Physics

• For decades, scientists have believed that we need to map this intricate connectivity in detail to understand how the structured patterns of activity defining our thoughts, feelings and behaviour emerge.
• The conventional view is that specific thoughts or sensations elicit activity in specific parts of the brain.

Function follows form

• We uncovered this close relationship between shape and function by examining the natural patterns of excitation that can be supported by the anatomy of the brain.
• In these patterns, called “eigenmodes”, different parts of the brain are all excited at the same frequency.
• In a similar way, the brain has its own preferred patterns of excitation, which are determined by its anatomical and physical properties.

A tale of two brains

• This perspective views the brain as a collection of discrete regions, each specialised for a specific function, such as vision or speech.
• An alternative view, embodied by an approach to modelling brain activity called neural field theory, eschews this division of the brain into discrete areas.

Comparing the two views

• To compare the two views of the brain, we tested how easily the conventional, discrete view and the continuous, wave-based view can explain more than 10,000 different maps of brain activity.
• We found that eigenmodes of brain shape – not connectivity – offer the most accurate account of these different activation patterns.

Brain waves and icebergs

• The model only uses the shape of the brain to constrain how the waves evolve through time and space.
• We also found that most of the 10,000 different brain maps that we studied were associated with activity patterns spanning nearly the entire brain.
• Rather than focusing solely on how signals pass between discrete regions, we should also investigate how waves of excitation travel through the brain.

A new approach to brain mapping

• Instead, typical brain mapping methods rely on complex statistics to quantify brain activity without any reference to the underlying physical and anatomical basis of those patterns.
• Our discovery also offers immediate practical benefits, since eigenmodes of brain shape are much simpler to quantify than those of brain connectivity.
• This new approach opens possibilities for studying how brain shape affects function through evolution, development and ageing, and in brain disease.

• They revealed the mechanism underlying each step in gene transcription, how proper gene transcription promotes health, and how dysregulation causes disease.
• At today's press conference in Hong Kong, The Shaw Prize Foundation announced the Shaw Laureates for 2023.
• The Shaw Prize consists of three annual prizes: Astronomy, Life Science and Medicine, and Mathematical Sciences, each bearing a monetary award of US$1.2 million.
• This will be the twentieth year that the Prize has been awarded and the presentation ceremony is scheduled for Sunday, 12 November 2023 in Hong Kong.

• The collaboration will leverage XtalPi's integrated AI capabilities and robotics platform to de novo design and deliver drug candidates for an undisclosed target.
• XtalPi has established itself as an industry leader in combining mutually informative AI "dry lab" algorithms with large-scale "wet lab" robotics to empower pharmaceutical innovation.
• XtalPi's AI + robotics platform is supercharging the shift of pharmaceutical R&D from labor-intensive trial-and-error research to a computation and automation-intensive model, empowering scientists to achieve more with less.
• We are honored that Lilly has chosen XtalPi's AI + robotics drug R&D platform as a partner in achieving more fruitful pharmaceutical innovation and bringing much-needed treatments to patients worldwide."

• Many systems that require strong security are either already underpinned by machine learning or soon will be.
• We propose that the integration of quantum computing in these models could yield new algorithms with strong resilience against adversarial attacks.

The dangers of data manipulation attacks

• However, they’re also highly vulnerable to data manipulation attacks, which can pose serious security risks.
• Data manipulation attacks – which involve the very subtle manipulation of image data – can be launched in several ways.
• Read more:
  AI to Z: all the terms you need to know to keep up in the AI hype age

  Either way, the consequences of data manipulation attacks can be severe.

How quantum computing can help

• In our article, we describe how integrating quantum computing with machine learning could give rise to secure algorithms called quantum machine learning models.
• Beyond this, quantum machine learning could allow for faster algorithmic training and more accuracy in learning features.

So how would it work?

• In classical computers, which follow the laws of classical physics, bits are represented as binary numbers – specifically 0s and 1s.
• Information in quantum computers is stored and processed as qubits (quantum bits) which can exist as 0, 1, or a combination of both at once.
• On the other, quantum computers could be used to generate powerful adversarial attacks, capable of easily deceiving even state-of-the-art conventional machine learning models.

Limitations to overcome

• The current evidence suggests we’re still some years away from quantum machine learning becoming a reality, due to limitations in the current generation of quantum processors.
• Today’s quantum computers are relatively small (with fewer than 500 qubits) and their error rates are high.
• This month the Australian government launched the National Quantum Strategy, aimed at growing the nation’s quantum industry and commercialising quantum technologies.