Magnetic Resonance Imaging (MRI) is an image modality highly susceptible to performance degradation due to subject motion. Artefacts caused by patient movements are a critical problem in many clinical and research MRI applications (involuntary subject's movement, no-collaborative patients affected by disease with tremor, fetal MRI). The result is a loss of image resolution, which prevents early diagnosis and increases the costs and time of healthcare due to repeated examinations.
This project proposes a novel high speed real-time prospective motion correction technique based on modern Deep Learning (DL) methods which provide a potential avenue for dramatically reducing the computation time and improving the convergence of retrospective motion correction overcoming the limits of the current state-of-the-art prospective methods.
Exploiting cutting-edge developments from the field of experimental high energy physics that allow to run sophisticated algorithms for reconstruction and inference of complex data at speeds much faster than the typical acquisition time of MRI sequences, and the know-how in developing and training innovative deep neural network (DNN) models in both foundamental and applied physics in medical imaging, the technique we propose has the potential to outclass current state-of-the-art motion correction methods with ground-breaking advancement in the clinical diagnostic based on MRI.
Innovative DNNs for motion correction in MRI will be developed using publicly available datasets. To fulfil the tight requirements of high accuracy and low latency required by the motion correction task, we will implement DNN compressione and simplification methods developed in high energy physics. The models will be then deployed into real-time co-processors, such as Field Programmable Gate Arrays (FPGAs) and GPUs, and performances with respect the motion-correction task will be evaluated.
The success of the proposed project will generate impact at multiple levels and scales.
First and foremost, a strong impact of the project is expected in the health field, with dramatic improvement in medical diagnostics and a consequent huge benefit for the citizens. Being able to have an MRI characterized by better image quality and spatial resolution for a high level of anatomical detail means being able to notice the smallest microstructural variations in tissues that occur due to the development of pathologies thus making early diagnosis. This opens up the possibility of acting promptly with early therapies, increasing the possibility of both healing and of more effectively counteracting the development of a disease. Due to its non ionizing property MRI with improved resolution will be particularly useful in clinical trials to test new drugs effectiveness and new therapeutic treatments. Therefore, in our opinion the project could also have an impact on the pharmaceutical industries as well as a significant impact on the biomedical imaging market.
We are confident that our project, which aims to correct in real time the effects of movements in MR images and therefore increase their quality and resolution, will be able to develop fetal MRI diagnostics. Better image quality and resolution will lead to early diagnosis. With continuous technological development, in utero surgery is now a reality. Fetal surgery greatly helps to improve the long-term outcomes of babies with specific birth defects. In our opinion, the early and more precise and sensitive fetal diagnosis that will occur thanks to the real-time correction of fetal movements during MRI acquisitions will help to further develop surgical and therapeutic strategies that can be achieved in the uterus, contributing to the improvement of human health and well-being.
A positive result of the project will also affect the total time of diagnostic MRI exams, eliminating the problem of repeating the exam due to poor image quality related to patient's movement. The costs of repeating scans, rescheduling appointments, and sedating patients are high. A recent study estimates the annual operational cost impact per scanner at about hundred thousand dollars. In other words, the project impact will be to reduce scans and increase productivity. A reduction in waiting times and medical costs together with an increase in earnings in private diagnostic centers due to a more numerous daily access of patients is expected. This condition will be particularly decisive for the development of fetal MRI, which to date has shown great potential but is dramatically affected by either the problem of several times repeating MRI exams due to fetal movements or the inability to apply highly sensitive tensor diffusion diagnostics protocols. In general, fetal MRI is characterized by a lower resolution, just when this would be more necessary for a correct prenatal diagnosis due to the smaller size of the organs and tissues compared to those of an adult.
The impact of the project goes beyond the field of medical imaging alone, opening up possible applications in other sectors of applied and basic science, and industrial technology. Exploitable Technologies (ET) are a possible outcome of the project, and may result in different technological transfer actions that can be evaluated and promoted with the Technological Transfer Office of the Sapienza Università di Roma, including open sourcing or patenting the identified ET, and contacting industrial stakeholders to investigate their interest in collaborative research to further develop that technology.
Beside the trivial identification of the proposed motion-correction system for MRI based on novel artificial intelligence techniques as the main ET, the development of the project will encompass the implementation of technologies and methodologies that could likely be exploited either in the fields of other medical imaging and treatment (e.g. real-time dose monitoring in hadron and radiotherapy) or in completely different areas for applications where the real-time AI-based processing of streams of data is needed to perform control or analytics tasks. These include, for example, banking and financial services, surveillance systems, cyber security, manufacturing control and automotive.