VUNO Publishes Study on its Deep Learning-based Gastric Cancer Pathology Solution in CCR
- 11. 17. 2020
VUNO Publishes Study on its Deep Learning-based Gastric Cancer Pathology Solution in
Clinical Cancer Research (CCR)
South Korean artificial intelligence (AI) developer VUNO Inc. announced that it published a study on VUNO Med®-PathGC AI™, a deep learning-based gastric cancer pathology solution of VUNO, in Clinical Cancer Research (CCR), which is one of the most world-renowned medical journals on oncology. The study, co-authored by the Department of Pathology at VUNO and Green Cross Laboratories, demonstrates the solution's outstandingly high level of performance and accuracy based on prospective clinical data, thus garnering significant attention.
(Jeonghyuk Park et al., A prospective validation and observer performance study of a deep learning algorithm for pathologic diagnosis of gastric tumors in endoscopic biopsies, Author Manuscript Published Online First on November 10, 2020; DOI: 10.1158/1078-0432.CCR-20-3159)
VUNO's Department of Pathology embarked on research to validate the performance and accuracy of its AI-based gastric cancer pathology solution VUNO Med®-PathGC AI™ in the real clinical settings. in the five-month-long prospective study, the performance of the algorithm was evaluated with 7,440 biopsy specimens, showing a superb performance with the accuracy of 100% and the specificity of 97% in detecting gastric carcinoma and adenoma.
In addition, an observer study with participation from six experienced pathologists who had a minimum of 5 years of surgical pathology experiences - was performed to compare the differences according to their use of an AI-based diagnostic supporting solution. With a variety of pathologic environments taken into consideration, Each pathologist was randomly designated to assess the three sets by a conventional microscope (Mic), a digital image viewer (DV), and algorithm-assisted digital image viewer (AADV)equipped with VUNO Med®-PathGC AI™. According to the results of the study, the group who utilized VUNO Med®-PathGC AI™ demonstrated a shorter review time per image by up to 58% compared to Mic group although no significant difference in accuracy, sensitivity, or specificity was observed among the three groups.
Kyu-Hwan Jung, Chief Technology Officer of VUNO, said, “VUNO has enthusiastically engaged in research and development in digital pathology with special focus on morphometry, diagnostic assistance, and prognostic and predictive biomarkers, etc. This research holds special significance as the first results in the field of diagnostic assistance.” He added, "We hope to see the fruits of various projects that are underway in cooperation with top-class medical organizations at home and abroad, following the recent research on prognostic biomarkers based on histopathology images, which was presented earlier this year at the annual meeting of AACR(American Association for Cancer Research), the ASCO (American Society of Clinical Oncology).”