PersImage
Connecting Qunatitative Imaging and in vitro Data Towards Personalized Medicine
Diseases such as cancer or neurodegeneration exhibit a rather complex and multilayer interplay of physiological and pathophysiological processes. A precise diagnosis, and a roadmap towards personalized medicine, however, depends on the collection, interpretation, fusion and deconvolution of multiple parameters which are an assembly of in vitro blood parameters, biopsy probes, physiological parameters, omics analysis, and in vivo images. In particular, for diagnosing complex diseases, non-invasive imaging has several favourable advantages over other diagnostic methods.
The combination of anatomical (magnetic resonance tomography (MRT) or computer tomography (CT)) and molecular imaging modalities (positron emission tomography (PET)) such as PET/CT or PET/MRI along with target-specific radiolabeled biomarkers (“tracers”) provide a powerful tool for exact spatial mapping and in vivo profiling of diseases. Multimodality imaging allows to temporarily assessing changes in disease progression or treatment success. Most important, multiparametric in vivo imaging modalities provide holistic information on the specific location, the molecular profile, and the spread of the disease. Obviously, such holistic, target-specific, and comprehensive information cannot be achieved by biopsy and pathology or by blood-derived molecular biomarkers. With the advent of novel functional and molecular imaging technologies such as functional MRI (fMRI), PET or (fluorescence) endoscopy, as well as target-specific imaging ligands (e.g. radiolabeled PET tracers), imaging offers -beyond high resolution morphology- quantitative in vivo parameters of metabolism, receptor expression, oxygenation or perfusion.
However, currently the information from medical imaging for clinical diagnosis is mainly based on visual alterations of organs (RECIST). Important functional and molecular information is still often disregarded and considered as too complex. The gap between the visualized (images) and the numerical information, that modern molecular and functional imaging modalities provide, is a major challenge to report relevant information, to track them over time, as well as to correlate these essential imaging (visual or numerical) information with other clinical parameters of the same patient, accessed through proteomics, metabolomics, genomics, pathology, blood biomarkers or individual physiology.
In this project the imaging protocols will be developed in the realm of oncology, focusing on gliomas, prostate tumors, lung tumors and rectum cancer.
Main Objectives
- Development of Standardized Preclinical and Clinical Imaging Protocols and SOPs
- Preclinical PET/MRI with a focus on clinical translation
- Clinical MR, CT, PET/CT, PET/MRI and endoscopy
- Specific endoscopy project (rectum cancer)
- Development of image texture analysis, extraction and segmentation methods
- Development of a dedicated workflow and database for segmented images and numerical imaging data (machine learning & texture analysis)