zhaoxuanma@yahoo.com
zhaoxuanma@yahoo.com
Zhaoxuan Ma is working as a senior principle algorithm engineer at Medtronic plc. His main focus is on medical image analysis and robotic control system. Besides, he develops apps and builds websites.
Robot arm control system design and related alogrithm development including but not limited to path planning, collision detection and system calibrition.
Design and implementation of state-of-the-art deep learning architecture on object detection, image classification and semantic segmentation.
End-to-end life cycle of IOS application development from concept to release and maintenance.
Immunohistochemistry (IHC) and fluorescent slide scanning. Cell and tissue based quantification like but not limit to immune infiltration, vessel and gland segmentation, lesion detection.
Develop web application for both front- and back-end including database interaction, authentication, UI design and server-side API development.
1. Robot guidance system
Development of robot control system and related algorithm including but not limited to robot-workstation communication service, path planning, collision detection and system calibration.
Vertebrae semantic segmentation algorithm development on 3D (CT/3D-carm) and 2D (X-ray) image data.
2. Surgical navigation system
Software and algorithm development of optical surgical navigation system.
1. Cervical cancer screening project
Cervical cancer algrithm development based on TBS system, including low-resolution abnormal cell detection; high-resolution multiple class classification; whole slide grading; cell count; slide quality evaluation.
Algorithm (with device partner) has been issued registration certificates for medical devices of class II; instrument along with algorithm deployed has been used in hundreds of medical center; beat or tie with clinical cervical cancer pathologist in "man vs. machine battle".
2. Smart microscope system project
Development of microscope control system in Android app and Wechat mini-program which can control smart microscope remotely and display video streaming.
Other local system function includes: OCR system for screening slide label; video streaming that supports telemedicine; multi-end live annotation and chat system; local and cloud data management system.
Development of machine learning tools to complement Gleason grading of prostate cancer.
Development of a deep learning pipeline to quantitate immune infiltrate in inflammatorybowel disease specimens, and colon gland histomorphometry.
Development of a deep learning approach to assess tumor growth patterns inadenocarcinoma of the lung.
Development of a novel 6-antibody staining protocol to quantitate immune infiltrates inhuman cancers.
Imaging and quantification of immunohistochemically stained tissues and cells.
Neural signal data processing and analysis software development.
System level (C-arm) image quality assessment, analysis and improvement on various image algorithms.
