The Department of Biomedical Informatics
presents
Tianye Niu, Ph.D.
Dr. Tianye Niu is the Director of the Medical Engineering Laboratory (MEL) of the Translational Medicine Institute of Zhejiang University, and Principal Investigator for the Natural Science Foundation of China and Ministry of Science and Technology of China. Dr. Niu’s research interests include cone-beam CT imaging, radiomics and medical image analysis. He has a number of active collaborations both nationally and internationally.
“Advances in cone-beam CT Imaging: Quantitative and Low-dose Schemes”
Monday June 08, 2015
1:00 pm – 2:00 pm
Biomedical Informatics Conference Room
HSC, Level 3, Room 3-045 A&B
Contact the Department of Biomedical Informatics at (631) 444-8459 with any questions regarding this event.
Abstract
X-ray cone-beam CT (CBCT) system is becoming an indispensable modality in the image guidance for radiation therapy and other clinical procedures. Major components of a CBCT scanner include x-ray tube and large-area flat panel detector. The system is designed as open-gantry geometry to facilitate its incorporation into treatment process. The hardware configuration of CBCT enables its on-board capability to detect the two-dimensional x-ray projections passing through a large illuminated volume of an object, and then to reconstruct three-dimensional volumetric images. Thus CBCT system acquires accurate patient geometry during treatment and is applied in routine clinical procedures, such as the patient setup in radiation therapy. Can CBCT be applied in advanced clinical applications, e.g., diagnosis or treatment evaluation? Current performance of CBCT are hindered by several major bottlenecks, including severe scatter and beamhardening effects, mono-energetic x-ray spectrum and high accumulated dose in repeated scans. Quantitative and safe CBCT imaging is on urgent demand for its advanced applications. In this presentation, quantitative CBCT schemes will be introduced using effective shading correction and dual-energy material decomposition. Shading correction method suppresses the severe artifacts in the CBCT images, and reduces the CT number error around 30 HU. Dual-energy imaging method is proposed for material decomposition in low-dose CBCT images and reduces the noise variation by 95%. Based on the compressed sensing theory, a low-dose iterative algorithm is developed to reconstruct CT images of high quality using no more than 30% of projections in a conventional scan to achieve the safe CBCT imaging. Future work will combine the schemes, and achieve the practical quantitative and safe CBCT imaging. These work have the great potential to facilitate the advanced use of CBCT and promote the clinical outcomes.