Abstract
Thyroid ultrasound is the first-line exam for assessing thyroid nodules and determining whether biopsy is warranted. In routine reporting, radiologists produce two coupled outputs: a nodule contour for measurement and a TI-RADS risk category based on sonographic criteria. Yet both contouring style and risk grading vary across readers, creating inconsistent supervision that can degrade standard learning pipelines. In this paper, we address this workflow with a clinically guided multitask framework that jointly predicts the nodule mask and TI-RADS category within a single model. To ground risk prediction in clinically meaningful evidence, we guide the classification embedding using a compact TI-RADS aligned radiomics target during training, while preserving complementary deep features for discriminative performance. However, under annotator variability, naive multitask optimization often fails not because the tasks are unrelated, but because their gradients compete within the shared representation. To make this competition explicit and controllable, we introduce RLAR, a representation-level adversarial gradient regularizer. Rather than performing parameter-level gradient surgery, RLAR uses each task's normalized adversarial direction in latent space as a geometric probe of task sensitivity and penalizes excessive angular alignment between task-specific adversarial directions. On a public TI-RADS dataset, our clinically guided multitask model with RLAR consistently improves risk stratification while maintaining segmentation quality compared to single-task training and conventional multitask baselines. Code and pretrained models will be released.
Abstract (translated)
甲状腺超声是评估甲状腺结节及判断是否需要活检的首选检查。在常规报告中，放射科医生会生成两种耦合输出：用于测量的结节轮廓，以及基于超声标准的TI-RADS风险分类。然而，不同阅片者的勾画风格和风险分级存在差异，导致监督不一致，可能削弱标准学习流程的效果。本文针对这一工作流程，提出一种临床引导的多任务框架，可在单一模型中联合预测结节掩膜与TI-RADS分类。为使风险预测基于有临床意义的证据，我们在训练中使用紧凑的TI-RADS对齐影像组学目标引导分类嵌入，同时保留互补的深度特征以提升判别性能。然而，在标注者存在变异的情况下，朴素的多任务优化常会失败——并非因为任务不相关，而是因为它们的梯度在共享表征中相互竞争。为使这种竞争显式且可控，我们引入RLAR（表征层面对抗梯度正则化器）。RLAR并非执行参数层面的梯度手术，而是利用各任务在隐空间归一化的对抗方向作为任务敏感性的几何探针，并对任务特定对抗方向间过大的角度对齐进行惩罚。在一个公开的TI-RADS数据集上，我们的临床引导多任务模型结合RLAR，在保持分割质量的同时，持续提升了风险分层效果，优于单任务训练和常规多任务基线。代码与预训练模型将公开发布。
URL
https://arxiv.org/abs/2603.21095
