---
title: "面向乳腺X光片迁移学习的“数据集知情迁移学习”框架：在乳腺癌筛查与病变诊断中的统一优化方法"
source: "arXiv AI/CL/LG（论文）"
source_tier: "t2"
url: "https://arxiv.org/abs/2607.26043v1"
permalink: "https://feed.astralor.com/items/itm_cf4c4fc54c51248c2394644c"
published: "2026-07-28T17:52:36.000Z"
score: 74
channel: "tech"
tags: ["DITL","VinDR-Mammo","乳腺X光分类","迁移学习"]
---

# 面向乳腺X光片迁移学习的“数据集知情迁移学习”框架：在乳腺癌筛查与病变诊断中的统一优化方法

## AI 点评

> 这篇论文最值得关注的不是某个特定任务上的SOTA，而是它用一种非常优雅的框架消解了迁移学习中‘调参依赖’和‘数据规模适应性’两个痛点。对于从事医疗AI落地的研究者和工程师而言，DITL提供了一个即插即用的优化思路，尤其适合那些数据质量参差不齐的实际场景。

## 速览

传统迁移学习在乳腺X光片分类中往往忽略数据集特性，且现有邻域信息方法局限于窄任务。本文提出数据集知情迁移学习（DITL）框架，通过自适应难度加权交叉熵与自适应邻域三元组损失，在无需超参数调优的情况下，在大型VinDR-Mammo数据集和小型ROI数据集上均取得显著性能提升，覆盖筛查到诊断全流程。

## 关键点

- 提出DITL框架，包含自适应难度加权交叉熵（A-DWCE）和自适应邻域表示三元组损失（A-NR-Triplet）。
- 无需超参数调优，计算开销可忽略，适用于从小型到大规模数据集。
- 在VinDR-Mammo数据集上全图像乳腺密度分类达到SOTA，准确率、F1和AUC显著提升（p<0.0001）。
- 在小ROI数据集上也获得一致且统计显著的性能增益，覆盖乳腺癌筛查与病变诊断。

## 全文

Enhancing classification performance in mammography remains a persistent challenge across both small curated datasets and large-scale clinical cohorts. Conventional transfer learning approaches often neglect dataset-specific characteristics, while recent neighborhood-informed methods have been restricted to narrow tasks with rigid formulations, limiting their scalability to population-level datasets. To address these challenges, we propose the Dataset-Informed Transfer Learning (DITL) framework, which integrates dataset-derived difficulty signals with neighborhood-based triplet supervision in a unified objective. DITL introduces two adaptive components: (i) Adaptive Difficulty-Weighted Cross-Entropy (A-DWCE), which assigns per-sample weights based on k-nearest neighbor label purity in a self-supervised feature space, and (ii) Adaptive Neighborhood Representation Triplet (A-NR-Triplet), which enforces intra-class compactness and inter-class separation using a learnable margin. Unlike focal loss, DITL requires no hyperparameter tuning, removes heuristic weighting and fixed margins, and incurs negligible computational overhead, yielding a robust and scalable optimization strategy. On the large-scale VinDR-Mammo dataset, DITL achieves state-of-the-art performance for whole-image breast density classification, with significant improvements across accuracy, F1-score, and AUC (p < 0.0001). Beyond large cohorts, DITL also delivers consistent, statistically significant gains on small ROI datasets (p < 0.0001). By bridging small-scale lesion analysis with large-scale density estimation, DITL establishes a clinically relevant, scalable, and generalizable framework for mammography classification, spanning the full breast cancer screening-to-diagnosis spectrum.

---

引用前请核对原文;本文由星际流动整理:https://feed.astralor.com/items/itm_cf4c4fc54c51248c2394644c