Computer Aided Surgery
Contactless Interface Software
Objective : Self Training & Prediction for contactless gesture recognition in Operating System
Development of hand gestures recognition using Machine learning for contactless interface system
Gesture : Hover, grab, click, one peak, and two peaks.
SW : Window OS, 64bit, WT PACS, 3DPACS, Common IMAGE Viewer
Sensor Device : Single Sensor or Multiple sensor Kinect, Leap Motion and MYO
Application Device : Amis-50(clinical device PC)
Published paper : Computer Methods and Programs in Biomedicine, Volume 161, July 2018, Pages 39-44
Deep learning based Chest X-ray-CAD system
Chest X-ray CAD (Computer aided diagnosis) system
For faster reading of radiologists, AI algorithm with deep learning diagnose X-rays in advance with attention
map. This SW automatically generate radiologic reports including lesion size, location, classification of possible
diseases, and attention map for indicating the diseases.
Pulmonary lesions : nodule, consolidation(mass), interstitial opacity, pleural effusion, pneumothorax
Algorithm : Convolutional neural net (CNN) with weakly supervision
Overview : lesion size, location, classification, and measurements
ENT surgery video
Monitoring simple mastoidectomy video
To develop monitoring and mentoring SW for physician with training procedures of simple mastoidectomy video with devices detection and anatomic landmark segmentation
Algorithm : CNN, LSTM, YOLO, U-Net, etc
Overview : scene prediction, device detection & landmark segmentation