Generating Structured Radiological Reports from Brain MRI Data Based on SigLIP2 Frozen Embeddings
Electronics and Control Systems · 18.04.2026 · 2 (88) · 38-44
DOI: 10.18372/1990-5548.88.20960 ↗Kurz gefasst
Fachartikel in der Zeitschrift Electronics and Control Systems, online erschienen am 18.04.2026. Verfasst von Andrew Sheruda. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Automatic generation of clinical reports from medical images is a relevant task capable of reducing the workload of radiologists and standardizing documentation. In this paper, we investigate an approach to generating structured reports from brain MRI data using a pre-trained multimodal SigLIP2 model as a feature extractor. We propose an architecture in which visual embeddings obtained from a frozen SigLIP2 are projected into the representation space of the GPT-2 language model for subsequent text generation. Experiments were conducted on the open-access BIOSE MRI dataset, containing 34 pairs of "MRI image + clinical report". It is shown that the proposed approach generates semantically meaningful reports, achieving quality comparable to more complex architectures with substantially lower computational costs. Additionally, the influence of pre-training SigLIP2 on a classification task (Brain3-Anomaly-SigLIP2 version) on generation quality is investigated. The results demonstrate the potential of using frozen vision encoders in medical generative tasks under data-scarce conditions.
Zitieren
APA
Sheruda, A. (2026). Generating Structured Radiological Reports from Brain MRI Data Based on SigLIP2 Frozen Embeddings. Electronics and Control Systems, 2(88), 38-44. https://doi.org/10.18372/1990-5548.88.20960
Vancouver
Sheruda A. Generating Structured Radiological Reports from Brain MRI Data Based on SigLIP2 Frozen Embeddings. Electronics and Control Systems. 2026;2(88):38-44. doi:10.18372/1990-5548.88.20960
BibTeX
@article{Sheruda2026,
author = {Sheruda, Andrew},
title = {Generating Structured Radiological Reports from Brain MRI Data Based on SigLIP2 Frozen Embeddings},
journal = {Electronics and Control Systems},
year = {2026},
volume = {2},
number = {88},
pages = {38--44},
doi = {10.18372/1990-5548.88.20960}
}
EUVIMED