Contact us


    Collection and Use of Personal Information. *


    Deep Bio’s AI-based Cancer Diagnostic Software Demonstrated Uropathologist-level Performance

    May 13, 2022

    Independent external validation study results of Deep Bio’s AI algorithm for prostate cancer diagnosis were published in the Modern Pathology


    SEOUL, SOUTH KOREA (PRWEB) MAY 13, 2022 – Deep Bio, a pioneer in medical AI for digital pathology and cancer diagnostics software, announced that the research results of an external validation study of its deep learning-based prostate cancer diagnostic support software, DeepDx® Prostate, were published in Modern Pathology, demonstrating high concordance to experts.


    The study, conducted in collaboration with Seoul National University Hospital (SNUH), aimed to validate the performance of DeepDx® Prostate for detecting and grading prostate cancer and evaluate its clinical value to general pathologists. In the study, 593 hematoxylin and eosin (H&E) stained whole-slide images of prostate biopsies, 130 normal and 463 adenocarcinomas, and their original pathology reports from the Department of Pathology of SNUH were included.


    To assess the performance of the algorithm, researchers compared both DeepDx® Prostate analysis results and the original reports against the reference standard established by three uropathology experts. For cancer detection, compared to the original hospital diagnoses, DeepDx® Prostate showed similar sensitivity, NPV and accuracy, and even improvement in specificity and PPV. For Gleason grade group assignments, the algorithm achieved higher concordance (0.713 kappa / 0.922 quadratic-weighted kappa) to the reference standard than the original reports (0.619 kappa / 0.873 quadratic-weighted kappa). Notably, the AI software outperformed the original report in the detection of Gleason patterns 4 and 5, achieving excellent agreement in quantifying areas of Gleason pattern 4.


    In addition to performance evaluation, the study also evaluated the utility of the algorithm in a clinical workflow. When a general pathologist evaluated cases with the aid of AI, the concordance of grade group between the user and the reference standard increased (kappa: 0.621 to 0.741 / quadratic-weighted kappa: 0.876 to 0.925), while the average analysis time per case saw a significant reduction from 55.7 to 36.8 seconds, about 34% reduction.


    “Accurate grading of prostate cancer biopsies is crucial to guiding disease management and treatment planning. However, pathologists often disagree in their diagnoses,” Jung Minsun, MD, PhD, at Department of Pathology, Yonsei University College of Medicine said. “DeepDx® Prostate can provide value here, by providing consistent results to reduce discrepancies among pathologists, as well as reduce time spent on the reading of the slides, resulting in improved diagnostic accuracy and efficiency at the same time,” added he.


    “It is meaningful that DeepDx® Prostate is proven to assist medical professionals at the skill level of uropathologist in cancer detection and grading of severity, which suggests that AI can be a potential screening process when diagnosing prostate cancer,” mentioned Sun-Woo Kim, CEO of Deep Bio. “We will continue to conduct research and develop AI assistance tools for other cancer types until they can be implemented in actual clinical settings.”


    Focusing not only on AI diagnostics, but also R&D, Deep Bio has been presenting its novel research results in prestigious science and technology journals including the Cancers, npj Digital Medicine, as well as in various international conferences and events. The company also continues to build its presence in the global market through overseas digital pathology solution providers in the US, Europe, and India, as well as conducting research cooperation with Stanford Medical School, Harvard Dana-Farber Cancer Institute, and other top research institutions in the US.


    About Deep Bio

    Deep Bio Inc. is an AI biotech company with in-house expertise in deep learning, pathology, life sciences, and pharmacotherapeutics. As the country’s first to obtain Korea’s MFDS (Ministry of Food and Drug Safety) approval of an AI-based cancer diagnostic support solution, Deep Bio envisions a suite of AI-based IVD SaMDs (In Vitro Diagnostics Software as a Medical Device) for diagnosis and prognosis of multiple cancers. Deep Bio is actively engaged in the research space and participating in ongoing collaborations with top US medical centers. To learn more, visit

    DeepDx® Prostate is a clinically-validated AI for prostate core needle biopsy tissue image analysis. Whole-slide images (WSIs) of H&E-stained biopsy tissue specimens are analyzed for prostate cancer, Gleason scores and grade groups. Extensively tested at US CLIA labs (> 500k cores in 2021), DeepDx® Prostate can alleviate the shortage of pathologists and the resultant increase in workload, while reducing diagnostic subjectivity and variability.” To learn more, visit