Paving the Way for Precision Medicine
How Robust and Standardized Scans Will Pave the Way to Quantitative Radiology for Precision Medicine
Dr. Hildegard Kaulen | 2017-05-18
For Konstantin Nikolaou, MD, the future belongs to quantitative radiology and image mining as it is a key factor in the success of precision medicine. Clinically relevant image information will only be useful if it can be tapped reliably and with consistent quality as well as be comparable across all centers. In Tübingen he spoked about how that will soon become reality thanks to robust and standardized scans.
The key to quantitative radiology is standardization
Professor Konstantin Nikolaou, who heads the Department of Diagnostic and Interventional Radiology at University Hospital in Tübingen, Germany is a man with a vision. He has no doubts that radiology will play an influential role in disease management in the coming years, a role that goes far beyond merely recognizing cancers and other diseases. “The future belongs to quantitative radiology and image mining,” he says. “As future information specialists, we will communicate with our clinical partners on an equal footing.”
Quantitative radiology seeks out objective and quantitative image information, which provides doctors with a more precise picture of the disease state and predicts the severity and progression of a disease, as well as its prognosis and response to treatment. Ultimately, this can only be achieved by using the combined forces of big data and artificial intelligence. However, clinically relevant image information will only be useful if it can be tapped reliably and with consistent quality. The results must also be comparable across all centers. For Professor Nikolaou, quantitative radiology is a key factor in the success of precision medicine, which aims to ensure that the right patient receives the right treatment at the right time, tailored to his or her current situation. Doctors can only do this if they have all of the available information. In addition, they need standardized image files of consistent quality, irrespective of how challenging the particular examination is and how experienced the user is. The prerequisite for quantitative radiology is standardization in imaging.
Robust scans using modern magnetic resonance imaging
In magnetic resonance imaging, however, achieving standardization is no easy feat. This technology operates without exposing patients to radiation and provides outstanding soft tissue contrast, but is also sensitive to variability in a patient’s anatomy and physiology, or the operator’s experience and training. Breathing difficulties, an irregular heartbeat or a patient’s individual body size and shape can result in motion- or other artifacts.. In everyday clinical practice, scans either are repeated or take longer than usual in order to compensate for motion artifacts or other problems. This results in considerable additional costs, which are not accounted for within the narrow constraints of the German diagnosis-related groups. Jalal Andre, MD, of the University of Washington in Seattle and his colleagues calculated that, on average, repeat scans of this nature result in additional annual costs of US$115,000 per scanner.
With the new MAGNETOM Vida1, featuring BioMatrix Technology and its new system architecture, Siemens Healthineers is addressing precisely these problems. The 3-Tesla MRI scanner is a world first, automatically adapting to patients’ anatomical and physiological characteristics to offer consistent and high-quality image information irrespective of the patient being examined and the person operating the system. MAGNETOM Vida was recently presented in Tübingen and at the European Congress of Radiology (ECR) in Vienna, Austria. Professor Nikolaou and his team have already been working with the system for a number of weeks. “ BioMatrix Technology is the first MRI innovation to automatically adapt to the patient and its individual biological characteristics. Available for the first time on MAGNETOM Vida, the technology will help us pave the way for precision medicine in MRI,” said Christoph Zindel, MD, Senior Vice President and Head of Magnetic Resonance Imaging at Siemens Healthineers, at the launch in Tübingen.
BioMatrix: a world first
BioMatrix Technology anticipates and responds to the challenges faced in each individual scan, so that the patient no longer needs to adapt to the system. Instead, the system adapts to the patient. This is an important prerequisite for standardizing MRI scans. “With MAGNETOM Vida, it takes us five minutes to scan a patient with a glioblastoma tumor before going into surgery,” said Professor Mike Notohamiprodjo, MD in Tübingen. “The scan is planned and conducted semi-automatically and produces robust and standardized image data.” Notohamiprodjo is the senior attending physician responsible for MRI in Professor Nikolaou’s department.
BioMatrix Technology operates using sensors, tuners, and interfaces. The respiratory sensors are built into the scanner table and automatically measure changes in the patient’s breathing. With this information, the operator is able to plan the scan better as they can make a more accurate assessment of the individual patient’s ability to hold his or her breath. This reduces the complexity of the scan and makes it easier to identify the optimal time for image acquisition. “We examine very young, very old, and very sick individuals,” says Professor Notohamiprodjo. “Many patients are unable to cooperate or hold their breath as a result of age or illness.” Professor Notohamiprodjo illustrated this using the example of a six-year-old girl with nephroblastoma examined recently. “The improved monitoring of the respiratory status meant we were able to obtain excellent images,” says the radiologist.
Professor Notohamiprodjo has also had positive experiences with the new acceleration technique Compressed Sensing, which allows free-breathing examinations of the heart or the abdomen, in particular the liver. This disruptive technique only acquires the most essential data to speed up the acquisition and uses new iterative reconstruction algorithms to create a high-quality, artifact-free image. Compressed Sensing GRASP-VIBE1 can be used to perform dynamic, free-breathing liver examinations. “At present, a dynamic liver examination involves four steps and a complex timing,” says Professor Notohamiprodjo. “The current procedure also requires the patients to hold their breath repeatedly. With the new application, the whole scan takes place at the touch of a button.”
BioMatrix Tuners provide a high level of consistency in the image files. For scans of the cervical spine area for example, intelligent coil systems balance out distortions caused by the patient’s anatomy, and the tuners also improve the quality and reproducibility of diffusion-weighted whole-body imaging. BioMatrix Interfaces simplify and speed up contact with the patient, resulting in a faster and more cost-effective scan. Once the operator has selected the region to be examined, the intelligent body models deposited move the patient to the correct scan position automatically. The GO technology assist the user intuitively through the entire imaging process from positioning, over scanning, post-processing to result distribution. The patient table also features a new motorized control system that allows even extremely overweight patients to be moved back and forth easily. “We used to put a lot of thought into which scanner best suits which patient,” says Professor Notohamiprodjo. “With MAGNETOM Vida, there is no need for these deliberations as the scanner can be used with every patient, even those who would previously have been unsuitable for an MRI.”
Image mining and decision support
The continuous growth of medical imaging leads to an exponential increase of imaging information. “We need a strategy for analyzing this data,” says Professor Nikolaou. He adds that the considerable decrease in the number of explorative laparotomies is testament to the fact that today far more information is extracted from the medical images than in the past. In cases of prostate cancer, multiparametric MRI can differentiate between high- and low-malignancy carcinomas. This, he says, raises the hope that it could be used to decide if patients need surgery or if it is best to wait and see.
Professor Nikolaou is certain that medical image data will soon be combined with data from pathology, laboratory medicine, and genetics to create a unique decision support system for precision medicine. Tübingen has its own center, set up under the German Excellence Initiative, which is driving advances in personalized medicine. Professor Nikolaou is aware that the quest for quantitative image characteristics will not succeed without image mining, big data and artificial intelligence. So-called ‘radiomics’ transforms images into higher dimensional pools of data that can be analyzed using intelligent algorithms. “In the field of radiology, we need to get ready for artificial intelligence,” says Professor Nikolaou. “To do this, we will need to change the way we work.”
About the Author
Dr. Hildegard Kaulen is a molecular biologist. After previous roles at the Rockefeller University in New York and Harvard Medical School in Boston, she has been working as a freelance science journalist for respected daily newspapers and science magazines since the mid 1990’s.
About Siemens Healthineers
Our new name expresses our mission and what we stand for: Helping healthcare providers become more successful in caring for their patients.
Our new name is unique. It embodies our pioneering spirit and our engineering expertise in the healthcare industry. It reflects the fact that this industry is increasingly driven by the skills of individuals. It also contains a commitment and a promise to our customers that we will use our pioneering spirit and our engineering expertise to help drive their success.
1510(k) pending. The product is still under development and not commercially available yet. Its future availability cannot be ensured.
The statements by Siemens’ customers described herein are based on results that were achieved in the customer's unique setting. Since there is no "typical" hospital and many variables exist (e.g., hospital size, case mix, level of IT adoption) there can be no guarantee that other customers will achieve the same results.