Efficient Augmented Intelligence Framework for Bladder Lesion Detection
JCO Clinical Cancer Informatics · 2023
DOI: 10.1200/cci.23.00031 ↗Kurz gefasst
Fachartikel in der Zeitschrift JCO Clinical Cancer Informatics, 2023. Verfasst von Okyaz Eminaga und 11 weiteren Autor:innen. Beteiligte Einrichtung: AI Vobis, Palo Alto, CA. Hauptthemen laut MEDLINE: Urinary Bladder Neoplasms. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 35 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: BLRD VA, NCI NIH HHS.
Abstract (Originalsprache)
PURPOSE Development of intelligence systems for bladder lesion detection is cost intensive. An efficient strategy to develop such intelligence solutions is needed. MATERIALS AND METHODS We used four deep learning models (ConvNeXt, PlexusNet, MobileNet, and SwinTransformer) covering a variety of model complexity and efficacy. We trained these models on a previously published educational cystoscopy atlas (n = 312 images) to estimate the ratio between normal and cancer scores and externally validated on cystoscopy videos from 68 cases, with region of interest (ROI) pathologically confirmed to be benign and cancerous bladder lesions (ie, ROI). The performance measurement included specificity and sensitivity at frame level, frame sequence (block) level, and ROI level for each case. RESULTS Specificity was comparable between four models at frame (range, 30.0%-44.8%) and block levels (56%-67%). Although sensitivity at the frame level (range, 81.4%-88.1%) differed between the models, sensitivity at the block level (100%) and ROI level (100%) was comparable between these models. MobileNet and PlexusNet were computationally more efficient for real-time ROI detection than ConvNeXt and SwinTransformer. CONCLUSION Educational cystoscopy atlas and efficient models facilitate the development of real-time intelligence system for bladder lesion detection.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- AI Vobis, Palo Alto, CA.
- Center for Artificial Intelligence in Medicine and Imaging, Stanford University School of Medicine, Stanford, CA.
- Department of Urology, Stanford University School of Medicine, Stanford, CA.
- Veterans Affairs Palo Alto Health Care System, Palo Alto, CA.
- Department of Radiation Oncology, Stanford University School of Medicine, Stanford, CA.
- Department of Urology, Muenster University Hospital, Muenster, Germany.
ORCID-Profile: Okyaz Eminaga, Mark Laurie, T Jessie Ge, Vinh La, Martin Bogemann, Hubert Lau, Joseph C Liao
Förderung
- BLRD VA
- NCI NIH HHS
Zitieren
APA
Eminaga, O., Lee, T. J., Laurie, M., Ge, T. J., La, V., Long, J., Semjonow, A., Bogemann, M., Lau, H., Shkolyar, E., Xing, L., & Liao, J. C. (2023). Efficient Augmented Intelligence Framework for Bladder Lesion Detection. JCO Clinical Cancer Informatics. https://doi.org/10.1200/cci.23.00031
Vancouver
Eminaga O, Lee TJ, Laurie M, Ge TJ, La V, Long J, et al. Efficient Augmented Intelligence Framework for Bladder Lesion Detection. JCO Clinical Cancer Informatics. 2023. doi:10.1200/cci.23.00031
BibTeX
@article{Eminaga2023,
author = {Eminaga, Okyaz and Lee, Timothy Jiyong and Laurie, Mark and Ge, T Jessie and La, Vinh and Long, Jin and Semjonow, Axel and Bogemann, Martin and Lau, Hubert and Shkolyar, Eugene and Xing, Lei and Liao, Joseph C},
title = {Efficient Augmented Intelligence Framework for Bladder Lesion Detection},
journal = {JCO Clinical Cancer Informatics},
year = {2023},
doi = {10.1200/cci.23.00031}
}
EUVIMED