SLIViT: a general AI framework for clinical-feature diagnosis from limited 3D biomedical-imaging data.

Autor:	Avram O; Department of Computational Medicine, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Computer Science, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Anesthesiology and Perioperative Medicine, University of California Los Angeles, Los Angeles, California, United States of America., Durmus B; Department of Computer Science, University of California Los Angeles, Los Angeles, California, United States of America., Rakocz N; Department of Computer Science, University of California Los Angeles, Los Angeles, California, United States of America., Corradetti G; Doheny Eye Institute, Pasadena, California, United States of America.; Department of Ophthalmology, University of California Los Angeles, Los Angeles, California, United States of America., An U; Department of Computational Medicine, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Computer Science, University of California Los Angeles, Los Angeles, California, United States of America., Nitalla MG; Doheny Eye Institute, Pasadena, California, United States of America.; Department of Ophthalmology, University of California Los Angeles, Los Angeles, California, United States of America., Rudas A; Department of Computational Medicine, University of California Los Angeles, Los Angeles, California, United States of America., Wakatsuki Y; Doheny Eye Institute, Pasadena, California, United States of America., Hirabayashi K; Doheny Eye Institute, Pasadena, California, United States of America., Velaga S; Doheny Eye Institute, Pasadena, California, United States of America., Tiosano L; Doheny Eye Institute, Pasadena, California, United States of America., Corvi F; Doheny Eye Institute, Pasadena, California, United States of America., Verma A; Doheny Eye Institute, Pasadena, California, United States of America.; Department of Ophthalmology and Visual Sciences, University of Louisville, Kentucky, United States of America., Karamat A; Doheny Eye Institute, Pasadena, California, United States of America., Lindenberg S; Doheny Eye Institute, Pasadena, California, United States of America., Oncel D; Doheny Eye Institute, Pasadena, California, United States of America., Almidani L; Doheny Eye Institute, Pasadena, California, United States of America., Hull V; Doheny Eye Institute, Pasadena, California, United States of America., Fasih-Ahmad S; Doheny Eye Institute, Pasadena, California, United States of America., Esmaeilkhanian H; Doheny Eye Institute, Pasadena, California, United States of America., Wykoff CC; Retina Consultants of Texas, Retina Consultants of America, Houston, Texas., Rahmani E; Department of Computational Medicine, University of California Los Angeles, Los Angeles, California, United States of America., Arnold CW; Departments of Radiology, University of California Los Angeles, Los Angeles, California, United States of America.; Departments of Bioengineering, University of California Los Angeles, Los Angeles, California, United States of America.; Departments of Pathology, University of California Los Angeles, Los Angeles, California, United States of America., Zhou B; Department of Computer Science, University of California Los Angeles, Los Angeles, California, United States of America., Zaitlen N; Department of Neurology, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Human Genetics, University of California Los Angeles, Los Angeles, California, United States of America., Gronau I; School of Computer Science, Reichman University, Herzliya, Israel., Sankararaman S; Department of Computational Medicine, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Computer Science, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Human Genetics, University of California Los Angeles, Los Angeles, California, United States of America., Chiang JN; Department of Computational Medicine, University of California Los Angeles, Los Angeles, California, United States of America., Sadda SR; Doheny Eye Institute, Pasadena, California, United States of America.; Department of Ophthalmology, University of California Los Angeles, Los Angeles, California, United States of America., Halperin E; Department of Computational Medicine, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Computer Science, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Anesthesiology and Perioperative Medicine, University of California Los Angeles, Los Angeles, California, United States of America.; Department of Human Genetics, University of California Los Angeles, Los Angeles, California, United States of America.
Jazyk:	angličtina
Zdroj:	Research square [Res Sq] 2023 Nov 21. Date of Electronic Publication: 2023 Nov 21.
DOI:	10.21203/rs.3.rs-3044914/v2
Abstrakt:	We present SLIViT, a deep-learning framework that accurately measures disease-related risk factors in volumetric biomedical imaging, such as magnetic resonance imaging (MRI) scans, optical coherence tomography (OCT) scans, and ultrasound videos. To evaluate SLIViT, we applied it to five different datasets of these three different data modalities tackling seven learning tasks (including both classification and regression) and found that it consistently and significantly outperforms domain-specific state-of-the-art models, typically improving performance (ROC AUC or correlation) by 0.1-0.4. Notably, compared to existing approaches, SLIViT can be applied even when only a small number of annotated training samples is available, which is often a constraint in medical applications. When trained on less than 700 annotated volumes, SLIViT obtained accuracy comparable to trained clinical specialists while reducing annotation time by a factor of 5,000 demonstrating its utility to automate and expedite ongoing research and other practical clinical scenarios. Competing Interests: Additional Declarations: Yes there is potential Competing Interest. Prof. Eran Halperin has an affiliation with Optum.
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu