From Gradient Clipping to Structural Refinement: Improving DPSGD for Medical Image Segmentation

• URL: http://arxiv.org/abs/2606.21763v1
• Date: Fri, 19 Jun 2026 21:26:37 GMT
• Title: From Gradient Clipping to Structural Refinement: Improving DPSGD for Medical Image Segmentation
• Abstract summary: Medical image segmentation is widely used for disease detection but relies on sensitive data, raising privacy concerns.<n> Differential privacy, typically implemented via Differential Private Gradient Descent (DPSGD), provides a solution.<n>Recent DPSGD variants, including Automatic clipping (Auto-S), Normalised SGD with perturbation (NSGD), and Per-sample adaptive clipping (PSAC), have shown promise in image classification.
• Score: 5.416422064108777
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Medical image segmentation is widely used for disease detection but relies on sensitive data, raising privacy concerns as trained models can leak information. Differential privacy, typically implemented via Differential Private Stochastic Gradient Descent (DPSGD), provides a solution, though at the cost of reduced utility. Recent DPSGD variants, including Automatic clipping (Auto-S), Normalised SGD with perturbation (NSGD), and Per-sample adaptive clipping (PSAC), have shown promise in image classification, but their behavior in medical segmentation remains underexplored. We evaluate these methods across binary and multi-class tasks and analyze gradient alignment, showing that prior assumptions, particularly for PSAC, do not consistently hold. We further demonstrate that combining clipping strategies with morphological refinement improves segmentation quality under privacy constraints. Finally, we propose an adaptive DP-Morph variant that captures class-specific structures and enhances performance in multi-class settings.

Related papers

• Revisiting Privacy Amplification by Subsampling in Selective Release DPSGD [4.723707756235521]
  We propose Differentially Private Selective Release based on Clipped Gradients (DPSR-CG)<n>We demonstrate that our DPSR-CG mechanism maintains strict privacy guarantees while achieving exceptional model performance.
  arXiv  Detail & Related papers  (2026-06-03T03:00:26Z)
• SPEGC: Continual Test-Time Adaptation via Semantic-Prompt-Enhanced Graph Clustering for Medical Image Segmentation [12.0982298854338]
  Continual Test-Time Adaptation (CTTA) aims to enable pre-trained models to adapt to continuously changing unlabeled domains.<n>We propose a CTTA via Semantic-Prompt-Enhanced Graph Clustering (SPEGC) for medical image segmentation.
  arXiv  Detail & Related papers  (2026-03-12T03:22:43Z)
• Differential privacy representation geometry for medical image analysis [1.828619151598757]
  Differential privacy (DP)'s effect in medical imaging is typically evaluated only through end-to-end performance.<n>We introduce Differential Privacy Representation Geometry for Medical Imaging (DP-RGMI), a framework that interprets DP as a structured transformation of representation space.<n>We show that DP is consistently associated with a utilization gap even when linear separability is largely preserved.
  arXiv  Detail & Related papers  (2026-03-01T13:30:36Z)
• Contrastive Graph Modeling for Cross-Domain Few-Shot Medical Image Segmentation [58.41482540044918]
  Cross-domain few-shot medical image segmentation (CD-FSMIS) offers a promising and data-efficient solution for medical applications.<n>We present Contrastive Graph Modeling (C-Graph), a framework that leverages the structural consistency of medical images as a reliable domain-transferable prior.
  arXiv  Detail & Related papers  (2025-12-25T14:00:17Z)
• PathSegDiff: Pathology Segmentation using Diffusion model representations [63.20694440934692]
  We propose PathSegDiff, a novel approach for histopathology image segmentation that leverages Latent Diffusion Models (LDMs) as pre-trained featured extractors.<n>Our method utilizes a pathology-specific LDM, guided by a self-supervised encoder, to extract rich semantic information from H&E stained histopathology images.<n>Our experiments demonstrate significant improvements over traditional methods on the BCSS and GlaS datasets.
  arXiv  Detail & Related papers  (2025-04-09T14:58:21Z)
• On the Convergence of DP-SGD with Adaptive Clipping [56.24689348875711]
  Gradient Descent with gradient clipping is a powerful technique for enabling differentially private optimization.<n>This paper provides the first comprehensive convergence analysis of SGD with quantile clipping (QC-SGD)<n>We show how QC-SGD suffers from a bias problem similar to constant-threshold clipped SGD but can be mitigated through a carefully designed quantile and step size schedule.
  arXiv  Detail & Related papers  (2024-12-27T20:29:47Z)
• Enhancing DP-SGD through Non-monotonous Adaptive Scaling Gradient Weight [15.139854970044075]
  We introduce Differentially Private Per-sample Adaptive Scaling Clipping (DP-PSASC) This approach replaces traditional clipping with non-monotonous adaptive gradient scaling. Our theoretical and empirical analyses confirm that DP-PSASC preserves gradient privacy and delivers superior performance across diverse datasets.
  arXiv  Detail & Related papers  (2024-11-05T12:47:30Z)
• Dual-scale Enhanced and Cross-generative Consistency Learning for Semi-supervised Medical Image Segmentation [49.57907601086494]
  Medical image segmentation plays a crucial role in computer-aided diagnosis. We propose a novel Dual-scale Enhanced and Cross-generative consistency learning framework for semi-supervised medical image (DEC-Seg)
  arXiv  Detail & Related papers  (2023-12-26T12:56:31Z)
• Differentially Private SGD Without Clipping Bias: An Error-Feedback Approach [62.000948039914135]
  Using Differentially Private Gradient Descent with Gradient Clipping (DPSGD-GC) to ensure Differential Privacy (DP) comes at the cost of model performance degradation. We propose a new error-feedback (EF) DP algorithm as an alternative to DPSGD-GC. We establish an algorithm-specific DP analysis for our proposed algorithm, providing privacy guarantees based on R'enyi DP.
  arXiv  Detail & Related papers  (2023-11-24T17:56:44Z)
• PCA: Semi-supervised Segmentation with Patch Confidence Adversarial Training [52.895952593202054]
  We propose a new semi-supervised adversarial method called Patch Confidence Adrial Training (PCA) for medical image segmentation. PCA learns the pixel structure and context information in each patch to get enough gradient feedback, which aids the discriminator in convergent to an optimal state. Our method outperforms the state-of-the-art semi-supervised methods, which demonstrates its effectiveness for medical image segmentation.
  arXiv  Detail & Related papers  (2022-07-24T07:45:47Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.