Funded under: POR-CreO
Computed Tomography To Nuclear Medicine
From 2018-02-01 duration 18 months
CompT-NM aims at developing and validating a diagnostic platform able to integrate nuclear medicine sensor (PET - Positron Emission Tomography or SPECT Single Photon Emission Computed Tomography) and CBCT (Cone Beam Computed Tomography) technology in a compact and portable device. This device will be used in clinical ambit and in pre-clinic experimentation for the anatomic-functional study of the brain.
In details, the goal is to realize a high performance device for the study of neurodegenerative (such as Alzheimer and Parkinson) and cerebrovascular (such as ischemic stroke) diseases that are constantly growing. This hybrid technology is yet employed in full-body oncologic diagnostic tool which only 5% of them are dedicated to neuro-vascular pathology. A dedicated diagnostic device will decrease costs, doses, waiting list and will be easy to use.
Recently, the research among functional imaging sensors has produced innovative solutions both in PET and SPECT sensors. In particular, new configuration allow photomultiplier removal decreasing complexity, fragility, weight and dimensions, making possible the development of a reliable and portable device.
The proposed configuration will structurally and functionally combine a nuclear imaging system with a reduced diameter CBCT. Thanks to this, the proposed device could be realized at relatively low cost and will be reliable, easy to use and will offer a higher detailed image. The integration between anatomic and functional imaging determines a performance jump because merges information coming from both technologiesc providing an immediate and complete view of the acquired data.
Project operative goals are:
- Technological state of the art
- Physical-virtual development platform
- System inegration
- Prototyping and validation
- Project Management
The development will follow the Lean Startup paradigm combined with Agile Development techniques. The realization of a development platform composed by a virtual and a physical simulator will speed up system behavior learning. Such platform will map system behavior in the foreseen use cases and will validate them with theoretical methods.
An artificial intelligence network will be developed. Such network will use Deep Learning algorithm basing on the data produced by a set of simulators to remotely monitor the developed devices providing also functionalities of predictive maintenance.