When taking oral drugs, transporter proteins found on cells that line the gastrointestinal tract facilitate their entry into the bloodstream. But for many drugs, it is not known which of those transporters they use to exit the digestive tract.
Identifying the transporters used by specific drugs could help to improve patient treatment because if two drugs rely on the same transporter, they can interfere with each other and should not be prescribed together.
Researchers at MIT, Brigham and Women’s Hospital, and Duke University have developed a multipronged strategy to identify the transporters used by different drugs, which appears in Nature Biomedical Engineering. Their approach, which makes use of both tissue models and machine-learning algorithms, has already revealed that a commonly prescribed antibiotic and a blood thinner can interfere with each other.
“One of the challenges in modelling absorption is that drugs are subject to different transporters. This study is all about how we can model those interactions, which could help us make drugs safer and more efficacious, and predict potential toxicities that may have been difficult to predict until now,” says Giovanni Traverso, an associate professor of mechanical engineering at MIT, a gastroenterologist at Brigham and Women’s Hospital, and the senior author of the study.
Learning more about which transporters help drugs pass through the digestive tract could also help drug developers improve the absorbability of new drugs by adding excipients that enhance their interactions with transporters.
Former MIT postdocs Yunhua Shi and Daniel Reker are the lead authors of the study.
Drug transport
Previous studies have identified several transporters in the GI tract that help drugs pass through the intestinal lining. Three of the most commonly used, which were the focus of the new study, are BCRP, MRP2, and PgP.
For this study, Traverso and his colleagues adapted a tissue model they had developed in 2020 to measure a given drug’s absorbability. This experimental setup, based on pig intestinal tissue grown in the laboratory, can be used to systematically expose tissue to different drug formulations and measure how well they are absorbed.
To study the role of individual transporters within the tissue, the researchers used short strands of RNA called siRNA to knock down the expression of each transporter. In each section of tissue, they knocked down different combinations of transporters, which enabled them to study how each transporter interacts with many different drugs.
“There are a few roads that drugs can take through tissue, but you don’t know which road. We can close the roads separately to figure out, if we close this road, does the drug still go through? If the answer is yes, then it’s not using that road,” Traverso says.
The researchers tested 23 commonly used drugs using this system, allowing them to identify transporters used by each of those drugs. Then, they trained a machine-learning model on that data, as well as data from several drug databases. The model learned to make predictions of which drugs would interact with which transporters, based on similarities between the chemical structures of the drugs.
Using this model, the researchers analysed a new set of 28 currently used drugs, as well as 1595 experimental drugs. This screen yielded nearly 2 million predictions of potential drug interactions. Among them was the prediction that doxycycline, an antibiotic, could interact with warfarin, a commonly prescribed blood-thinner. Doxycycline was also predicted to interact with digoxin, which is used to treat heart failure, levetiracetam, an antiseizure medication, and tacrolimus, an immunosuppressant.
Identifying interactions
To test those predictions, the researchers looked at data from about 50 patients who had been taking one of those three drugs when they were prescribed doxycycline. This data, which came from a patient database at Massachusetts General Hospital and Brigham and Women’s Hospital, showed that when doxycycline was given to patients already taking warfarin, the level of warfarin in the patients’ bloodstream went up, then went back down again after they stopped taking doxycycline.
That data also confirmed the model’s predictions that the absorption of doxycycline is affected by digoxin, levetiracetam, and tacrolimus. Only one of those drugs, tacrolimus, had been previously suspected to interact with doxycycline.
“These are drugs that are commonly used, and we are the first to predict this interaction using this accelerated in silico and in vitro model,” Traverso says. “This kind of approach gives you the ability to understand the potential safety implications of giving these drugs together.”
Novel toolkit translates neuroimaging data into audiovisual formats to aid interpretation
Simple audiovisualisation of wide field neural activity. Adapted from Thibodeaux et al., 2024, PLOS ONE, CC-BY 4.0
Complex neuroimaging data can be explored through translation into an audiovisual format – a video with accompanying musical soundtrack – to help interpret what happens in the brain when performing certain behaviours. David Thibodeaux and colleagues at Columbia University, US, present this technique in the open-access journal PLOS ONE on February 21, 2024. Examples of these beautiful “brain movies” are included below.
Recent technological advances have made it possible for multiple components of activity in the awake brain to be recorded in real time. Scientists can now observe, for instance, what happens in a mouse’s brain when it performs specific behaviours or receives a certain stimulus. However, such research produces large quantities of data that can be difficult to intuitively explore to gain insights into the biological mechanisms behind brain activity patterns.
Prior research has shown that some brain imaging data can be translated into audible representations. Building on such approaches, Thibodeaux and colleagues developed a flexible toolkit that enables translation of different types of brain imaging data – and accompanying video recordings of lab animal behaviour – into audiovisual representations.
The researchers then demonstrated the new technique in three different experimental settings, showing how audiovisual representations can be prepared with data from various brain imaging approaches, including 2D wide-field optical mapping (WFOM) and 3D swept confocally aligned planar excitation (SCAPE) microscopy.
The toolkit was applied to previously-collected WFOM data that detected both neural activity and brain blood flow changes in mice engaging in different behaviours, such as running or grooming. Neuronal data was represented by piano sounds that struck in time with spikes in brain activity, with the volume of each note indicating magnitude of activity and its pitch indicating the location in the brain where the activity occurred. Meanwhile, blood flow data were represented by violin sounds. The piano and violin sounds, played in real time, demonstrate the coupled relationship between neuronal activity and blood flow. Viewed alongside a video of the mouse, a viewer can discern which patterns of brain activity corresponded to different behaviours.
The authors note that their toolkit is not a substitute for quantitative analysis of neuroimaging data. Nonetheless, it could help scientists screen large datasets for patterns that might otherwise have gone unnoticed and are worth further analysis.
The authors add: “Listening to and seeing representations of [brain activity] data is an immersive experience that can tap into this capacity of ours to recognise and interpret patterns (consider the online security feature that asks you to “select traffic lights in this image” – a challenge beyond most computers, but trivial for our brains)…[It] is almost impossible to watch and focus on both the time-varying [brain activity] data and the behavior video at the same time, our eyes will need to flick back and forth to see things that happen together. You generally need to continually replay clips over and over to be able to figure out what happened at a particular moment. Having an auditory representation of the data makes it much simpler to see (and hear) when things happen at the exact same time.”
This is a pseudo-colored image of high-resolution gradient-echo MRI scan of a fixed cerebral hemisphere from a person with multiple sclerosis.
Credit: Govind Bhagavatheeshwaran, Daniel Reich, National Institute of Neurological Disorders and Stroke, National Institutes of Health
A key feature of multiple sclerosis (MS) is that it causes the patient’s own immune system to attack and destroy the myelin sheaths in the central nervous system. To date, it hasn’t been possible to visualise the myelin sheaths well enough to use this information for the diagnosis and monitoring of MS. Now researchers have developed a new magnetic resonance imaging (MRI) procedure that maps the condition of the myelin sheaths more accurately than was previously possible.
The researchers successfully tested the procedure on healthy people for the first time, and published their results in Magnetic Resonance in Medicine.
In the future, the MRI system with its special head scanner could help doctors to recognise MS at an early stage and better monitor the progression of the disease.
This technology, developed by the researchers at ETH Zurich and University of Zurich, led by Markus Weiger and Emily Baadsvik from the Institute for Biomedical Engineering, could also facilitate the development of new drugs for MS. But it doesn’t end there: the new MRI method could also be used by researchers to better visualise other solid tissue types such as connective tissue, tendons and ligaments.
Quantitative myelin maps
Conventional MRI devices capture only inaccurate, indirect images of the myelin sheaths because these devices typically work by reacting to water molecules in the body that have been stimulated by radio waves in a strong magnetic field.
But the myelin sheaths, which wrap around the nerve fibres in several layers, consist mainly of fatty tissue and proteins. That said, there is some water – known as myelin water – trapped between these layers.
Standard MRIs build their images primarily using the signals of the hydrogen atoms in this myelin water, rather than imaging the myelin sheaths directly.
The ETH researchers’ new MRI method solves this problem and measures the myelin content directly.
It puts numerical values on MRI images of the brain to show how much myelin is present in a particular area compared to other areas of the image.
A number 8, for instance, means that the myelin content at this point is only 8 percent of a maximum value of 100, which indicates a significant thinning of the myelin sheaths.
Essentially, the darker the area and the smaller the number in the image, the more the myelin sheaths have been reduced.
This information ought to enable doctors to better assess the severity and progression of MS.
Measuring signals within millionths of a second
It is difficult however to image the myelin sheaths directly, since the signals that the MRI triggers in the tissue are very short-lived; the signals that emanate from the myelin water last much longer.
“Put simply, the hydrogen atoms in myelin tissue move less freely than those in myelin water. That means they generate much briefer signals, which disappear again after a few microseconds,” Weiger says, adding: “And bearing in mind a microsecond is a millionth of a second, that’s a very short time indeed.” A conventional MRI scanner can’t capture these fleeting signals because it doesn’t take the measurements fast enough.
To solve this problem, the researchers used a specially customised MRI head scanner that they have developed over the past ten years together with the companies Philips and Futura.
This scanner is characterised by a particularly strong gradient in the magnetic field.
“The greater the change in magnetic field strength generated by the three scanner coils, the faster information about the position of hydrogen atoms can be recorded,” Baadsvik says.
Generating such a strong gradient calls for a strong current and a sophisticated design.
As the researchers scan only the head, the magnetic field is more contained and concentrated than with conventional devices.
In addition, the system can quickly switch from transmitting radio waves to receiving signals; the researchers and their industry partners have developed a special circuit for this purpose.
The researchers have already successfully tested their MRI procedure on tissue samples from MS patients and on two healthy individuals. Next, they want to test it on MS patients themselves. Whether the new MRI head scanner will make its way into hospitals in the future now depends on the medical industry. “We’ve shown that our process works,” Weiger says. “Now it’s up to industry partners to implement it and bring it to market.”
Researchers have developed a way to use ultrasound to estimate the risk of delivering a baby preterm. The new method measures microstructural changes in a woman’s cervix using quantitative ultrasound. The method works as early as 23 weeks into a pregnancy, according to the research, which is published in the American Journal of Obstetrics & Gynecology Maternal Fetal Medicine.
The current method for assessing a woman’s risk of preterm birth is based solely on whether she has previously given birth prematurely. This means there has been no way to assess risk in a first-time pregnancy.
“Today, clinicians wait for signs and symptoms of a preterm birth,” such as a ruptured membrane, explained lead author Barbara McFarlin, a professor emeritus of nursing at University of Illinois Chicago.
“Our technique would be helpful in making decisions based on the tissue and not just on symptoms.”
The new method is the result of more than 20 years of collaboration between researchers in nursing and engineering at UIC and University of Illinois Urbana-Champaign. In a study of 429 women who gave birth without induction at the University of Illinois Hospital, the new method was effective at predicting the risk of preterm births during first-time pregnancies.
And for women who were having a subsequent pregnancy, combing the data from quantitative ultrasound with the woman’s delivery history was more effective at assessing risk than just using her history.
The new approach differs from a traditional ultrasound where a picture is produced from the data received.
In quantitative ultrasound, a traditional ultrasound is performed but the radio frequency data itself is read and analysed to determine tissue characteristics.
The study is the culmination of a research partnership that began in 2001 when McFarlin was a nursing PhD student at UIC. Having previously worked as a nurse midwife and sonographer, she had noticed that there were differences in the appearance of the cervix in women who went on to deliver preterm.
She was interested in quantifying this and discovered that “no one was looking at it.”
She was put in touch with Bill O’Brien, a UIUC professor of electrical and computer engineering, who was studying ways to use quantitative ultrasound data in health research.
Together, over the past 22 years, they established that quantitative ultrasound could detect changes in the cervix and, as McFarlin had suspected long ago, that those changes help predict the risk of preterm delivery.
The preterm birth rate hovers around 10-15% of pregnancies, O’Brien said.
“That’s a very, very high percentage to not know what is happening,” he said.
If a clinician could know at 23 weeks that there was a risk of preterm birth, they would likely conduct extra appointments to keep an eye on the foetus, the researchers said.
But since there had previously been no routine way to assess preterm birth risk this early, there have been no studies to show what sort of interventions would be helpful in delaying labour.
This study, O’Brien explains, will allow other researchers to “start studying processes by which you might be able to prevent or delay preterm birth.”
Diffuse gliomas are malignant brain tumours that cannot be optimally examined by means of conventional MRI imaging. So-called amino acid PET (positron emission tomography) scans are better able to image the activity and spread of gliomas. An international team of researchers from the RANO Working Group have drawn up the first ever international criteria for the standardised imaging of gliomas using amino acid PET. It has published its results in the journal The Lancet Oncology.
PET uses a radioactive tracer to measure metabolic processes in the body. Amino acid PET is used in the diagnosis of diffuse gliomas, with tracers that work on a protein basis (amino acids) and accumulate in brain tumours.
The Response Assessment in Neuro-Oncology (RANO) Working Group is an international, multidisciplinary consortium founded to develop standardised new response criteria for clinical studies relating to brain tumours.
Under the joint leadership of nuclear physician Nathalie Albert from LMU and oncologist Professor Matthias Preusser from the Medical University of Vienna, the RANO group has developed new criteria for assessing the success of therapies for diffuse gliomas.
Nathalie Albert explains: “PET imaging with radioactively labelled amino acids has proven extremely valuable in neuro-oncology and permits reliable representation of the activity and extension of gliomas. Although amino acid PET has been used for years, it had not been evaluated in a structured manner before now. In contrast to MRI-based diagnostics, there have been no criteria for interpreting these PET images.” According to the researchers, the new criteria allow PET to be used in clinical studies and everyday clinical practice and create a foundation for future research and the comparison of treatments for improved therapies.
New criteria for PET examinations of brain tumours
Diffuse gliomas are malignant brain tumorus that cannot be optimally examined by means of conventional MRI imaging. So-called amino acid PET scans are better able to image the activity and spread of gliomas.
These malignant brain tumours develop out of glial cells and are generally aggressive and difficult to treat.
The RANO group has developed criteria that permit evaluation of the success of treatment using PET. Called PET RANO 1.0, these PET-based criteria open up new possibilities for the standardised assessment of diffuse gliomas.
Coup and contrecoup brain injury. Credit: Scientific Animations CC4.0
Researchers have created a new brain imaging method that allows to be diagnosed, even when existing imaging techniques like magnetic resonance imaging (MRI) The technique involves loading gadolinium, a standard MRI contrast agent, into ‘backpacks’ that are attached macrophages. mTBIs cause inflammation, attracting macrophages there. Coupling the gadolinium contrast agent to these cells enables MRI to reveal brain inflammation and increase the number of correctly diagnosed mTBI cases, improving patient care. The method is described in a new paper in Science Translational Medicine.
“70-90% of reported TBI cases are categorised as ‘mild,’ yet as many as 90% of mTBI cases go undiagnosed, even though their effects can last for years and they are known to increase the risk of a host of neurological disorders including depression, dementia, and Parkinson’s disease,” said senior author Samir Mitragotri, PhD, in whose lab the research was performed. “Our cell-based imaging approach exploits immune cells’ innate ability to travel into the brain in response to inflammation, enabling us to identify mTBIs that standard MRI imaging would miss.”
Using immune cells to identify inflammation
Most of us know someone who has had a concussion (another name for an mTBI), sometimes even more than one. But the vast majority of people who experience an mTBI are never properly diagnosed. Without that diagnosis, they can exacerbate their injuries by returning to normal activity before they’re fully recovered, which can lead to further damage. Some studies even suggest that repeated mTBIs can lead to chronic traumatic encephalopathy (CTE), the neurodegenerative disease that has been found to afflict more than 90% of professional American football players.
Because the effects of mTBI are believed to be caused by “invisible” brain inflammation, members of the Mitragotri lab decided to leverage their experience with immune cells to create a better diagnostic. “Our previous projects have focused on controlling the behaviour of immune cells or using them to deliver drugs to a specific tissue. We wanted to exploit another innate ability of immune cells – homing to sites of inflammation in the body – to carry imaging agents into the brain, where they can provide a visible detection signal for mTBI,” said first author Lily Li-Wen Wang, Ph.D.. Wang is a former Research Fellow in the Mitragotri Lab at the Wyss Institute and SEAS who is now a scientist at Landmark Bio.
Gadolinium needs water to show up on MRI
The team planned to use their cellular backpack technology to attach gadolinium molecules to macrophages, known to infiltrate the brain in response to inflammation. But right away, they ran into a problem: in order to function as a contrast agent for MRI scans, gadolinium needs to interact with water. Their original backpack microparticles are made of a hydrophobic polymer called PLGA. So Wang and her co-authors started developing a new backpack made out of a hydrogel material that could be manufactured at a large scale in the lab.
After years of hard work, they finally created a new hydrogel backpack that could produce a strong gadolinium-mediated MRI signal, attach stably to both mouse and pig macrophages, and maintain their cargo for a sustained period of time in vitro. They named their new microparticles M-GLAMs, short for “macrophage-hitchhiking Gd(III)-Loaded Anisotropic Micropatches.” Now, it was time to test them in a more realistic setting, for which they partnered with researchers and clinicians at Boston Children’s Hospital.
First, they injected mouse M-GLAMs macrophages into mice to see if they could visualize them in vivo. They were especially interested to see if they accumulated in the kidney, as existing gadolinium-based contrast agents like Gadavist® can cause health risks for patients with kidney disease. Their M-GLAMs did not accumulate in the mice’s kidneys, but persisted in their bodies for over 24 hours with no negative side effects. In contrast, mice injected with Gadavist® showed substantial accumulation of the contrast agent in their kidneys within 15 minutes of injection, and the substance was fully cleared from their bodies within 24 hours.
Then, they tested porcine M-GLAMs in a pig model of mTBI. They injected the M-GLAMs into the animals’ blood two days after a mock mTBI, then used MRI to evaluate the concentration of gadolinium in the brain. They focused on a small region called the choroid plexus, which is known as a major conduit of immune cells into the brain. Pigs that received the M-GLAMs displayed a significant increase in the intensity of gadolinium present in the choroid plexus, while those injected with Gadavist® did not, despite confirmation of increased inflammation macrophage density in the brains of both groups. The animals showed no toxicity in any of their major organs following administration of the treatments.
“Another important aspect of our M-GLAMs is that we are able to achieve better imaging at a much lower dose of gadolinium than current contrast agents – 500-1000-fold lower in the case of Gadavist®,” said Wang. “This could allow the use of MRI for patients who are currently unable to tolerate existing contrast agents, including those who have existing kidney problems.”
Researchers from the University of Birmingham have designed and developed a novel diagnostic device to detect traumatic brain injury (TBI) by shining a safe laser into the eye.
The technique is radically different from other diagnostic methods and is expected to be developed into a hand-held device for use in the critical ‘golden hour’ after traumatic brain injury, when life critical decisions on treatment must be made.
The device, described in Science Advances, incorporates a class 1, CE marked, eye-safe laser and a unique Raman spectroscopy system, which uses light to reveal the biochemical and structural properties of molecules by detecting how they scatter light, to detect the presence and levels of known biomarkers for brain injury.
There is an urgent need for new technologies to improve the timeliness of TBI diagnosis. TBI is caused by sudden shock or impact to the head, which can cause mild to severe injury to the brain, and rapid intervention is necessary to prevent further irreversible damage.
Diagnosis at the point of injury is difficult. Moreover, radiological investigations such as X-ray or MRI are very expensive and slow to show results.
Birmingham researchers, led by Professor Pola Goldberg Oppenheimer from the School of Chemical Engineering, designed and developed the novel diagnostic hand-held device to assess patients as soon as injury occurs.
It is fast, precise and non-invasive for the patient, causing no additional discomfort, can provide information on the severity of the trauma, and will be suitable to be used on-site to assess TBI.
Professor Pola Goldberg Oppenheimer said: “Early diagnosis of TBI is crucial, as life-critical decisions on treatment must be made with the first ‘golden hour’ after injury. However current diagnostic procedure relies on observation by ambulance crews, and MRI or CT scans at a hospital – which may be some distance away.”
The device works by scanning the retina where the optic nerve sits. Since the optic nerve is so closely linked to the brain, it carries the same biological information in the form of protein and lipid biomarkers.
These biomarkers exist in a very tightly regulated balance, meaning even the slightest change may have serious effects on the ‘brain-health’. TBI causes these biomarkers to change, indicating that something is wrong.
Previous research has demonstrated the technology can accurately detect the changes in animal brain and eye tissues with different levels of brain injuries — picking up the slightest changes.1,2,3
The device detailed in the current paper detects and analyses the composition and balance of these biomarkers to create ‘molecular fingerprints’.
The current study details the development, manufacture, and optimisation of a proof-of-concept prototype, and its use in reading biochemical fingerprints of brain injury on the optic nerve, to see whether it is a viable and effective approach for initial ‘on the scene’ diagnosis of TBI.
The researchers constructed a phantom eye to test its alignment and ability to focus on the back of the eye, used animal tissue to test whether it could discern between TBI and non-TBI states, and also developed decision support tools for the device, using AI, to rapidly classify TBIs.
The device is now ready for further evaluation including clinical feasibility and efficacy studies, and patient acceptability.
The researchers expect the diagnostic device to be developed into a portable technology which is suitable for use in point-of-care conditions capable to rapidly determine whether TBI occurs as well as classify whether it is mild, moderate or severe, and therefore, direct triage appropriately and in timely manner.
Cross-sectional diagram of the NexGen 7T scanner, showing the new Impulse head-only gradient coil (green) and receiver-transmit coil (white) resting on a movable bed (brown) and connected to an electronic interface (blue) containing nearly a thousand wires (blue) that extend out of the magnet. Credit: Bernhard Gruber, MGH Harvard
An intense international effort to improve the resolution of magnetic resonance imaging (MRI) for studying the human brain has culminated in an ultra-high resolution 7 Tesla scanner that records up to 10 times more detail than current 7T scanners and over 50 times more detail than current 3T scanners, the mainstay of most hospitals.
This next generation or NexGen 7T functional MRI (fMRI) scanner can resolve features 0.4mm across, compared to the 2–3mm typical of today’s standard 3T fMRIs. It is described in a paper published in Nature Methods.
“The NexGen 7T scanner is a new tool that allows us to look at the brain circuitry underlying different diseases of the brain with higher spatial resolution in fMRI, diffusion and structural imaging, and therefore to perform human neuroscience research at higher granularity,” said David Feinberg, the director of the project to build the scanner. “The ultra-high resolution scanner will allow research on underlying changes in brain circuitry in a multitude of brain disorders, including degenerative diseases, schizophrenia and developmental disorders, including autism spectrum disorder.”
The improved resolution will help neuroscientists probe the neuronal circuits in different regions of the brain’s neocortex and allow researchers to track signals propagating from one area of the cortex, and perhaps discover underlying causes of developmental disorders. This could lead to better ways of diagnosing brain disorders, perhaps by identifying new biomarkers that would allow diagnosis of mental disorders earlier or, more specifically, in order to choose the best therapy.
“Normally, MRI is not fast enough at all to see the direction of the information being passed from one area of the brain to another,” Feinberg said. “The scanner’s higher spatial resolution can identify activity at different depths in the brain’s cortex to indirectly reveal brain circuitry by differentiating activity in different cell layers of the cortex.”
This is possible because neuroscientists have found in vision brain areas that the superficial and deepest cortex layers incorporate ‘top-down’ circuits, that is, they receive information from higher cortical brain areas, whereas the middle cortex involves ‘bottom-up’ circuitry, receiving sensory input. Pinpointing the fMRI activity to a specific depth in the cortex lets neuroscientists track the flow of information throughout the brain and cortex.
With the higher spatial resolution, neuroscientists will be able to home in on the activity of something on the order of 850 individual neurons within a single voxel – a 3D pixel – instead of the 600 000 recorded with standard hospital MRIs, said Silvia Bunge, a UC Berkeley professor of psychology who is one of the first to use the NexGen 7T to conduct research on a human brain.
“We were able to look at the layer profile of the prefrontal cortex, and it’s beautiful,” said Bunge, who studies abstract reasoning. “It’s so exciting to have this state-of-the-art, world-class machine.”
For William Jagust, a UC Berkeley professor of public health who studies the brain changes associated with Alzheimer’s disease, the improved resolution could finally help connect the dots between observed changes due to Alzheimer’s that occur in the brain – abnormal clumps of protein called beta amyloid and tau – and changes in memory.
“We know that part of the memory system in the brain degenerates as we get older, but we know little about the actual changes to the memory system – we can only go so far because of the resolution of our current MRI systems,” said Jagust. “With this new scanner, we think we’re going to be able to take apart a lot more carefully exactly where things have gone wrong. This could help with diagnosis or predicting outcomes in normal people.”
Jack Gallant, a UC Berkeley professor of psychology, hopes the scanner will help neuroscientists discover how functional changes in the brain lead to developmental and mental disorders such as dyslexia, autism and schizophrenia, or that result from neurological disorders, such as dementia and stroke.
“Mental disorders have an enormous impact on individuals, families and society. Together they represent about 10% of the US GDP. Mental disorders are fundamentally disorders of brain function, but functional measures are not used currently to diagnose most brain disorders or to look to see if a treatment’s working. Instead, these disorders are diagnosed behaviourally. This is a weak approach, because there are a lot of different mental brain states that can lead to exactly the same behaviour,” Gallant said. “What we need is more powerful MRI machines like this so that we can map, at high resolution, how information is represented in the brain. To me this is the big potential clinical benefit of ultra-high resolution MRI.”
Funding initiatives lead to ‘quantum leap’
The breakthrough came about through $22 million of funding from various government and private sector sources.
Incorporating newly developed hardware technology from those groups, Siemens collaborated with Feinberg’s team to rebuild a conventional 7 Tesla MRI scanner delivered to UC Berkeley in 2000 to improve the spatial resolution in pictures captured during brain scans.
“There’s been a large increase throughout the world of sites that use 7T MRI scanners, but they were mostly for development and were difficult to use,” said Nicolas Boulant, a physicist visiting from the NeuroSpin project at the University of Paris in Saclay, where he leads the team that operates the world’s only 11.7 Tesla MRI scanner, the strongest magnetic field employed to date. “David’s team really put together many ingredients to make a quantum leap at 7 Tesla, to go beyond what was achievable before and gain performance.”
Boulant hopes to adapt some of the new ingredients in the NexGen 7T – in particular, redesigned gradient coils – to eventually achieve even better resolution with the 11.7 Tesla MRI scanner. The gradient coils generate a rising magnetic field across the brain so that each part of the brain sees a different field strength, which helps to precisely map brain activity.
“The higher the magnetic field, the more difficult it is to really grab the potential promised by these higher-field MRI scanners to see finer details in the human brain,” he said. “You need all this peripheral equipment, which needs to be on steroids to meet those promises. The NexGen 7T is really a game-changer when you want to do neuro MRI.”
To reach higher spatial resolution, the NexGen 7T scanner had to be designed with a greatly improved gradient coil and with larger receiver array coils – which pick up the brain signals – using from 64 to 128 channels to achieve a higher signal-to-noise ratio (SNR) in the cortex and faster data acquisition. All these improvements were combined with a higher signal from the ultra-high field 7T magnet to achieve cumulative gains in the scanner performance.
The extremely powerful gradient coil is the first to be made with three layers of wire windings. Designed by Peter Dietz at Siemens in Erlangen, Germany, the “Impulse” gradient has 10 times the performance of gradient systems in current 7T scanners. Mathias Davids, then a physics graduate student at Heidelberg University in Mannheim, Germany, and a member of Feinberg’s team, collaborated with Dietz in performing physiologic modelling to allow a faster gradient slew rate – a measure of how quickly the magnetic field changes across the brain – while remaining under the neuronal stimulation thresholds of the human body.
“It’s designed so that the gradient pulses can be turned on and off much quicker – in microseconds – to record the signals much quicker, and also so the much higher amplitude gradients can be utilised without stimulating the peripheral nerves in the body or stimulating the heart, which are physiologic limitations,” Feinberg said.
A second key development in the scanner, Feinberg said, is the 128-channel receiver system that replaces the standard 32 channels. The large receiver coil arrays built by Shajan Gunamony of MR CoilTech in Glasgow, UK, gave a higher signal-to-noise ratio in the cerebral cortex and also provided higher parallel imaging acceleration for faster data acquisition to encode large image matrices for ultra high resolution fMRI and structural MRI.
To take advantage of the new hardware technology, Suhyung Park, Rüdiger Stirnberg, Renzo Huber, Xiaozhi Cao and Feinberg designed new pulse sequences of precisely timed gradient pulses to rapidly achieve ultra high resolution. The smaller voxels, measured in units of cubic millimetres and less than 0.1 microlitre, provide a 3D image resolution that is 10 times higher than that of previous 7T fMRIs and 125 times higher than the typical hospital 3T MRI scanners used for medical diagnosis.
Voxel-perfect resolution
The most common MRI scanners employ superconducting magnets that produce a steady magnetic field of 3 Tesla – 90 000 times stronger than Earth’s magnetic field and 3000 times stronger than a fridge magnet.
“A 3T fMRI scanner can resolve spatial details with a resolution of about 2 to 3mm. The cortical circuits that underpin thought and behaviour are about 0.5mm across, so standard research scanners cannot resolve these important structures,” Gallant said.
In contrast, fMRI focuses on blood flow in arteries and veins and can vividly distinguish oxygenated haemoglobin funnelling into working areas of the brain from deoxygenated haemoglobin in less active areas. This allows neuroscientists to determine which areas of the brain are engaged during a specific task.
But again, the 3mm resolution of a 3T fMRI can distinguish only large veins, not the small ones that could indicate activity within microcircuits.
The NexGen 7T will allow neuroscientists to pinpoint activity within the thin cortical layers in the grey matter, as well as within the narrow column circuits that are organised perpendicular to the layers. These columns are of special interest to Gallant, who studies how the world we see is represented in the visual cortex. He has actually been able to reconstruct what a person is seeing based solely on recordings from the brain’s visual cortex.
“The machine that David has built, in theory, should get down to 500 microns, or something like that, which is way better than anything else – we’re very near the scale you would want if you’re getting signals from a single column, for example,” Gallant said. “It’s fantastic. The whole thing about MRI is how big is the little volumetric unit, the voxel […] that’s the only thing that matters.”
For the moment, NexGen 7T brain scanners must be custom-built from regular 7T scanners but should be a lot cheaper than the $22 million required to build the first one.
Feinberg said that UC Berkeley’s NexGen 7T scanner technology will be disseminated by Siemens and MR CoilTech Ltd.
“My view is that we may never be able to understand the human brain on the cellular synaptic circuitry level, where there are more connections than there are stars in the universe,” Feinberg said. ” But we are now able to see signal patterns of brain circuits and begin to tease apart feedback and feed forward circuitry in different depths of the cerebral cortex. And in that sense, we will soon be able to understand the human brain organisation better, which will give us a new view into disease processes and ultimately allow us to test new therapies. We are seeking a better understanding and view of brain function that we can reliably test and reproducibly use noninvasively.”
A 2000-year-old practice by Chinese herbalists – examining the human tongue for signs of disease – is now being embraced by computer scientists using machine learning and artificial intelligence.
Tongue diagnostic systems are fast gaining traction due to an increase in remote health monitoring worldwide, and a new paper in AIP Conference Proceedings provides more evidence of the increasing accuracy of this technology to detect disease.
Engineers from Middle Technical University (MTU) in Baghdad and the University of South Australia (UniSA) used a USB web camera and computer to capture tongue images from 50 patients with diabetes, renal failure and anaemia, comparing colours with a data base of 9000 tongue images.
Using image processing techniques, they correctly diagnosed the diseases in 94 per cent of cases, compared to laboratory results. A voicemail specifying the tongue colour and disease was also sent via a text message to the patient or nominated health provider.
MTU and UniSA Adjunct Associate Professor Ali Al-Naji and his colleagues have reviewed the worldwide advances in computer-aided disease diagnosis, based on tongue colour.
“Thousands of years ago, Chinese medicine pioneered the practice of examining the tongue to detect illness,” Assoc Prof Al-Naji says.
“Conventional medicine has long endorsed this method, demonstrating that the colour, shape, and thickness of the tongue can reveal signs of diabetes, liver issues, circulatory and digestive problems, as well as blood and heart diseases.
“Taking this a step further, new methods for diagnosing disease from the tongue’s appearance are now being done remotely using artificial intelligence and a camera – even a smartphone.
“Computerised tongue analysis is highly accurate and could help diagnose diseases remotely in a safe, effective, easy, painless, and cost-effective way. This is especially relevant in the wake of a global pandemic like COVID, where access to health centres can be compromised.”
Diabetes patients typically have a yellow tongue, cancer patients a purple tongue with a thick greasy coating, and acute stroke patients present with a red tongue that is often crooked.
A 2022 study in Ukraine analysing tongue images of 135 COVID patients via a smartphone showed that 64% of patients with a mild infection had a pale pink tongue, 62% of patients with a moderate infection had a red tongue, and 99% of patients with a severe COVID infection had a dark red tongue.
Previous studies using tongue diagnostic systems have accurately diagnosed appendicitis, diabetes, and thyroid disease.
“It is possible to diagnose with 80% accuracy more than 10 diseases that cause a visible change in tongue colour. In our study we achieved a 94% accuracy with three diseases, so the potential is there to fine tune this research even further,” Assoc Prof Al-Naji says.
A new artificial intelligence (AI)-based method can provide as much information on subtle neurodegenerative changes in the brain captured by computed tomography (CT) as compared to magnetic resonance imaging (MRI). The method, reported in the journal Alzheimer’s & Dementia, could enhance diagnostic support, particularly in primary care, for conditions such as dementia and other brain disorders.
Compared to MRI, which requires powerful superconducting magnetics and their associated cryogenic cooling, computed tomography (CT) is a relatively inexpensive and widely available imaging technology. CT is considered inferior to MRI when it comes to reproducing subtle structural changes in the brain or flow changes in the ventricular system. Certain imaging must therefore currently be carried out by specialist departments at larger hospitals equipped with MRI.
AI trained on MRI images
Created with deep learning, a form of AI, the software has been trained to transfer interpretations from MRI images to CT images of the same brains. The new software can provide diagnostic support for radiologists and other professionals who interpret CT images.
“Our method generates diagnostically useful data from routine CT scans that, in some cases, is as good as an MRI scan performed in specialist healthcare,” says Michael Schöll, a professor at Sahlgrenska Academy who led the work involved in the study, carried out in collaboration with researchers at Karolinska Institutet, the National University of Singapore, and Lund University
“The point is that this simple, quick method can provide much more information from examinations that are already carried out on a routine basis within primary care, but also in certain specialist healthcare investigations. In its initial stage, the method can support dementia diagnosis, however, it is also likely to have other applications within neuroradiology.”
Reliable decision-making support
This is a well-validated clinical application of AI-based algorithms, and has the potential to become a fast and reliable form of decision-making support that effectively reduces the number of false negatives. The researchers believe that this solution can improve diagnostics in primary care, optimising patient flow to specialist care.
“This is a major step forward for imaging diagnosis,” says Meera Srikrishna, a postdoctor at the University of Gothenburg and lead author of the study.
“It is now possible to measure the size of different structures or regions of the brain in a similar way to advanced analysis of MRI images. The software makes it possible to segment the brain’s constituent parts in the image and to measure its volume, even though the image quality is not as high with CT.”
Applications for other brain diseases
The software was trained on images of 1117 people, all of whom underwent both CT and MRI imaging. The current study mainly involved healthy older individuals and patients with various forms of dementia. Another application that the team is now investigating is for normal pressure hydrocephalus (NPH).
With NPH, the team has obtained new results indicating that the method can be used both during diagnosis and to monitor the effects of treatment. NPH is a condition that occurs particularly in older people, whereby fluid builds up in the cerebral ventricular system and results in neurological symptoms. About two percent of all people over the age of 65 are affected. Because diagnosis can be complicated and the condition risks being confused with other diseases, many cases are likely to be missed.
“NPH is difficult to diagnose, and it can also be hard to safely evaluate the effect of shunt surgery to drain the fluid in the brain,” continues Michael. “We therefore believe that our method can make a big difference when caring for these patients.”
The software has been developed over the course of several years, and development is now continuing in cooperation with clinics in Sweden, the UK, and the US together with a company, which is a requirement for the innovation to be approved and transferred to healthcare.
