Saved in:
| Main Authors: | , , , , , , , , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.11653 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866914469544722432 |
|---|---|
| author | Wong, David Isik, Zeynep Wang, Bin Tliba, Marouane Durak, Gorkem Keles, Elif Aktas, Halil Ertugrul Chetouani, Aladine Topel, Cagdas Gennaro, Nicolo Vendrami, Camila Lopes Trabzonlu, Tugce Agirlar Rahsepar, Amir Ali Perronne, Laetitia Antalek, Matthew Ozturk, Onural Okur, Gokcan Gordon, Andrew C. Pyrros, Ayis Miller, Frank H. Borhani, Amir Savas, Hatice Hart, Eric Krupinski, Elizabeth Bagci, Ulas |
| author_facet | Wong, David Isik, Zeynep Wang, Bin Tliba, Marouane Durak, Gorkem Keles, Elif Aktas, Halil Ertugrul Chetouani, Aladine Topel, Cagdas Gennaro, Nicolo Vendrami, Camila Lopes Trabzonlu, Tugce Agirlar Rahsepar, Amir Ali Perronne, Laetitia Antalek, Matthew Ozturk, Onural Okur, Gokcan Gordon, Andrew C. Pyrros, Ayis Miller, Frank H. Borhani, Amir Savas, Hatice Hart, Eric Krupinski, Elizabeth Bagci, Ulas |
| contents | We introduce GazeVaLM, a public eye-tracking dataset for studying clinical perception during chest radiograph authenticity assessment. The dataset comprises 960 gaze recordings from 16 expert radiologists interpreting 30 real and 30 synthetic chest X-rays (generated by diffusion based generative AI) under two conditions: diagnostic assessment and real-fake classification (Visual Turing test). For each image-observer pair, we provide raw gaze samples, fixation maps, scanpaths, saliency density maps, structured diagnostic labels, and authenticity judgments. We extend the protocol to 6 state-of-the-art multimodal LLMs, releasing their predicted diagnoses, authenticity labels, and confidence scores under matched conditions - enabling direct human-AI comparison at both decision and uncertainty levels. We further provide analyses of gaze agreement, inter-observer consistency, and benchmarking of radiologists versus LLMs in diagnostic accuracy and authenticity detection. GazeVaLM supports research in gaze modeling, clinical decision-making, human-AI comparison, generative image realism assessment, and uncertainty quantification. By jointly releasing visual attention data, clinical labels, and model predictions, we aim to facilitate reproducible research on how experts and AI systems perceive, interpret, and evaluate medical images. The dataset is available at https://huggingface.co/datasets/davidcwong/GazeVaLM. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_11653 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | GazeVaLM: A Multi-Observer Eye-Tracking Benchmark for Evaluating Clinical Realism in AI-Generated X-Rays Wong, David Isik, Zeynep Wang, Bin Tliba, Marouane Durak, Gorkem Keles, Elif Aktas, Halil Ertugrul Chetouani, Aladine Topel, Cagdas Gennaro, Nicolo Vendrami, Camila Lopes Trabzonlu, Tugce Agirlar Rahsepar, Amir Ali Perronne, Laetitia Antalek, Matthew Ozturk, Onural Okur, Gokcan Gordon, Andrew C. Pyrros, Ayis Miller, Frank H. Borhani, Amir Savas, Hatice Hart, Eric Krupinski, Elizabeth Bagci, Ulas Computer Vision and Pattern Recognition We introduce GazeVaLM, a public eye-tracking dataset for studying clinical perception during chest radiograph authenticity assessment. The dataset comprises 960 gaze recordings from 16 expert radiologists interpreting 30 real and 30 synthetic chest X-rays (generated by diffusion based generative AI) under two conditions: diagnostic assessment and real-fake classification (Visual Turing test). For each image-observer pair, we provide raw gaze samples, fixation maps, scanpaths, saliency density maps, structured diagnostic labels, and authenticity judgments. We extend the protocol to 6 state-of-the-art multimodal LLMs, releasing their predicted diagnoses, authenticity labels, and confidence scores under matched conditions - enabling direct human-AI comparison at both decision and uncertainty levels. We further provide analyses of gaze agreement, inter-observer consistency, and benchmarking of radiologists versus LLMs in diagnostic accuracy and authenticity detection. GazeVaLM supports research in gaze modeling, clinical decision-making, human-AI comparison, generative image realism assessment, and uncertainty quantification. By jointly releasing visual attention data, clinical labels, and model predictions, we aim to facilitate reproducible research on how experts and AI systems perceive, interpret, and evaluate medical images. The dataset is available at https://huggingface.co/datasets/davidcwong/GazeVaLM. |
| title | GazeVaLM: A Multi-Observer Eye-Tracking Benchmark for Evaluating Clinical Realism in AI-Generated X-Rays |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.11653 |