When are radiology reports useful for training medical image classifiers?

• URL: http://arxiv.org/abs/2510.24385v2
• Date: Wed, 29 Oct 2025 08:55:13 GMT
• Title: When are radiology reports useful for training medical image classifiers?
• Abstract summary: We conduct a systematic study of how radiology reports can be used during both pre-training and fine-tuning.<n>Our findings reveal that pre-training through explicit image-text alignment can be detrimental in settings where it's not.<n>These results provide actionable insights into when and how to leverage privileged text data to train medical image classifiers.
• Score: 17.157946138233758
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Medical images used to train machine learning models are often accompanied by radiology reports containing rich expert annotations. However, relying on these reports as inputs for clinical prediction requires the timely manual work of a trained radiologist. This raises a natural question: when can radiology reports be leveraged during training to improve image-only classification? Prior works are limited to evaluating pre-trained image representations by fine-tuning them to predict diagnostic labels, often extracted from reports, ignoring tasks with labels that are weakly associated with the text. To address this gap, we conduct a systematic study of how radiology reports can be used during both pre-training and fine-tuning, across diagnostic and prognostic tasks (e.g., 12-month readmission), and under varying training set sizes. Our findings reveal that: (1) Leveraging reports during pre-training is beneficial for downstream classification tasks where the label is well-represented in the text; however, pre-training through explicit image-text alignment can be detrimental in settings where it's not; (2) Fine-tuning with reports can lead to significant improvements and even have a larger impact than the pre-training method in certain settings. These results provide actionable insights into when and how to leverage privileged text data to train medical image classifiers while highlighting gaps in current research.

Related papers

• Uncertainty-aware Medical Diagnostic Phrase Identification and Grounding [72.18719355481052]
  We introduce a novel task called Medical Report Grounding (MRG)<n>MRG aims to directly identify diagnostic phrases and their corresponding grounding boxes from medical reports in an end-to-end manner.<n>We propose uMedGround, a robust and reliable framework that leverages a multimodal large language model to predict diagnostic phrases.
  arXiv  Detail & Related papers  (2024-04-10T07:41:35Z)
• Mining Gaze for Contrastive Learning toward Computer-Assisted Diagnosis [61.089776864520594]
  We propose eye-tracking as an alternative to text reports for medical images. By tracking the gaze of radiologists as they read and diagnose medical images, we can understand their visual attention and clinical reasoning. We introduce the Medical contrastive Gaze Image Pre-training (McGIP) as a plug-and-play module for contrastive learning frameworks.
  arXiv  Detail & Related papers  (2023-12-11T02:27:45Z)
• Controllable Chest X-Ray Report Generation from Longitudinal Representations [13.151444796296868]
  One strategy to speed up reporting is to integrate automated reporting systems. Previous approaches to automated radiology reporting generally do not provide the prior study as input. We introduce two novel aspects: (1) longitudinal learning -- we propose a method to align, leverage the current and prior scan information into a joint longitudinal representation which can be provided to the multimodal report generation model; (2) sentence-anatomy dropout -- a training strategy for controllability in which the report generator model is trained to predict only sentences from the original report which correspond to the subset of anatomical regions given as input.
  arXiv  Detail & Related papers  (2023-10-09T17:22:58Z)
• Vision-Language Modelling For Radiological Imaging and Reports In The Low Data Regime [70.04389979779195]
  This paper explores training medical vision-language models (VLMs) where the visual and language inputs are embedded into a common space. We explore several candidate methods to improve low-data performance, including adapting generic pre-trained models to novel image and text domains. Using text-to-image retrieval as a benchmark, we evaluate the performance of these methods with variable sized training datasets of paired chest X-rays and radiological reports.
  arXiv  Detail & Related papers  (2023-03-30T18:20:00Z)
• Learning to Exploit Temporal Structure for Biomedical Vision-Language Processing [53.89917396428747]
  Self-supervised learning in vision-language processing exploits semantic alignment between imaging and text modalities. We explicitly account for prior images and reports when available during both training and fine-tuning. Our approach, named BioViL-T, uses a CNN-Transformer hybrid multi-image encoder trained jointly with a text model.
  arXiv  Detail & Related papers  (2023-01-11T16:35:33Z)
• Weakly Supervised Learning Significantly Reduces the Number of Labels Required for Intracranial Hemorrhage Detection on Head CT [7.713240800142863]
  Machine learning pipelines, in particular those based on deep learning (DL) models, require large amounts of labeled data. This work studies the question of what kind of labels should be collected for the problem of intracranial hemorrhage detection in brain CT. We find that strong supervision (i.e., learning with local image-level annotations) and weak supervision (i.e., learning with only global examination-level labels) achieve comparable performance.
  arXiv  Detail & Related papers  (2022-11-29T04:42:41Z)
• RadTex: Learning Efficient Radiograph Representations from Text Reports [7.090896766922791]
  We build a data-efficient learning framework that utilizes radiology reports to improve medical image classification performance with limited labeled data. Our model achieves higher classification performance than ImageNet-supervised pretraining when labeled training data is limited.
  arXiv  Detail & Related papers  (2022-08-05T15:06:26Z)
• Self-Supervised Learning as a Means To Reduce the Need for Labeled Data in Medical Image Analysis [64.4093648042484]
  We use a dataset of chest X-ray images with bounding box labels for 13 different classes of anomalies. We show that it is possible to achieve similar performance to a fully supervised model in terms of mean average precision and accuracy with only 60% of the labeled data.
  arXiv  Detail & Related papers  (2022-06-01T09:20:30Z)
• Breaking with Fixed Set Pathology Recognition through Report-Guided Contrastive Training [23.506879497561712]
  We employ a contrastive global-local dual-encoder architecture to learn concepts directly from unstructured medical reports. We evaluate our approach on the large-scale chest X-Ray datasets MIMIC-CXR, CheXpert, and ChestX-Ray14 for disease classification.
  arXiv  Detail & Related papers  (2022-05-14T21:44:05Z)
• FlexR: Few-shot Classification with Language Embeddings for Structured Reporting of Chest X-rays [37.15474283789249]
  We propose a method to predict clinical findings defined by sentences in structured reporting templates. The approach involves training a contrastive language-image model using chest X-rays and related free-text radiological reports. Results show that even with limited image-level annotations for training, the method can accomplish the structured reporting tasks of severity assessment of cardiomegaly and localizing pathologies in chest X-rays.
  arXiv  Detail & Related papers  (2022-03-29T16:31:39Z)
• Contrastive Cross-Modal Pre-Training: A General Strategy for Small Sample Medical Imaging [17.530514174860826]
  We propose a form of weak supervision to improve the image interpretation performance of a neural network. We use an image-text matching task to train a feature extractor and then fine-tune it in a transfer learning setting for a supervised task. The end result is a neural network that automatically interprets imagery without requiring textual reports during inference.
  arXiv  Detail & Related papers  (2020-10-06T22:20:29Z)
• Auxiliary Signal-Guided Knowledge Encoder-Decoder for Medical Report Generation [107.3538598876467]
  We propose an Auxiliary Signal-Guided Knowledge-Decoder (ASGK) to mimic radiologists' working patterns. ASGK integrates internal visual feature fusion and external medical linguistic information to guide medical knowledge transfer and learning.
  arXiv  Detail & Related papers  (2020-06-06T01:00:15Z)

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.