Exploring New Solutions for Limitations in Neurology Imaging
Julien Savatovsky, M.D.
Rothschild Foundation Hospital, Paris, France
Rothschild Foundation Hospital, Paris, France
Earlier this year, Dr. Julien Savatovsky, Neuroradiologist and the Deputy Head of Diagnostic Neuroradiology at the Foundation Adolphe de Rothschild Hospital, Paris, France, started to evaluate Canon Medical System’s AI reconstruction algorithms (see article). From his ongoing studies, the results continue to indicate that the technology holds great potential in advancing MR imaging beyond its current limits to overcome some of healthcare’s most pressing challenges. Dr. Savatovsky describes his latest findings with Canon Medical System’s Advanced intelligent Clear-IQ Engine (AiCE) and Precise IQ Engine (PIQE). His most recent focus has been further exploration of the potential of PIQE, as well as some broader studies with AiCE.
Dr. Savatovsky contributed to the development of the latest reconstruction algorithms and has evaluated them on the Canon Vantage Orian 1.5T and Vantage Galan 3T MRI systems that the Neuroradiology Department of the Foundation Adolphe de Rothschild Hospital acquired in 2021.
“Our hospital pioneers care, research, and cooperation on eye and brain diseases and its Neurologists and Neurosurgeons are continually involved in advancing research through the use of cutting-edge imaging techniques. We need both higher quality images and more comprehensive examinations. In addition, sequences that dive deep inside the mechanism of diseases, such as myelin imaging, are becoming available,” he explained. “With these new developments, examination time can be significantly longer.”
“However, as elsewhere, the challenging economic situation along with decreasing reimbursements in France, means that we are required to increase the number of scans that we carry out per day,” he added. “We now scan far more patients per day than 10 years ago, and aim to scan approximately 30 patients per day per MRI device. That means that it needs to perform faster MRI examinations. We also need to run a greater range of sequences, then every sequence needs to be shorter.”
“Our hospital pioneers care, research, and cooperation on eye and brain diseases and its Neurologists and Neurosurgeons are continually involved in advancing research through the use of cutting-edge imaging techniques. We need both higher quality images and more comprehensive examinations. In addition, sequences that dive deep inside the mechanism of diseases, such as myelin imaging, are becoming available,” he explained. “With these new developments, examination time can be significantly longer.”
“However, as elsewhere, the challenging economic situation along with decreasing reimbursements in France, means that we are required to increase the number of scans that we carry out per day,” he added. “We now scan far more patients per day than 10 years ago, and aim to scan approximately 30 patients per day per MRI device. That means that it needs to perform faster MRI examinations. We also need to run a greater range of sequences, then every sequence needs to be shorter.”
Overcoming compromises
Issues in MRI, such as the well-known ‘triangle of compromise’ in imaging between signal-to-noise ratio (SNR), spatial resolution, and acquisition time, limit efforts to improve efficiency. For example, for given settings, adjustment to obtain a better spatial resolution has required compromise through decreased SNR and longer acquisition time.
“We have been stuck with this compromises for many years” said Dr. Savatovsky. “When we first acquired AiCE and PIQE, we wondered if the SNR and contrast-to-noise ratio (CNR) would really be modified, and whether we would be able to go even further by improving the resolution and then the sharpness of the image. We are very enthusiastic about the new Deep Learning Reconstruction (DLR) reconstruction techniques and AiCE, which carries the promise of offering us better image quality.”
“We have been stuck with this compromises for many years” said Dr. Savatovsky. “When we first acquired AiCE and PIQE, we wondered if the SNR and contrast-to-noise ratio (CNR) would really be modified, and whether we would be able to go even further by improving the resolution and then the sharpness of the image. We are very enthusiastic about the new Deep Learning Reconstruction (DLR) reconstruction techniques and AiCE, which carries the promise of offering us better image quality.”
Impact on disease detection
Image quality has a significant impact on detection and diagnosis.
“Image quality is important. We are not just ‘picky’ about having a nice image. It can actually have an impact on whether we see the disease or not,” emphasized Dr. Savatovsky. “When we make images using AiCE, we see additional lesions.”
“Image quality is important. We are not just ‘picky’ about having a nice image. It can actually have an impact on whether we see the disease or not,” emphasized Dr. Savatovsky. “When we make images using AiCE, we see additional lesions.”
Improved sharpness and resolution
Dr. Savatovsky has specifically explored if resolution and sharpness could be improved with PIQE and illustrates the differences found through images that compare classic filters, AiCE and PIQE.
“There is a great improvement of the sharpness of the image without any loss of the SNR, even in faster acquisition sequences,” he said.
“There is a great improvement of the sharpness of the image without any loss of the SNR, even in faster acquisition sequences,” he said.
“In countless examples, even in high resolution sequences, you tend to have a gain at some point by using the PIQE reconstructions.”
“It works for any anatomy,” he continued. “We can see the hippocampus more sharply, for example.”
“It works for any anatomy,” he continued. “We can see the hippocampus more sharply, for example.”
“The lesions are better depicted with PIQE in this Sagittal T2 from a spinal cord case on an MS patient.”
Reduced acquisition time
Another use of PIQE that Dr. Savatovsky has explored is to reduce acquisition time, as well as improving image quality.
“When you alter acquisition spatial resolution, you gain a lot of time. It's one of the main drivers to make the acquisition faster, to actually make a bigger voxel,” he remarked. “However, if you apply now PIQE to the lower spatial resolution sequence you notice that there is not much difference with the regular sequence at high spatial resolution. So, we effectively ‘double tweak’ the triangle of compromise with the same acquisition, both improving SNR and spatial resolution, which is really useful.”
“When you alter acquisition spatial resolution, you gain a lot of time. It's one of the main drivers to make the acquisition faster, to actually make a bigger voxel,” he remarked. “However, if you apply now PIQE to the lower spatial resolution sequence you notice that there is not much difference with the regular sequence at high spatial resolution. So, we effectively ‘double tweak’ the triangle of compromise with the same acquisition, both improving SNR and spatial resolution, which is really useful.”
Integrating PIQE into clinical practice
The PIQE algorithm is currently available on all 2D sequences that the Hospital runs.
“We have integrated PIQE into all of our 2D acquisitions in our clinical practice, even the Dixon sequences,” remarked Dr. Savatovsky.
“We have PIQE in 2D and we wanted to know if it could be applicable in 3D sequences. This is currently a ‘work in progress’, but the preliminary results are promising in certain applications.”
“For example, I am reconstructing my 3D T2 scans with PIQE at 1.5 Tesla for vestibular schwannoma explorations. When PIQE is applied, the image is sharper and it's very useful to interpret the cranial nerves, or even the fine structure within the inner ear, such as the basilar membrane.”
“PIQE makes the image sharper without any loss of SNR,” he added.
“We have integrated PIQE into all of our 2D acquisitions in our clinical practice, even the Dixon sequences,” remarked Dr. Savatovsky.
“We have PIQE in 2D and we wanted to know if it could be applicable in 3D sequences. This is currently a ‘work in progress’, but the preliminary results are promising in certain applications.”
“For example, I am reconstructing my 3D T2 scans with PIQE at 1.5 Tesla for vestibular schwannoma explorations. When PIQE is applied, the image is sharper and it's very useful to interpret the cranial nerves, or even the fine structure within the inner ear, such as the basilar membrane.”
“PIQE makes the image sharper without any loss of SNR,” he added.
Solving a difficult equation
Solutions that can tackle both old and new challenges in Neurology Imaging have potential to bring many benefits and contribute to advancing this field.
“I think AiCE and PIQE help solve one of our most difficult equations to enable us to offer higher and higher quality image while maintaining a high throughput of patients,” said Dr. Savatovsky. “It is vital for us both for economic reasons and within complicated context to maintain a very good image quality as well as good care for our patients.”
“I think AiCE and PIQE help solve one of our most difficult equations to enable us to offer higher and higher quality image while maintaining a high throughput of patients,” said Dr. Savatovsky. “It is vital for us both for economic reasons and within complicated context to maintain a very good image quality as well as good care for our patients.”
“I think AiCE and PIQE help solve one of our most difficult equations to enable us to offer higher and higher quality image while maintaining a high throughput of patients.”
Julien Savatovsky, M.D.
Deputy head of Diagnostic Neuroradiology at the Rothschild Foundation Hospital, Paris, France
Dr. Julien Savatovsky
Dr. Savatovsky is Deputy Head of the Imaging Department (Diagnostic Neuroradiology) at the Foundation Adolphe de Rothschild Hospital, Paris, France.
Dr. Savatovsky was trained in Medicine at the University of Paris-VI Pierre and Marie. He completed his training with a two-year fellowship in the Neuroradiology Department of the Pitié-Salpêtrière Hospital Group. He continued his hospital career as a hospital practitioner in neuroradiology at the Foundation Adolphe de Rothschild Hospital.
He specializes in nervous system imaging and head and neck diseases. In order to give access to high-quality imaging to patients, Dr. Savatovsky uses specifically optimized 3T MRI equipment and has advanced training in MRI techniques (including sequence development and tuning) and in advanced post-processing.
In addition to his clinical work and the management of the Imaging Department, Dr. Savatovsky has been involved in research for several years. He has contributed to over a hundred articles published in international journals.
His main areas of research include:
He is a lecturer at various Parisian universities and at the European Course in Neuroradiology. He has also contributed to several radiology textbooks.
Disclaimer
Some features presented in this article may not be commercially available on all systems shown or may require the purchase of additional options. Due to local regulatory processes, some commercial features included in this publication may not be available in some countries. Please contact your local representative from Canon Medical Systems for details and the most current information.
For more information
Due to medical device regulatory reasons, not all products/services displayed on this Canon Medical Systems webpage are available in all countries or markets. Future availability of the products/service cannot also be guaranteed. Don't hesitate to contact your local Canon Medical Systems representative for further details.
