Our first product line, DeepDx-Prostate, uses deep learning to recognize and/or localize acinar adenocarcinoma of the prostate in scanned, H&E-stained histopathology slides. The deep learning model used for slide image analysis is trained using a myriad of slides containing highly varied patterns of benign and malignant tissue, making it robust to variations introduced during slide preparation and scanning.
DeepDx-Prostate Basic is a slide viewer and a binary screening solution that recognizes cancer in slides, allowing pathologists to concentrate on cases with known malignancy.
It enables pathologists to process more cases in less time and reduces the risk of misdiagnosis due to mental fatigue accumulated through long hours spent pouring over the eyepiece of a traditional microscope.
DeepDx-Prostate Pro distinguishes cancer severity, coloring tissue regions based on Gleason pattern classification results and suggests a Gleason score. Users can also consult AI on hand-picked regions to obtain quick feedback on slides with large areas of tissue.
Pre-filled pathology reports are generated once slides are analyzed using the AI from DeepDx-Prostate. Clinically significant information included in the pathology report includes:
- Gleason score
- Relative proportions of each Gleason pattern
- Percentage involvement of cancer in tissue
- Image samples representative of the cancer found in the slide
DeepDx-Prostate Connect is DeepDx-Prostate put on the web. It has the same powerful AI and functionalities from DeepDx-Prostate Pro and allows the controlled sharing of slides, discussions, and annotations amongst medical experts, by streamlining communication and lowering barriers that hinder cooperation.
DeepDx-Prostate Connect is also supported on portable devices so that users can contribute to the case regardless of their physical location.