Is Synthetic Image Augmentation Useful for Imbalanced Classification Problems? Case-Study on the MIDOG2025 Atypical Cell Detection Competition

• URL: http://arxiv.org/abs/2509.02612v1
• Date: Sat, 30 Aug 2025 23:37:20 GMT
• Title: Is Synthetic Image Augmentation Useful for Imbalanced Classification Problems? Case-Study on the MIDOG2025 Atypical Cell Detection Competition
• Abstract summary: The MIDOG 2025 challenge extends prior work on mitotic figure detection by introducing a new Track 2 on atypical mitosis classification.<n>This task aims to distinguish normal from atypical figures in histopathology images.<n>We investigated two complementary backbones: (i) ConvNeXt-Small, pretrained on ImageNet, and (ii) a histopathology-specific ViT from Lunit trained via self-supervision.
• Score: 2.4582766828986076
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: The MIDOG 2025 challenge extends prior work on mitotic figure detection by introducing a new Track 2 on atypical mitosis classification. This task aims to distinguish normal from atypical mitotic figures in histopathology images, a clinically relevant but highly imbalanced and cross-domain problem. We investigated two complementary backbones: (i) ConvNeXt-Small, pretrained on ImageNet, and (ii) a histopathology-specific ViT from Lunit trained via self-supervision. To address the strong prevalence imbalance (9408 normal vs. 1741 atypical), we synthesized additional atypical examples to approximate class balance and compared models trained with real-only vs. real+synthetic data. Using five-fold cross-validation, both backbones reached strong performance (mean AUROC approximately 95 percent), with ConvNeXt achieving slightly higher peaks while Lunit exhibited greater fold-to-fold stability. Synthetic balancing, however, did not lead to consistent improvements. On the organizers' preliminary hidden test set, explicitly designed as an out-of-distribution debug subset, ConvNeXt attained the highest AUROC (95.4 percent), whereas Lunit remained competitive on balanced accuracy. These findings suggest that both ImageNet and domain-pretrained backbones are viable for atypical mitosis classification, with domain-pretraining conferring robustness and ImageNet pretraining reaching higher peaks, while naive synthetic balancing has limited benefit. Full hidden test set results will be reported upon challenge completion.

Related papers

• From Pixels to Explanations: Interpretable Diabetic Retinopathy Grading with CNN-Transformer Ensembles, Visual Explainability and Vision-Language Models [7.782282985072339]
  This study presents a methodology that combines strong discriminative models with multimodal explanations.<n>For interpretability, we produced Grad-CAM++ visual attribution maps and short textual rationales.<n>For explanation quality, Grad-CAM++ offered plausible but coarse localization, and VLM rationales were generally grade-consistent.
  arXiv  Detail & Related papers  (2026-04-25T00:21:11Z)
• Towards Syn-to-Real IQA: A Novel Perspective on Reshaping Synthetic Data Distributions [74.00222571094437]
  Blind Image Quality Assessment (BIQA) has advanced significantly through deep learning, but the scarcity of large-scale labeled datasets remains a challenge.<n>We make a key observation that representations learned from synthetic datasets often exhibit a discrete and clustered pattern that hinders regression performance.<n>We introduce a novel framework SynDR-IQA, which reshapes synthetic data distribution to enhance BIQA generalization.
  arXiv  Detail & Related papers  (2026-01-01T06:11:16Z)
• Medical Image Classification on Imbalanced Data Using ProGAN and SMA-Optimized ResNet: Application to COVID-19 [1.270952934304684]
  This study proposes a progressive generative adversarial network to generate synthetic data to supplement the real ones.<n>The proposed model achieves 95.5% and 98.5% accuracy for 4-class and 2-class imbalanced classification problems, respectively.
  arXiv  Detail & Related papers  (2025-12-30T13:26:01Z)
• Improving Diagnostic Performance on Small and Imbalanced Datasets Using Class-Based Input Image Composition [0.2676349883103403]
  Small, imbalanced datasets and poor input image quality can lead to high false predictions rates with deep learning models.<n>This paper introduces Class-Based Image Composition, an approach that allows us to reformulate training inputs through a fusion of multiple images of the same class into combined visual composites.
  arXiv  Detail & Related papers  (2025-11-05T22:34:06Z)
• Robust Atypical Mitosis Classification with DenseNet121: Stain-Aware Augmentation and Hybrid Loss for Domain Generalization [16.048949963804123]
  We present a DenseNet-121-based framework tailored for atypical mitosis classification in the MIDOG 2025 (Track 2) setting.<n>Our method integrates stain-aware augmentation (Macenko), geometric and intensity transformations, and imbalance-aware learning via weighted sampling with a hybrid objective combining class-weighted binary cross-entropy and focal loss.<n> Trained end-to-end with AdamW and evaluated across multiple independent domains, the model demonstrates strong generalization under scanner and staining shifts, achieving balanced accuracy 85.0%, AUROC 0.927, sensitivity 89.2%, and specificity 80.9% on the official test set.
  arXiv  Detail & Related papers  (2025-10-26T11:24:55Z)
• Adaptive Learning Strategies for Mitotic Figure Classification in MIDOG2025 Challenge [7.3323821474776]
  We investigated three variants of adapting the pathology foundation model UNI2-h for the MIDOG2025 Track 2 challenge.<n>We found that visual prompt tuning (VPT) substantially improved generalization, and that further integrating test-time augmentation (TTA) with Vahadane and Macenko stain normalization provided the best robustness.
  arXiv  Detail & Related papers  (2025-09-01T22:42:53Z)
• ConvNeXt with Histopathology-Specific Augmentations for Mitotic Figure Classification [1.398256265458105]
  We propose a solution based on the lightweight ConvNeXt architecture to maximize domain coverage.<n>On the preliminary leaderboard, our model achieved a balanced accuracy of 0.8961, ranking among the top entries.
  arXiv  Detail & Related papers  (2025-08-29T13:18:32Z)
• Efficient Fine-Tuning of DINOv3 Pretrained on Natural Images for Atypical Mitotic Figure Classification in MIDOG 2025 [1.7259725776748482]
  Atypical mitotic figures (AMFs) are markers of abnormal cell division associated with poor prognosis.<n>The MIDOG 2025 challenge introduces a benchmark for AMF classification across multiple domains.<n>We evaluate the recently published DINOv3-H+ vision transformer, pretrained on natural images.
  arXiv  Detail & Related papers  (2025-08-28T17:45:22Z)
• CytoDiff: AI-Driven Cytomorphology Image Synthesis for Medical Diagnostics [0.9350583142732543]
  We introduce CytoDiff, a stable diffusion model fine-tuned with LoRA weights and guided by few-shot samples that generates high-fidelity synthetic white blood cell images.<n>Using a small, highly imbalanced real dataset, the addition of 5,000 synthetic images per class improved ResNet classifier accuracy from 27% to 78% (+51%)<n>Similarly, CLIP-based classification accuracy increased from 62% to 77% (+15%)
  arXiv  Detail & Related papers  (2025-07-07T14:49:05Z)
• Multi-Level Collaboration in Model Merging [56.31088116526825]
  This paper explores the intrinsic connections between model merging and model ensembling.<n>We find that even when previous restrictions are not met, there is still a way for model merging to attain a near-identical and superior performance similar to that of ensembling.
  arXiv  Detail & Related papers  (2025-03-03T07:45:04Z)
• Merging synthetic and real embryo data for advanced AI predictions [69.07284335967019]
  We train two generative models using two datasets-one we created and made publicly available, and one existing public dataset-to generate synthetic embryo images at various cell stages.<n>These were combined with real images to train classification models for embryo cell stage prediction.<n>Our results demonstrate that incorporating synthetic images alongside real data improved classification performance, with the model achieving 97% accuracy compared to 94.5% when trained solely on real data.
  arXiv  Detail & Related papers  (2024-12-02T08:24:49Z)
• Forgery-aware Adaptive Transformer for Generalizable Synthetic Image Detection [106.39544368711427]
  We study the problem of generalizable synthetic image detection, aiming to detect forgery images from diverse generative methods. We present a novel forgery-aware adaptive transformer approach, namely FatFormer. Our approach tuned on 4-class ProGAN data attains an average of 98% accuracy to unseen GANs, and surprisingly generalizes to unseen diffusion models with 95% accuracy.
  arXiv  Detail & Related papers  (2023-12-27T17:36:32Z)
• Rethinking Semi-Supervised Medical Image Segmentation: A Variance-Reduction Perspective [51.70661197256033]
  We propose ARCO, a semi-supervised contrastive learning framework with stratified group theory for medical image segmentation. We first propose building ARCO through the concept of variance-reduced estimation and show that certain variance-reduction techniques are particularly beneficial in pixel/voxel-level segmentation tasks. We experimentally validate our approaches on eight benchmarks, i.e., five 2D/3D medical and three semantic segmentation datasets, with different label settings.
  arXiv  Detail & Related papers  (2023-02-03T13:50:25Z)
• Fine-Grained Hard Negative Mining: Generalizing Mitosis Detection with a Fifth of the MIDOG 2022 Dataset [1.2183405753834562]
  We describe a candidate deep learning solution for the Mitosis Domain Generalization Challenge 2022 (MIDOG) Our approach consists in training a rotation-invariant deep learning model using aggressive data augmentation. Our model ensemble achieved a F1-score of.697 on the final test set after automated evaluation.
  arXiv  Detail & Related papers  (2023-01-03T13:06:44Z)
• Stochastic Transformer Networks with Linear Competing Units: Application to end-to-end SL Translation [46.733644368276764]
  We introduce an end-to-end SLT model that does not entail explicit use of glosses. This is in stark contrast to existing end-to-end models that use gloss sequence groundtruth. We demonstrate that our approach can reach the currently best reported BLEU-4 score on the ENIX 2014T benchmark.
  arXiv  Detail & Related papers  (2021-09-01T15:00:52Z)
• Vision Transformers for femur fracture classification [59.99241204074268]
  The Vision Transformer (ViT) was able to correctly predict 83% of the test images. Good results were obtained in sub-fractures with the largest and richest dataset ever.
  arXiv  Detail & Related papers  (2021-08-07T10:12:42Z)

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.