Functional Brain Image Understanding for Differential Diagnosis of Dementia

Summary

Intelligent analysis of functional brain images and extraction of pathologic patterns from large-group study for early diagnosis of dementia types

Supervisor(s)

Professor David Feng, Dr Yong Xia

Research Location

Computer Science

Program Type

Masters/PHD

Synopsis

With the increasing ageing of the population, dementia has become an important world-wide public health problem. Even though no definitive cure has been found for this disease, early diagnosis is useful for choosing appropriate management and developing new treatment, which can effectively slow-down the progression of dementia. Functional imaging techniques of positron emission tomography (PET) is able to detect subtle changes in cerebral metabolism before there are changes on anatomical imaging and before a clinical diagnosis of probable dementia can be made. This ability gives PET a distinct advantage over anatomical imaging techniques in the evaluation of neurodegenerative disorders like dementia at the early stages of the disease process. However, since the functional changes in the early course of dementia can be subtle, there are some overlapped patterns in images of different dementia types and lacks of clearly defined boundaries between dementia patients and normal aged controls. This project aims to develop a functional brain image understanding system, with the identification and extraction of intrinsic pathologic patterns via analysing a large group of brain PET studies, to provide discriminatory classification of dementia types, and to offer computer-aided early diagnosis of dementia through case-based reasoning or evidence-based medicine.

Keywords

Functional brain imaging, dementia, image analysis, Image understanding, pattern classification

Opportunity ID

The opportunity ID for this research opportunity is: 1195

Other opportunities with Professor David Feng

• Advanced computer modelling of biological systems using insight knowledge
• Discovery of new image-derived features for computer-aided diagnosis
• Automated 3-Dimensional Biomedical Registration for Whole-body Images from Combined PET/CT Scanners - Automatic Registration for 3D Whole-body Images from Combined PET/CT Scanners
• Deformable Registration for Temporal Lung Volumes
• Kinetic Characterization and Mapping for Whole-body Molecular Image Retrieval
• Multi-dimensional Biomedical Data Visualization
• Image Representation using Multi-dimensional Biomedical Functional and Anatomical Features
• Automatic Image Content Annotation
• Web Image Annotation
• Data Management for Automated Identification and Classification of Plant Images
• Novel Image Retouching Techniques
• Automatic Video Content Annotation
• Intelligent Access to Digital TV Content
• Semantic Multimedia Information Retrieval
• Multimedia Streaming with Peer-to-Peer Techniques
• Intelligent Multimodality Molecular Image Segmentation
• Multimodality Medical Image Segmentation
• Medical Image Mining for Computer-Aided Diagnosis
• Disease Map - Big Data driven modelling and derivation of diseases and treatment response
• Machine learning in Multiscale Image-Omics
• Content-based Retrieval and Management of Multi-dimensional Biomedical Imaging Data
• Object-based Volumetric Texture Feature Extraction for Biomedical Image Retrieval
• Semantic-driven Multi-modal Biomedical Data Visualisation
• Computerised image analysis of musculoskeletal diagnostics and surgical planning

Other opportunities with Dr Yong Xia

• Multimodality Medical Image Segmentation
• Medical Image Mining for Computer-Aided Diagnosis