Radiology and Bioimaging, Neuroscience

Radiological Imaging and Informatics

Revelation of Fundamental Mechanisms and Origin of Intelligence in Neural Information Processing by Biomedical Imaging and Computational Intelligence Techniques

Health Sciences Course

  • Master / Doctoral Degree

Faculty

HOMMA, NoriyasuHOMMA, Noriyasu
HOMMA, Noriyasu

Professor, Ph.D.

  • TEL

    +81-22-717-7914

  • Mail

    secretary*rii.med.tohoku.ac.jp (Please convert "*" into "@".)

Research Theme

  • Revelation of the information processing mechanisms in the neuronal circuit of cortico-basal ganglia loop
  • Development of highly intelligent medical instruments
  • Improvement of X-ray CT image quality
Research Keywords:

basal ganglia, Parkinson’s disease, calcium, image-guided radiation therapy, computer-aided diagnosis

Technical Keywords:

calcium imaging, MRI, X-ray CT, computational intelligence, machine learning

Laboratory Introduction

By using multi-modal imaging techniques, we aim to accurately visualize wide range of biomedical phenomena such as microscopic neurophysiological activities as well as macroscopic whole brain ones. The images can then be utilized for not only extracting useful information for medical and life sciences, but also revealing mechanisms of neural information processing and implementing them. For example, the basal ganglia receives the input from cortex and sends the output to cortex, and is thought to play a crucial role in motor control, reward prediction, learning, and memory. The basal ganglia also related to the symptoms of Parkinson’s disease. However, the fundamentals of information processing in the basal ganglia and the pathophysiological mechanisms of Parkinson’s disease are not fully elucidated. Therefore, we conduct the studies for revelation of the information processing mechanisms in the neuronal circuit of cortico-basal ganglia loop, and pathophysiological characteristic of Parkinson’s disease. In addition to the neurophysiological analysis, we also build computational models of neural information processing inspired by highly intellectual skills of human specialists such as radiologists. Based on the models and some machine learning techniques, we develop intelligent computer-aided systems for medical image diagnosis, interventional radiology, and image-guided therapy.

Figure 1. Visualization of the neuronal activity changes by qAIM- MRI

Figure 1. Visualization of the neuronal activity changes by qAIM- MRI

Figure 2. Novel intelligent fluoroscopic imaging: Tumor extraction (right) from the X-ray image (left)

Figure 2. Novel intelligent fluoroscopic imaging: Tumor extraction (right) from the X-ray image (left)

Recent Publications

  • Zhang X, et al, Tracking tumor boundary in MV-EPID images without implanted markers: A feasibility study, Medical Physics, 42(5): 2510-2523, 2015
  • Kikuta S, et al. Quantitative activation-induced manganese-enhanced MRI reveals severity of Parkinson’s disease in mice. Scientific Reports 5: 12800, 2015