SMARTAI : Objecting finding and classification
SMARTAI is a deep learning-based analysis software technology that can present high sensitivity and accuracy by learning the diagnostic results of the Lateral Flow assay through position analysis. This technology is designed as a two-step CNN model algorithm for object detection and classification, and various data training and data augmentation are performed to improve performance, and the algorithm is optimized so that it is not affected by the surrounding include cases such as indoors/outdoors, lighting conditions, and shade/sunlight.
SMARTAI-LFA can find a kit in a smartphone image through YOLOv3, acquire a test line area containing information about positive/negative results, and transfer the acquired image to ResNet-18 to obtain positive/negative results.
One of the excellent features of SMARTAI-LFA is detecting only the test line. and it minimizes the variation of results due to color and brightness and the environment by minimizing the variation of results due to color and brightness and the environment by repeated training and provides accurate results in a short time of less than 15 seconds.
This allows users to use diagnostic reagents with high accuracy and consistent results.
TIMESAVER : Revolutionizing Point-of-Care Testing with Rapid Diagnosis
TIMESAVER technology is at the forefront of innovation in the field of point-of-care testing (POCT). By leveraging deep learning techniques, we have developed an algorithm that analyzes color changes in rapid kits and predicts the results at an early stage of color change. The core deep learning algorithm is based on an architecture composed of YOLO, CNN-LSTM, and fully connected (FC) layers. Combined with smart AI-based verification, TIMESAVER technology enables a significant reduction in diagnostic time.
