AI spots five ways your grey matter goes grey; CERN’s search for new physics; Oxygen gives 2D materials a glow-up

Brains, beams, bright materials, and more in this edition of Science Notebook.

MRI images of the brain: With ageing, some parts of the brain degenerate at the same time with other regions, while other parts do so independently. | Photo Credit: Getty Images/iStockphoto

Detecting dementia through brain patterns

AN analysis of 49,482 brain scans has revealed five dominant patterns of wasting, or atrophy, of regions in the brain that can be linked to ageing and neurodegenerative diseases. This international study was conducted by a team of researchers led by Christos Davatzikos, a biomedical-imaging specialist at the University of Pennsylvania and an author of the paper on the work that was published in Nature Medicine.

The brain anatomy changes with ageing and disease and these changes can be seen in magnetic resonance imaging (MRI) scans, with some areas shrivelling or undergoing structural alterations. But the changes are subtle. “The human eye is not able to perceive patterns of systematic brain changes,” said Davatzikos.

Earlier studies showed that machine-learning methods can extract the subtle fingerprints of ageing from MRI data, but their scope was limited and often included data from a small sample. To identify broader patterns, Davatzikos’ team enlarged the scope of its study, which took roughly eight years to complete. The researchers used a deep-learning method called Surreal-GAN developed by Zhijian Yang, the first author of the study. The algorithm was trained on brain MRIs of 1,150 healthy people aged 20-49 and 8,992 older adults, including many with cognitive decline. The algorithm so trained could recognise recurring features of ageing brains. This enabled the creation of an internal model of brain structures that change at the same time and those that tend to change independently.

The model was applied to MRI scans from 49,482 people participating in different ageing and neurological health studies. This analysis yielded patterns of brain atrophy that linked different age-related brain degeneration to combinations of five dominant patterns. There was, though, some variability between individuals with the same condition.

Dementia and its precursor, mild cognitive impairment, for example, had links to three of the five patterns. The researchers also found evidence that the patterns they identified could reveal the likelihood of more future brain degeneration. “If you want to predict progression from cognitively normal status to mild cognitive impairment, one [pattern] was the most predictive by far,” said Davatzikos. “At later stages, the addition of a second enriches your prediction, which makes sense because this kind of captures the propagation of the pathology.” The team also identified patterns linked to diseases such as Parkinson’s and Alzheimer’s. One combination of three patterns was highly predictive of mortality. The authors also found associations between certain patterns and disease risk factors such as alcohol intake and smoking and various genetic and blood-based markers.

Display showing a collision event recorded by the ATLAS detector at an energy of 13.6 TeV, featuring two candidate dis- placed electrons each represented by a track (blue lines) that points to an energy de- posit in the ATLAS calorimeter (green). The inset shows an axial view of the detector, illustrating the electron tracks displaced from the interaction point (red circle) by a few mm. | Photo Credit: ATLAS/CERN

The search for new physics at CERN’s LHC

DESPITE its immense success in describing the fundamental building blocks of matter and their interactions, the Standard Model of particle physics is known to be incomplete. Experiments worldwide are therefore searching for signs of new physics phenomena that would guide physicists towards a more comprehensive theory.

At the 42nd International Conference on High Energy Physics in July in Prague, the ATLAS collaboration, one of the international collaborative experiments at the Large Hadron Collider (LHC) in CERN, Geneva, presented its first results from searches for new physics at record collision energies of 13.6 tera (or 10¹²) electronvolts (TeV), targeting long-lived particles (LLPs) created in proton–proton collisions.

Most searches for new physics look for new particles that decay “promptly” and produce decay products that emanate from the LHC’s proton–proton interaction points. However, beyond-the-Standard-Model physics theories, such as supersymmetry, also predict LLPs that would produce decay products away from the interaction point. Such “displaced” particles require dedicated innovative techniques to reconstruct particle tracks that may have eluded detection in earlier searches.

ATLAS has released the result of a new search for a pair of LLPs, each of which decays into an electron, muon, or tau lepton, resulting in two particle tracks that are “displaced” from the interaction point, a rare signature that could be indicative of new physics. In particular, ATLAS looked for a new signature where one of the long-lived particles travelled far enough before decaying so that only a single electron was detected. This is the first ATLAS search of this type of proton–proton collision data from LHC’s Run 3. For this, ATLAS researchers had enhanced the online collision-event selection, the “trigger”, with the reconstruction of displaced tracks, which enabled the present search for new LLPs.

The event-yields in all search regions were consistent with Standard Model expectations. The results have set the strictest limits yet on the long-lived supersymmetric partners of electrons, muons, and tau leptons. With more data from the LHC and its future upgrade, the High-Luminosity LHC, the quest to find LLPs, magnetic monopoles, and other hypothetical particles will continue—all the while further refining their search techniques and developing new experimental strategies, said the CERN release.

A graphic representation of oxygen-assisted chemical vapour deposition in molybdenum disulphide samples. | Photo Credit: Keerthana S. Kumar/IISc

Adding oxygen to get brighter 2D materials

TWO-DIMENSIONAL (2D) transition metal dichalcogenides (TMDs) like molybdenum disulphide (MoS₂), tungsten disulphide, molybdenum diselenide, and tungsten diselenide are promising candidates for optical, optoelectronic, and energy applications. The key challenges that need to be addressed for commercialisation are scalability, repeatability, and quality control. Mechanical exfoliation (ME) and chemical vapour deposition (CVD) are the two most common synthesis methods used to obtain monolayer TMDs. CVD provides repeatability and scalability but leads to a high density of defects compared with exfoliation.

The researchers—Akshay Singh and collaborators of the IISc, Bengaluru— reported that MoS₂ samples created using oxygen-assisted CVD (O-CVD) are superior in quality over regular ME samples. For example, O-CVD samples showed about 300 per cent increase in room temperature photoluminescence. The team differentiated the effect of defects, oxygen, and strain on the optical properties of MoS₂ by studying samples synthesised using ME and O-CVD after their hexagonal boron nitride covering and encapsulation, which are used to prevent contamination and degradation of the 2D devices. The team unravelled the key effect of oxygen complexes on optical properties and showed the beneficial impact of introducing oxygen in CVD. They showed that O-CVD can be used to synthesise high-quality materials for next generation optoelectronics. According to the IISc release, the study has implications for applications in quantum technologies as well.