×
  • Select the area you would like to search.
  • ACTIVE INVESTIGATIONS Search for current projects using the investigator's name, institution, or keywords.
  • EXPERTS KNOWLEDGE BASE Enter keywords to search a list of questions and answers received and processed by the ADNI team.
  • ADNI PDFS Search any ADNI publication pdf by author, keyword, or PMID. Use an asterisk only to view all pdfs.
Principal Investigator  
Principal Investigator's Name: Manuel Jorge Cardoso
Institution: KCL
Department: Biomedical Engineering
Country:
Proposed Analysis: A major problem in medical imaging research is data scarcity and availability. Governance, privacy concerns, and cost-of-acquisition all restrict access to medical imaging data, which, compounded by the data-hungry nature of deep learning algorithms, limits progress in the field. Recently, generative models have been used to synthesise photorealistic natural images, providing a potential solution to data scarcity. We have developed a 3D conditioned generative model of the human brain, trained at the necessary scale to generate diverse, realistic-looking, morphologically correct, high-resolution samples, which can be conditioned on patient characteristics (e.g. age and pathology). We show that the samples preserve the biological and disease phenotype and are realistic enough to be used with well-established image analysis tools. We would now like to use the ADNI dataset to show that such tools can also learn to generate synthetic data in a way that preserves the longitudinal progression of disease. The analysis will comprise training the model with ADNI data, using different subgroups and disease/age/gender conditionings, and demonstrate that the synthetic data preserves the morphological characteristics of the underlying dataset (e.g. the VBMs look similar, the regional volumes have the same distribution, etc).
Additional Investigators