Application of YOLOv8 and a model based on vision transformers and UNet for LVNC diagnosis: advantages and limitations

De Haro, Salvador; González Férez, Pilar; García, José M.; Bernabé García, Gregorio

Publication:
Application of YOLOv8 and a model based on vision transformers and UNet for LVNC diagnosis: advantages and limitations

dc.contributor.author	De Haro, Salvador
dc.contributor.author	González Férez, Pilar
dc.contributor.author	García, José M.
dc.contributor.author	Bernabé García, Gregorio
dc.contributor.department	Ingeniería y Tecnología de Computadores	es
dc.date.accessioned	2025-05-14T08:59:54Z
dc.date.available	2025-05-14T08:59:54Z
dc.date.issued	2025-04-25
dc.description	© 2025 The Author(s), under exclusive license to Springer Nature Switzerland AG.
dc.description.abstract	Hypertrabeculation or left ventricular non-compaction (LVNC) is a cardiac condition that has recently been recognized. While several methods exist for accurately measuring the trabeculae in the ventricle, there is still no consensus within the medical community regarding the optimal approach. In previous work, we introduced DL-LVTQ, a tool based on a UNet convolutional neural network designed to quantify the trabeculae in the left ventricle. In this paper, we present an expanded dataset that includes new patients affected by a cardiomyopathy known as Titin, necessitating the retraining of the models involved in our study on this updated dataset to accurately infer future patients with this condition. We also introduce ViTUNet, a hybrid architecture that aims to merge the benefits of UNet and Vision Transformers for precise segmentation of the left ventricle. Furthermore, we train a YOLOv8 model to detect the left ventricle and integrate it with the hybrid model to focus segmentation on a region of interest around the ventricle. Regarding the precision quality achieved by ViTUNet using YOLOv8, results are quite similar to those obtained by the DL-LVTQ tool, suggesting that the dataset is a limiting factor in our improvement. To substantiate this, we conduct a detailed analysis of the MRI slices in the current dataset. By identifying and removing problematic slices, results significantly improve. The introduction of a YOLOv8 model alongside a deep learning model presents a promising approach.
dc.embargo.terms	Si
dc.format	application/pdf	es
dc.format.extent	11	es
dc.identifier.doi	https://doi.org/10.1007/978-3-031-87873-2_14
dc.identifier.eissn	Print: 978-3-031-87872-5
dc.identifier.eissn	Electronic: 978-3-031-87873-2
dc.identifier.uri	http://hdl.handle.net/10201/154560
dc.language	eng	es
dc.publisher	Springer
dc.relation	This work has been partially funded by Grant TED2021-129221B-I00 funded by MCIN/AEI/10.13039/501100011033 and by the “European Union NextGenerationEU/PRTR”.	es
dc.relation.ispartof	Practical Applications of Computational Biology and Bioinformatics, 18th International Conference (PACBB 2024), p.p. 132–142	es
dc.relation.publisherversion	https://link.springer.com/chapter/10.1007/978-3-031-87873-2_14
dc.rights	info:eu-repo/semantics/embargoedAccess	es
dc.subject	Left ventricular non-compaction diagnosis
dc.subject	UNet
dc.subject	Vision
dc.subject	Transformers
dc.subject	YOLOv8
dc.subject	MRI Image segmentation
dc.subject	Data analysis
dc.title	Application of YOLOv8 and a model based on vision transformers and UNet for LVNC diagnosis: advantages and limitations	es
dc.type	info:eu-repo/semantics/article	es
dspace.entity.type	Publication	es