Applications in the areas of diagnostics and neuroscience
Published February 2021
•
Copyright © IOP Publishing Ltd 2021
Pages 6-1 to 6-10
You need an eReader or compatible software to experience the benefits of the ePub3 file format.
Download complete PDF book, the ePub book or the Kindle book
Permissions
Abstract
Chapter 6 deals exclusively with the applications of superconductors in bio-magnetism, particularly some recent ones towards diagnostics and neuroscience. The successful use of high-T c superconductors is also highlighted, wherever it has been accomplished.
In chapter 4, we have discussed applications based largely on bulk superconducting materials, such as wires/cables for the power sector dealing with energy transmission and storage, as high-field magnets wound from superconducting cables/wires, such as for MRI machines, maglev, particle accelerators and fusion reactors; or as thin films, such as in RF cavities. The largest commercial market for superconducting magnets has, of course, been for their application in MRI scanners.
In chapter 5, we described the low field applications of superconductors, such as the use of sensitive SQUID sensors (figure 5.7) in mapping the very low magnetic fields generated by the human heart using the technique called magnetocardiography (MCG), or in the brain known as magneto-encephalography (MEG). We also tried to focus on the specific capabilities of MEG, when compared with other scanning techniques such as EEG, fMRI, and fNIRS and other functional brain imaging techniques, including magnetic resonance spectroscopy (MRS) and invasive EEG (iEEG) (figure 5.8).
In this short chapter, we endeavor to take up the recent progress in brain imaging techniques based on superconducting materials, such as MRI and MEG, to investigate how brain function supports mental activities, i.e. cognitive neuroscience. For instance, we shall discuss how rt-fMRI-NF (real-time functional magnetic resonance imaging neurofeedback) has been emerging, not just as a diagnostic tool, but also for therapy, i.e. treatment of neuro-psychiatric disorders; and, the recent dramatic growth of human brain-mapping after the advent of fMRI BOLD ('blood oxygen level dependent contrast') imaging. Similarly, psycho-physiological MEG studies are beginning to assume a role in suppressing the spontaneous brain activity in favor of the evoked activity.
6.1. Brain imaging and cognitive neuroscience
Cognitive neuroscience deals with the study of how brain function supports mental activities. Among the different brain imaging techniques, the oldest is x-ray computed tomography (CT), which was introduced in 1971 by Godfrey Hounsfield. It deals with obtaining three-dimensional transaxial tomographic images by passing highly focused x-ray beams through the brain and recording their attenuation [1]. However, CT is an anatomical tool, whereas 'functional brain imaging' for humans falls in the purview of the techniques known as positron emission tomography (PET), functional magnetic resonance imaging (fMRI), electroencephalography (EEG), electrocorticography (ECoG), magnetoencephalography (MEG) and, optical imaging with near-infrared spectroscopy (NIRS).
MEG employs the unique sensitivity of SQUID magnetometry, using which magnetic fields of the order of 10−11 Oe can be detected, which translates into measurement of magnetic flux with a resolution of the order of 10−5 ϕ0 in a bandwidth of 1 Hz. MEG has unique applications in diagnostics through brain imaging related to neuro-disorders, plus as a general tool for biomagnetism. SQUID magnetometry also finds applications in magneto-telluric detection, in the search for oil basins and in basic research for precise study of magnetic properties.
Both PET and MRI (explained in chapter 4), have the capability to do functional brain imaging because any changes in the cellular activity cause local blood flow changes in the brain, quite similar to the earlier non-tomographic techniques.
Positron emission tomography (PET) employs radionuclides with short half-lives, viz. 15O (2 min), 13N (10 min), 11C (20 min) and 18F (110 min), which decay by emission of positrons, providing a link between biology and medicine, fortuitously, because carbon, nitrogen and oxygen are the building blocks of most biological molecules and fluorine can be substituted for hydrogen in some instances. While working with these positron-emitting radionuclides, if an image of the density of a transverse section of the body could be reconstructed from the measured attenuation of highly-focused x-ray beams projected through the section (i.e. x-ray CT), then the distribution of a radionuclide within the section (especially ones that decayed by positron emission) could be accurately and quantitatively reconstructed from its emissions, which became the basis of PET.
MRI gained quick popularity because it doesn't use any ionizing radiation and yields superb images of the human body with much greater detail and variety than CT because of its high sensitivity to soft tissues. Owing to the existence of a correlation between the neuronal activation and cerebral blood flow, the use of any region of the brain leads to (calls upon) an increased blood flow to that region of the brain. The fMRI (functional MRI), which measures brain activity by detecting changes associated with blood flow, has turned out to be a serious player in the functional mapping of the human brain, as discussed in an overview by Glover [2].
6.1.1. rt-fMRI-NF
In recent years, a non-invasive MRI-based technique called rt-fMRI-NF (real-time functional magnetic resonance imaging neurofeedback) has been emerging as a tool for the treatment of neuro-psychiatric disorders. Under rt-fMRI-NF, we have, perhaps for the first time, the potential to use a neuroscience-based intervention to train human brain function towards healthier patterns [3]. Real-time functional magnetic resonance imaging neurofeedback (rt-fMRI-NF) presents fMRI signals to participants in a real-time manner to change their behaviors. Any resultant alterations in comportments observed after real-time fMRI as neurofeedback are postulated to be caused by neural plasticity driven by the induction of specific targeted activities at the neuronal level (targeted neural plasticity model).
6.1.2. Blood oxygenation level dependent (BOLD) MRI
Using contrast agents and rapid data acquisition strategies, one can use MRI to measure changes in brain blood volume produced by physiological manipulations of brain blood flow. High contrast MRI images relevant to functional mapping of the human brain can yield excellent results when combined. However, one can administer a contrast agent only a limited number of times. This limitation could be addressed due to the property of deoxyhemoglobin itself in a magnetic field. The magnetic susceptibilities of oxygenated and deoxygenated hemoglobin differ significantly. Unlike oxyhemoglobin, the deoxyhemoglobin is paramagnetic and, hence, can function as an MRI contrast agent.
The deoxygenated hemoglobin that is present in the veins acts like a little magnet due to its paramagnetic behavior. Under a magnetic field, such as during an MRI scan, owing to the presence of the exposed iron in the hemoglobin molecule, and its presence in large veins in the resting state, the deoxygenated hemoglobin makes the veins stand out as dark lines. However, the presence of oxygen in the oxygenated hemoglobin, found in arteries, 'neutralizes' this effect of the iron such that oxygenated blood can be present in a magnetic field without disrupting it. BOLD imaging, however has its limitations because the cerebral blood flow (CBF), being an indirect marker of activity cannot visualize the active cortex directly. Besides, any increased activity of the brain takes time (about 5 s) to generate increased CBF, and further, the blood flow can be regulated in only a small region (a few mm dia).
Functional brain imaging with PET, or with fMRI are both based on changes in the circulation in the brain, and on metabolisms that are associated with activity changes in both neurons and astrocytes. The profile of these changes was first detailed by PET, which showed that increases in blood flow and glucose utilization far exceeded that of oxygen utilization. Consequently, the amount of oxygen available in the brain increases, causing the relative percentage of paramagnetic deoxyhemoglobin to decrease. The fMRI signal arises because of this change in the relative amount of deoxyhemoglobin (deoxyHb). Dark areas of veins (containing deoxyHb) disappear when the subject breathes on 100% oxygen. In the human fMRI experiment, deoxyHb can decrease when blood flow increases more than oxygen consumption, leading to an enhancement of the fMRI BOLD signal. While on 100% oxygen, the venous structures disappear, which is called the 'blood oxygen level dependent contrast' or BOLD contrast, as discussed by Ogawa et al [4], which provides an additional feature to magnetic resonance imaging. The human brain-mapping has seen a dramatic growth, since the advent of fMRI BOLD imaging.
6.2. Neuro-diseases
6.2.1. Tourette syndrome
Although the critical potential of rt-fMRI-NF is quite general, being applicable to treat conditions arising from depression to Parkinson's disease, it is quite useful as a clinical intervention for Tourette syndrome (TS). TS is a neurodevelopmental disorder characterized by childhood onset of motor and vocal tics affecting an estimated 14/1000 children, with peak severity afflicting adolescents, who undergo repetitive movements or vocalizations known as tics. Recently, Yale researchers trained 21 adolescents aged 11 to 19 years with TS to control their tics through rt-fMRI-NF, which allows them to monitor the function of their own brain in real time. TS has been linked with dysfunction in motor corticostriatal–thalamo-cortical loops, and the supplementary motor area (SMA) is particularly a key node in the dysfunctional neural circuit underlying the chronic tics of TS. Stimulation of the SMA can produce movements as well as urges to move, similar to tics and the premonitory urges experienced by patients with TS. The imaging technique rt-fMRI-NF has been used by Sukhodolsky et al [5] to track the feedback from the supplementary motor area (SMA), the brain region associated with tics in Tourette syndrome.
To date, the tics in TS are treated using behavior therapy and pharmaceuticals, but not every case responds to those efforts. The rt-fMRI-NF intervention has great potential, therefore, for treating TS, since NF can potentially harness feedback learning to train targeted control over the neural circuitry underlying tic generation.
Neuro-feedback is a low-risk, drug-free treatment, and although functional neuroimaging is expensive, the time commitment for neurofeedback scans is less than that spent on a course of behavioral therapy for tics. Being a new technique, the neurofeedback protocol is yet to be optimized, and the efficacy of this intervention for TS is expected to increase with time.
Neurofeedback lets patients see activity in a particular brain region while thinking of specific cues. In this way, a patient learns which cues trigger the most activity, and can therefore trigger the development of more connections. As a brain scanner measures brain activity, the patients imagine motor tasks while looking at a bar representing activity in the SMA region.
Neurofeedback training may increase brain connectivity in patients with Huntington's disease (HD), too, even after some degeneration, with chances to improve behavior and movement abnormalities.
6.2.2. rt-fMRI-NF for ADHD
Critical underlying neurofunctional deficits can be targeted by using fMRI-NF [6], which teaches participants (even children) to self-regulate 'blood oxygen level dependent' (BOLD) response in specific brain regions based on real-time feedback of their brain activation [7].
The advantages of fMRI-NF are no known side effects and potential longer-term neuroplastic effects. Neurofeedback using fMRI has several advantages over EEG-neurofeedback. Due to its superior spatial resolution, it can target key neurofunctional biomarkers, such as the inferior frontal cortex or the basal ganglia, which cannot directly be reached with EEG-Neurofeedback, as observed by Rubia [6]. Although more costly per session, self-regulation is typically achieved much faster with fMRI-neurofeedback than with EEG-neurofeedback. The comparison between pre- and post-fMRI-neurofeedback of a typically dysfunctional frontal region in ADHD adolescents showed a significant decrease in the ADHD symptoms.
To rule out behavioral changes in conventional real-time fMRI neurofeedback studies due to any alternative accounts, including the placebo effect and physiological artifacts, a decoded neurofeedback (DecNef) system has been developed by Shibata et al [8] by an integration of implicit neurofeedback and fMRI multivariate analysis. DecNef provides strong evidence for the targeted neural plasticity and refutes the possibility of any artifacts causing it. It has also been shown by employing DecNef that the targeted neural plasticity occurs at the neuronal level during DecNef training.
6.3. The salience network (SN)
A collection of parts of the human brain, comprising the anterior insula (AI) and dorsal anterior cingulate cortex (dACC), is known as the salience network (SN), or more technically, the cingulo-opercular network. Along with its interconnected brain networks, the SN plays an important role in the conduct of complex human behavioral functions, such as communication, self-awareness and social behavior by collating cognitive, sensory and emotional information. Several psychiatric disorders, such as schizophrenia, Alzheimer's disease, post-traumatic stress disorder, anxiety disorders and frontotemporal dementia have been linked with the dysfunction of the SN [9]. During anxiety disorders, the anterior insula (AI) node of the SN may become hyperactive which can lead to a worried state of mind. A reduction of salience of social stimuli related to eyes, gaze and face in the case of autistic people is thought to be responsible for relatively impaired social skills in them, as observed by Menon [10].
fMRI studies have, in fact, pointed to the severely affected SN across several psychiatric disorders, because the SN coordinates the neural resources of the brain in response to the detection of the behaviorally relevant stimuli by it [11].
6.4. SN and the mesolimbic dopamine system
Dopamine neurons also play a role in the identification of behaviorally relevant environmental stimuli. Mesolimbic dopamine neurons are instrumental in assigning salience to relevant environmental stimuli [12, 13]. Dysfunction of this system is also observed in many neuropsychiatric illnesses [14, 15]. Mesolimbic dopamine signaling plays a role in the modulation of the salience network which, therefore, calls for the development of an integrative understanding of SN and the mesolimbic dopamine system. Using positron emission tomography (PET) to measure dopamine release capacity and dopamine synthesis capacity in the ventral striatum; and resting-state fMRI to investigate salience network functional connectivity in some individuals, it has been shown that the dopamine synthesis capacity is associated with greater salience network connectivity, particularly for the brain regions that act as information-processing hubs. In contrast, the dopamine release capacity was associated with weaker salience network connectivity. These findings demonstrate a close relationship between the salience network and mesolimbic dopamine system, and its relevance to neuropsychiatric illnesses which produce aberrant functioning of both these systems, as reported by McCutcheon et al [16].
6.5. Magnetic resonance perfusion
In patients with risk factors for coronary artery disease, a common strategy employed for diagnosis is invasive angiography to visualize the presence and extent of coronary artery disease, supported by the assessment of fractional flow reserve (FFR) to guide the need for subsequent revascularization. One can alternately go for a noninvasive test for the detection of coronary artery disease, viz. myocardial-perfusion cardiovascular magnetic resonance imaging (MRI), which has a high concordance with FFR for ischemia detection, as explained by Nagel et al [17]. Cardiovascular MRI has been associated with a lower incidence of invasive angiography than testing based on clinical risk assessment. In the group undergoing cardiovascular-MRI, myocardial perfusion cardiovascular MRI was performed with the use of scanners that had a magnetic field strength of 1.5 T. Myocardial perfusion was assessed with the first pass of gadobutrol (in the form of Gadovist, from Bayer, Germany) typically at a dose of 0.075 mmol kg−1 of body weight during adenosine infusion.
6.6. BIO-interface
6.6.1. Studies on dog brains and function
Humans have bred different lineages of domestic dogs for different tasks, like hunting, herding, guarding, or companionship. In a first study of its kind, scientists have documented how brain structure varies across dog breeds and corresponds to the specific behavior each breed is known for. These behavioral differences have arisen from their underlying neuro-anatomical differences, which were the subject of an MRI study made in 2019 by Hecht et al [18]. It hints that though the brain structure is related to function, there does exist an interaction between genetically determined neuro-anatomy and learned behavior.
6.6.2. Plant biomechanics
As the fruit of a witch hazel plant dries out, the top part of the woody capsule around the seed splits open. The middle part of the capsule constricts, as if it were squeezed by fingers, until the seed, about the size of a pumpkin seed, breaks free and flies out at about 28 miles h−1. Just before it shoots out, one hears the sound of a crack which is caused by a squeezing, because there is no other explosive mechanism to it. As it flies, the seed is spinning fast at about 12 000–25 000 revolutions per minute, which enables the seed to travel far, landing several yards away. Interestingly, the seeds from different fruits on the same witch hazel plant can spin in different directions! In addition to using high-speed cameras to study the seeds flying out, the researchers have also put the plants inside MRI machines, recently [19], to examine the hidden structures within the hazel plant and the fruit without having to cut it apart, and to get insights into the ballistic seed dispersal. By studying how the witch hazel launches its seeds by applying a torque to make them spin, one could inspire engineers to design a sensor that can detect when the humidity falls below a critical level, resulting in the opening of a valve.
Repetitive MRI scans on the same plant as it grows over time can give a lead on the mechanism of the movements of other plants, too, such as that of the Venus flytrap that is famous for catching bugs to trap and eat them.
6.7. Signal-space projection/separation for MEG data
Raw MEG data are a combination of brain activity, biological interference, and technical noise from outside. The signals that are not brain-related are widely suppressed by recording data in a magnetically shielded room, filtering it, and with software noise cancellation. Two prevalent methods for the suppression of external interference are signal-space projection [20], and signal space separation [21, 22]. A digital highpass filter eliminates possible baseline drifts, and by band-pass filtering the focus is put on a certain frequency range.
The brain signals themselves can arise from two sources: (a) spontaneous activity, and, (b) activity elicited by a stimulus, in the case of neurocognitive experiments. Spontaneous activity is ongoing and can be measured without external stimulation. The challenge lies in trying to separate the stimulus-related brain signals from spontaneous activity signals, because both originate in the brain. Stimulus-related activity can, however, be distinguished by making a comparative analysis of the responses obtained under evoked activity and induced activity, as discussed below.
6.8. Evoked and induced responses
Evoked responses have a fixed latency and are phase-locked to a stimulus. The stimulus can either be external, presented visually or auditorily, or it can be any response of the subject, like a button press or an eye blink, for example. To compute the generators of the event-related field (ERF), averaging proves to be useful. It is a standard procedure in data processing of many psycho-physiological MEG studies that has been applied very successfully to suppress spontaneous brain activity in favor of the evoked activity.
Induced responses have a variable latency and are not necessarily phase-locked, so averaging them might lead to an attenuation. The brain does not always respond identically to repeated stimuli and there might be a great deal of information in the data which gets lost during the averaging process.
6.9. Consequences of deprivation of sleep
Acute deprivation of sleep can have disastrous effects, and even cause death. Researchers have investigated electromyograms (EMG), electroencephalograms (EEG) and also electrooculograms (EOG) to study both the rapid eye movement (REM) and non-REM (NREM) sleep phases [23], to investigate the origin and functions of sleep.
Sleep-induced changes in the brain are thought to correlate with consciousness changes in the brain. Sleep is vitally important to the body's regulation of oxidation, mainly in the gut. Deprivation of sleep has been seen to be associated with building up of reactive oxygen species (ROS) in the intestines [24, 25]. MEG holds promise to elucidate the hidden causes of acute lack of sleep in some people.
6.10. Non-destructive imaging of soft tissue using synchrotron radiation
The synchrotron light source is about a hundred billion times more intense than the x-ray equipment that we come across in a hospital. Synchrotron radiation (SR) is emitted from an electron traveling at almost the speed of light when a magnetic field bends its path. Storage ring based synchrotron light sources offer synchrotron radiation, i.e. x-ray beams which have brightness and coherence boosted orders of magnitude over conventional x-ray sources. It allows high-resolution observations across multiple length scales, within very short and concise time spans. Other than deep x-ray penetration power and wide x-ray energy tunability, the improved brightness of the state-of-the-art synchrotron sources offers microscopic and imaging modalities. Synchrotron x-ray sources are now being used as a microscope to conduct virtual histology, because SR can be tuned to much shorter wavelengths than that of the normal light in a room, allowing synchrotron x-rays to make cellular level study of soft tissue, without making any incisions.
For the first time, peripheral nerve samples from different subjects: a healthy person, a type 1 diabetes patient, and a type 2 diabetes patient have been studied using x-ray phase contrast holographic nano-tomography at European Synchrotron Radiation Facility (ESRF) at Grenoble, using an x-ray beam of 17 keV energy [26]. Such studies hold significance in how diabetes affects the growth of nerve fibers in the arms and legs, an important aspect in diabetes neuropathy. With the availability of compact synchrotrons, additional insights on such topics will unfold in the coming years.
6.11. Carbon-ion radiotherapy
In order to treat a deep-seated tumor with well-localized dose distributions, carbon ions obtained from a superconducting synchrotron are considered to be a good candidate for heavy-ion radiotherapy. The carbon-beam intensity, which depends on the volume and shape of the target, and the efficiency of the irradiation method, needs to be a few 108 particles per second in order to obtain a biological dose rate of 5 Gy E min−1, which roughly equates to a physical dose of 2 Gy min−1 [27, 28]. The National Health Insurance scheme of Japan has approved carbon-ion radiotherapy (C-RT) for treating 'bone and soft tissue tumor', 'prostate tumor' and 'head and neck tumor'.
References
- [1]Raichle M E 2008 A brief history of human brain mapping Trends Neurosci. 32 118–26
- [2]Glover G H 2011 Overview of functional magnetic resonance imaging Neurosurg. Clin. N. Am. 22 133–9
- [3]Linden D E J and Turner D L 2016 Real-time functional magnetic resonance imaging neurofeedback in motor neurorehabilitation Curr. Opin. Neurol. 29 412–8
- [4]Ogawa S, Lee T M, Kay A R and Tank D W 1990 Brain magnetic resonance imaging with contrast dependent on blood oxygenation Proc. Natl. Acad. Sci. 87 9868–72
- [5]Sukhodolsky D G et al 2020 Randomized, sham-controlled trial of real-time fMRI neurofeedback for tics in adolescents with Tourette syndrome Biol. Psychiatry 87 1063–70
- [6]Rubia K 2018 Cognitive neuroscience of attention deficit hyperactivity disorder (ADHD) and its clinical translation Front. Hum. Neurosci. 12 1–23
- [7]Rubia K, Criaud M, Wulff M, Alegria A, Brinson H, Barker G, Stahl D and Giampietro V 2019 Functional connectivity changes associated with fMRI neurofeedback of right inferior frontal cortex in adolescents with ADHD Neuroimage 188 43–58
- [8]Shibata K, Lisi G, Cortese A, Watanabe T, Sasaki Y and Kawato M 2019 Toward a comprehensive understanding of the neural mechanisms of decoded neurofeedback NeuroImage 188 539–56
- [9]Menon V 2011 Large-scale brain networks and psychopathology: a unifying triple network model Trends Cogn. Sci. 15 483–506
- [10]Menon V 2015 Salience network Brain Mapping: An Encyclopedic Reference vol 2 , ed A W Toga (Cambridge, MA: Academic) pp 597–611
- [11]McTeague L M, Huemer J, Carreon D M, Jiang Y, Eickhoff S B and Etkin A 2017 Identification of common neural circuit disruptions in cognitive control across psychiatric disorders Am. J. Psychiatry 174 676–85
- [12]Howes O D and Nour M M 2016 Dopamine and the aberrant salience hypothesis of schizophrenia World Psychiatry 15 3–4
- [13]Takahashi Y K, Batchelor H M, Liu B, Khanna A, Morales M and Schoenbaum G 2017 Dopamine neurons respond to errors in the prediction of sensory features of expected rewards Neuron 95 1395–405
- [14]Volkow N D, Wise R A and Baler R 2017 The dopamine motive system: implications for drug and food addiction Nat. Rev. Neurosci. 18 741–52
- [15]Salamone J D and Correa M 2012 The mysterious motivational functions of mesolimbic dopamine Neuron 76 470–85
- [16]McCutcheon R A et al 2018 Mesolimbic dopamine function is related to salience network connectivity: an integrative positron emission tomography and magnetic resonance study Biol. Psychiatry 85 368–78
- [17]Nagel E et al 2019 Magnetic resonance perfusion or fractional flow reserve in coronary disease New Engl. J. Med. 380 2418–28
- [18]Hecht E E, Smaers J B, Dunn W D, Kent M, Preuss T M and Gutman D A 2019 Significant neuroanatomical variation among domestic dog breeds J. Neurosci. 39 7748–58
- [19]Poppinga S et al 2019 A seed flying like a bullet: ballistic seed dispersal in Chinese witch-hazel (Hamamelis mollis OLIV., Hamamelidaceae) J. R. Soc. Interface 16 20190327
- [20]Uusitalo M A and Ilmoniemi R J 1997 Signal-space projection method for separating MEG or EEG into components Med. Biol. Eng. Comput. 35 135–40
- [21]Taulu S, Kajola M and Simola J 2003 The signal space separation method Abstract Book of the NFSI2003 Conf. (Chieti, Italy) p A79
- [22]Taulu S and Kajola M 2005 Presentation of electromagnetic multichannel data: the signal space separation method J. Appl. Phys. 97 124905-1–10
- [23]Hobson J A 2005 Sleep is of the brain, by the brain and for the brain Nature 437 1254–6
- [24]Greenwood V 2020 Why sleep deprivation kills Quanta https://www.quantamagazine.org/why-sleep-deprivation-kills-20200604/
- [25]Vaccaro A, Dor Y K, Nambara K, Pollina E A, Lin C, Greenberg M E and Rogulja D 2020 Sleep loss can cause death through accumulation of reactive oxygen species in the gut Cell 181 1307–28
- [26]Dahlin L B, Rix K R, Dahl V A, Dahl A B, Jensen J N, Cloetens P, Pacureanu A, Mohseni S, Thomsen N O B and Bech M 2020 Three-dimensional architecture of human diabetic peripheral nerves revealed by x-ray phase contrast holographic nanotomography Sci. Rep. 10 7592
- [27]Malouff T D, Mahajan A, Krishnan S, Beltran C, Seneviratne D S and Trifiletti D M 2020 Carbon ion therapy: a modern review of an emerging technology Front. Oncol. 10 82
- [28]Ohno T et al 2011 Carbon ion radiotherapy at the Gunma University Heavy Ion Medical Center: new facility set-up Cancers 3 4046–60