Development of a hardware-based registration system for the multimodal medical images by USB cameras

Michiaki Iwata, Hiroshi Watabe, Kazuhiro Koshino, Akihide Yamamoto, Kotaro Minato, Hidehiro Iida

Research output: Contribution to journalArticlepeer-review

Abstract

There are severalmedicalimaging scanners and each modality has different aspect for visualizing inside of human body. By combining these images, diagnostic accuracy could be improved, and therefore, several attempts for multimodal image registration have been implemented. One popular approach is to use hybrid image scanners such as PET/CT and SPECT/CT. However, these hybrid scanners are expensive and not fully available. We developed multimodal image registration system with USB cameras, which is inexpensive and applicable to any combinations of existed conventional imaging scanners. The multiple USB cameras will determine the three dimensional positions of a patient while scanning. Using information of these positions and rigid body transformation, the acquired image is registered to the common coordinate which is shared with another scanner. For each scanner, reference marker is attached on gantry of the scanner. For observing the reference markers position by the USB cameras, the location of the USB cameras can be arbitrary. In order to validate the system, we scanned a cardiac phantom with different positions by PET and MRI scanners. Using this system, images from PET and MRI were visually aligned, and good correlations between PET and MRI images were obtained after the registration. The results suggest this system can be inexpensively used for multimodal image registrations.

Original languageEnglish
Pages (from-to)574-581
Number of pages8
JournalTransactions of Japanese Society for Medical and Biological Engineering
Volume47
Issue number6
Publication statusPublished - 2009 Dec 1
Externally publishedYes

Keywords

  • Image registration
  • Multimodality
  • USB camera

ASJC Scopus subject areas

  • Biomedical Engineering

Fingerprint Dive into the research topics of 'Development of a hardware-based registration system for the multimodal medical images by USB cameras'. Together they form a unique fingerprint.

Cite this