Model Using Deep Features and Learnable Functions for End-to-End MRI Prostate Zonal Segmentation.”
🖊️ A L Y Hung, K Zhao, K Pang, H Zheng, X Du, Q Miao, D Terzopoulos, K Sung.
⬇️

A Neural Conditional Random Field Model Using Deep Features and Learnable Functions for End-to-End MRI Prostate Zonal Segmentation

A L Y Hung, K Zhao, K Pang, H Zheng, X Du, Q Miao, D Terzopoulos, K Sung

doi.org

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 25

🎯 Authors investigate transfer learning to improve the robustness of deformable image registration under unforeseen domain shifts, by pretraining on synthetic data and fine-tuning on target domains.
🔎 Free to read: doi.org/10.59275/j.m...

Robust deformable image registration using synthetic data and transfer learning

I D Kolenbrander, M Maspero, J P Pluim

doi.org

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 25

🚨 New publication alert:
📢 “Robust deformable image registration using synthetic data and transfer learning.”
🖊️ I D Kolenbrander, M Maspero, J PW Pluim.
⬇️

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 22

🎯 Authors propose a synthetic data framework for stroke lesion segmentation, extending SynthSeg with lesion-specific augmentations, and achieve strong out-of-domain generalization.
🔎 Free to read: doi.org/10.59275/j.m...

Synthetic Data for Robust Stroke Segmentation

L Chalcroft, I Pappas, C J Price, J Ashburner

doi.org

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Aug 22

🚨 New publication alert:
📢 “Synthetic Data for Robust Stroke Segmentation.”
🖊️ L Chalcroft, I Pappas, C J Price, J Ashburner
⬇️

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Jul 14

🎯 Authors present the first BraTS challenge on pediatric brain tumor segmentation using multi-institutional data, aimed at supporting clinical trial workflows and fostering collaboration in pediatric neuro-oncology.
🔎 Free to read: doi.org/10.59275/j.m...

BraTS-PEDs: Results of the Multi-Consortium International Pediatric Brain Tumor Segmentation Challenge 2023

A Fathi Kazerooni, N Khalili, X Liu, D Haldar, Z Jiang, A Zapaishchykova, J Pavaine, L M Shah, B V Jones, N Sheth, S P Prabhu, A S McAllister, W Tu, K K Nandolia, A F Rodriguez, I S Shaikh, M Sanchez-...

doi.org

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Jul 14

🚨 New publication alert:
📢 “BraTS-PEDs: Results of the Multi-Consortium International Pediatric Brain Tumor Segmentation Challenge 2023.”
🖊️ A F Kazerooni, N Khalili, X Liu, D Haldar,..., A Resnick, @spyridonbakas.bsky.social, A Vossough, M G Linguraru.
⬇️

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Jul 11

As a reminder, MELBA and MIDL have a formal strategic partnership, enabling journal-to-conference and conference-to-journal tracks that support the seamless dissemination of high-quality research.
We’re proud to contribute to this collaborative and open-access ecosystem!

1 1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Jul 11

🎙️ MELBA at MIDL 2025 🎙️

MELBA had the pleasure of hosting the final session of Day 2 at MIDL 2025, featuring two outstanding oral presentations.

This session highlighted the strong synergy between MELBA and MIDL. ⬇️

1 1 1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Jun 12

🎯 Authors show that transformer design choices from natural images often hurt performance in medical imaging. On mammography and chest CT, simpler architectures outperform complex ones, highlighting the value of simplicity.
🔎 Free to read: doi.org/10.59275/j.m...

Understanding differences in applying DETR to natural and medical images

Y Xu, Y Shen, C Fernandez-Granda, L Heacock, K J Geras

doi.org

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Jun 12

🚨 New publication alert:
📢 “Understanding differences in applying DETR to natural and medical images.”
🖊️ Y Xu, Y Shen, C Fernandez-Granda, L Heacock, K J Geras.
⬇️

1

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Jun 11

🎯 Authors introduce BrainMorph, a keypoint-based foundation model for multi-modal brain MRI registration, trained on 100,000+ 3D scans and enabling robust, interpretable, and scalable alignment in challenging scenarios.
🔎 Free to read: doi.org/10.59275/j.m...

BrainMorph: A Foundational Keypoint Model for Robust and Flexible Brain MRI Registration

A Q Wang, R Saluja, H Kim, X He, A Dalca, M R Sabuncu

doi.org

2 2

Machine Learning for Biomedical Imaging @melbajournal.bsky.social · Jun 11

🚨 New publication alert:
📢 “BrainMorph: A Foundational Keypoint Model for Robust and Flexible Brain MRI Registration.”
🖊️ A Q Wang, R Saluja, H Kim, X He, @adriandalca.bsky.social , @msabuncu.bsky.social.
⬇️

1 3 4