Periodic Reporting for period 3 - ResolveStroke (Stroke diagnostic imaging performed with ultrafast ultrasound localization microscopy (uULM))
Berichtszeitraum: 2021-09-01 bis 2023-02-28
The ERC ResolveStroke project attempts to create a new imaging technique that can map all the blood vessels within the adult human brain (brain angiography). It exploits an innovative approach called ultrasound localization microscopy (ULM), which tracks micrometric contrast agents traveling among red blood cells. As it relies on ultrasound, we can envision such a technique to be implemented in a small portable imaging scanner, which could have a major impact on the treatment of strokes.
Strokes represent one of the main causes of death across the world. It is also the primary cause of handicaps. This disease can be caused by a blockage of a blood vessel within the brain (ischemic stroke) or a rupture of a blood vessel (hemorrhagic stroke). Stroke needs to be treated within a few hours to lessen its impact. Unfortunately, treatment can only be provided after diagnostic is performed imaging - with MRI and CT currently - to distinguish ischemic and hemorrhagic stroke. Unfortunately, access to these imaging tools is limited worldwide and, due to this situation, the great majority of stroke patients remain untreated.
The project ResolveStroke aims to transform ULM into an emergency diagnostic tool that can be brought to the patient rapidly, removing the current hurdle preventing rapid treatment. Its objectives are 1) convert current 2D ULM into a full 3D ultrasound angiography 2) image through the animal model and human skull 3) Provide an angiography at the micrometric scale 4) Define what distinguishes ischemic and hemorrhagic stroke on ULM images 5) Demonstrate its use in models and humans.
Strokes represent one of the main causes of death across the world. It is also the primary cause of handicaps. This disease can be caused by a blockage of a blood vessel within the brain (ischemic stroke) or a rupture of a blood vessel (hemorrhagic stroke). Stroke needs to be treated within a few hours to lessen its impact. Unfortunately, treatment can only be provided after diagnostic is performed imaging - with MRI and CT currently - to distinguish ischemic and hemorrhagic stroke. Unfortunately, access to these imaging tools is limited worldwide and, due to this situation, the great majority of stroke patients remain untreated.
The project ResolveStroke aims to transform ULM into an emergency diagnostic tool that can be brought to the patient rapidly, removing the current hurdle preventing rapid treatment. Its objectives are 1) convert current 2D ULM into a full 3D ultrasound angiography 2) image through the animal model and human skull 3) Provide an angiography at the micrometric scale 4) Define what distinguishes ischemic and hemorrhagic stroke on ULM images 5) Demonstrate its use in models and humans.
The project ERC ResolveStroke was oriented toward the conception of two ULM angiographs, one for animal models and another for humans. Each needed several scientific and technological breakthroughs performed in a developmental phase, along with a second phase designed to demonstrate its use. Four Ph.D. students are currently working on the subject, along with a postdoctoral fellow.
The 3D ULM scanner was conceived with a programmable scanner connected to matrix array transducers. One is centered around a 10 MHz working frequency and can be used for smaller animals, the second being centered around a 1.5 MHz working frequency. Sequences and algorithms were conceived to image and track microbubbles within artificial models, along with small animal models.
We have already published an article where 2D ULM has been exploited to characterize stroke in an animal model [Hingot, Brodin et al. Theranostics 2020]. It demonstrated that ischemia could be rapidly identified with ULM. However, its planar character impedes the capacity to quantify blood flow and its modifications linked to stroke. Moreover, long acquisition and plane selection imply that we already know what we are trying to observe before the start of the imaging sequence. 3D ULM is bound to resolve these issues.
The 3D ULM system for the small animal has, to date, demonstrated its capacity at providing a brain angiography through the skull of the animal model. These results have been presented in conferences [Chavignon et al. IUS 2020, Rotterdam Contrast Symposium 2021]. The 3D ULM system for human remains in development, but new sequences have been implemented to resolve the issues implicated in divergent ultrafast imaging through the human skull.
We are also attempting to open the field of ultrasound localization microscopy worldwide. O.Couture piloted a review article with many researchers across the world to establish standards [Christensen-Jeffries et al. 2020]. Moreover, we are currently submitting an article that comprises large datasets, user-friendly code, and standardization metrics to help other laboratories in implementing and characterizing ULM.
The 3D ULM scanner was conceived with a programmable scanner connected to matrix array transducers. One is centered around a 10 MHz working frequency and can be used for smaller animals, the second being centered around a 1.5 MHz working frequency. Sequences and algorithms were conceived to image and track microbubbles within artificial models, along with small animal models.
We have already published an article where 2D ULM has been exploited to characterize stroke in an animal model [Hingot, Brodin et al. Theranostics 2020]. It demonstrated that ischemia could be rapidly identified with ULM. However, its planar character impedes the capacity to quantify blood flow and its modifications linked to stroke. Moreover, long acquisition and plane selection imply that we already know what we are trying to observe before the start of the imaging sequence. 3D ULM is bound to resolve these issues.
The 3D ULM system for the small animal has, to date, demonstrated its capacity at providing a brain angiography through the skull of the animal model. These results have been presented in conferences [Chavignon et al. IUS 2020, Rotterdam Contrast Symposium 2021]. The 3D ULM system for human remains in development, but new sequences have been implemented to resolve the issues implicated in divergent ultrafast imaging through the human skull.
We are also attempting to open the field of ultrasound localization microscopy worldwide. O.Couture piloted a review article with many researchers across the world to establish standards [Christensen-Jeffries et al. 2020]. Moreover, we are currently submitting an article that comprises large datasets, user-friendly code, and standardization metrics to help other laboratories in implementing and characterizing ULM.
Before the start of the project ResolveStroke, ULM was a fairly recent and limited approach. We had demonstrated its use on models’ brain but following the removal of its skull [Errico et al. Nature 2015]. The technique was entirely in 2-dimensions, which highly limited its potential use as a diagnostic tool. Indeed, within a single plane of imaging, it is impossible to quantify blood flow, correct 3D motion, avoid projection artefacts and alleviate user-dependency. Moreover, covering the entire brain vasculature took hours of imaging since every planar acquisition took several minutes.
Thank to the ERC ResolveStroke, we are currently equipped with an ULM system that can perform a 3D angiogram of the majority of the model’s brain within minutes. The approach is fully transcranial and can be performed longitudinally. It provides a precise mapping of the vasculature at the micrometric scale down to a depth of about 1 cm. Ischemic and hemorrhagic stroke can be distinguished on these images. However, we are currently trying to define the precise imaging biomarkers that could provide staging during diagnostic imaging.
The 3DULM angiograph for human is obviously not as advanced as the animal model scanner. Indeed, such apparatus needs to traverse a thicker skull and pass all the safety regulations related to human applications. Currently, our apparatus was shown to map in 3D artificial blood vessel through an artificial human skull down to a depth of 14 cm. Although its resolution is worse than its small animal counterpart, it is still far superior than conventional ultrasound imaging. The next step is to study its capacity within a larger animal before performing the necessary safety test and ethical examinations that will open the way to human applications as the end stage of the ERC ResolveStroke.
Thank to the ERC ResolveStroke, we are currently equipped with an ULM system that can perform a 3D angiogram of the majority of the model’s brain within minutes. The approach is fully transcranial and can be performed longitudinally. It provides a precise mapping of the vasculature at the micrometric scale down to a depth of about 1 cm. Ischemic and hemorrhagic stroke can be distinguished on these images. However, we are currently trying to define the precise imaging biomarkers that could provide staging during diagnostic imaging.
The 3DULM angiograph for human is obviously not as advanced as the animal model scanner. Indeed, such apparatus needs to traverse a thicker skull and pass all the safety regulations related to human applications. Currently, our apparatus was shown to map in 3D artificial blood vessel through an artificial human skull down to a depth of 14 cm. Although its resolution is worse than its small animal counterpart, it is still far superior than conventional ultrasound imaging. The next step is to study its capacity within a larger animal before performing the necessary safety test and ethical examinations that will open the way to human applications as the end stage of the ERC ResolveStroke.