Summary
Radiomics - the extraction of clinically relevant information from complex medical imaging data via mathematics and data science - is on the verge of becoming a main player in clinical research and medicine. However, the current radiomics workflow mostly relies on feature engineering and black box machine learning, lagging behind the state-of-the-art in explainable artificial intelligence. I will exploit my broad expertise in mathematics, informatics, and biomedical imaging to implement intelligent radiomics by integrating the whole imaging value chain - ranging from imaging hardware, over image formation, to medical interpretation of the data - into an intelligent software environment. EchoLux is a paradigm shift from black box machine learning to truly intelligent radiomics.
EchoLux will be realized in three steps: 1) Modelling the imaging system via a digital twin approach using dedicated physical phantom data and a reference dataset acquired from healthy volunteers; 2) Development and integration of a medical knowledge base that captures the effects of disease on the imaged tissue; 3) Integration of system and medical tissue models into a Bayesian reasoning framework that is able to perform explainable diagnostics and medical knowledge discovery.
The clinical use case is optoacoustic imaging of peripheral neuropathy. I recently demonstrated that the internal structure and vascular supply of peripheral nerves can be visualized in unprecedented detail by exploiting the molecular contrast of optoacoustic imaging. The EchoLux approach, thus, has the potential to enable early detection of pathological changes in peripheral nerves, e.g., in conjunction with diabetes.
EchoLux links the clinical imaging data simultaneously back to the hardware and forward to the medical decision making. This holistic and interdisciplinary approach will leverage radiomics in general and boost the clinical value of optoacoustic imaging.
EchoLux will be realized in three steps: 1) Modelling the imaging system via a digital twin approach using dedicated physical phantom data and a reference dataset acquired from healthy volunteers; 2) Development and integration of a medical knowledge base that captures the effects of disease on the imaged tissue; 3) Integration of system and medical tissue models into a Bayesian reasoning framework that is able to perform explainable diagnostics and medical knowledge discovery.
The clinical use case is optoacoustic imaging of peripheral neuropathy. I recently demonstrated that the internal structure and vascular supply of peripheral nerves can be visualized in unprecedented detail by exploiting the molecular contrast of optoacoustic imaging. The EchoLux approach, thus, has the potential to enable early detection of pathological changes in peripheral nerves, e.g., in conjunction with diabetes.
EchoLux links the clinical imaging data simultaneously back to the hardware and forward to the medical decision making. This holistic and interdisciplinary approach will leverage radiomics in general and boost the clinical value of optoacoustic imaging.
Unfold all
/
Fold all
More information & hyperlinks
| Web resources: | https://cordis.europa.eu/project/id/101041936 |
| Start date: | 01-07-2022 |
| End date: | 30-06-2027 |
| Total budget - Public funding: | 1 499 976,00 Euro - 1 499 976,00 Euro |
Cordis data
Original description
Radiomics - the extraction of clinically relevant information from complex medical imaging data via mathematics and data science - is on the verge of becoming a main player in clinical research and medicine. However, the current radiomics workflow mostly relies on feature engineering and black box machine learning, lagging behind the state-of-the-art in explainable artificial intelligence. I will exploit my broad expertise in mathematics, informatics, and biomedical imaging to implement intelligent radiomics by integrating the whole imaging value chain - ranging from imaging hardware, over image formation, to medical interpretation of the data - into an intelligent software environment. EchoLux is a paradigm shift from black box machine learning to truly intelligent radiomics.EchoLux will be realized in three steps: 1) Modelling the imaging system via a digital twin approach using dedicated physical phantom data and a reference dataset acquired from healthy volunteers; 2) Development and integration of a medical knowledge base that captures the effects of disease on the imaged tissue; 3) Integration of system and medical tissue models into a Bayesian reasoning framework that is able to perform explainable diagnostics and medical knowledge discovery.
The clinical use case is optoacoustic imaging of peripheral neuropathy. I recently demonstrated that the internal structure and vascular supply of peripheral nerves can be visualized in unprecedented detail by exploiting the molecular contrast of optoacoustic imaging. The EchoLux approach, thus, has the potential to enable early detection of pathological changes in peripheral nerves, e.g., in conjunction with diabetes.
EchoLux links the clinical imaging data simultaneously back to the hardware and forward to the medical decision making. This holistic and interdisciplinary approach will leverage radiomics in general and boost the clinical value of optoacoustic imaging.
Status
SIGNEDCall topic
ERC-2021-STGUpdate Date
09-02-2023
Images
No images available.
Geographical location(s)
