Visualizing Neuroscience Through AI: A Systematic Review

Introduction

Intelligence is described in terms of how an object responds to a circumstance or action. Human intelligence is characterized by abilities in data manipulation and accurate pattern identification. Humans can learn to match patterns by retraining their neural networks in the brain. We call this human intellect. The human brain is made up of one million neurons connected in various ways. Since trained neurons are required for learning, seeing, and analyzing, these skills make up intelligence. In order to infuse systems with human intelligence and teach them to observe, learn from, and respond in accordance with that intelligence, artificial intelligence (AI) tries to imbue them with human cognition. For scientists, mathematicians, and researchers working on AI, the mechanical, anatomical, and functional properties of the brain have served as an inspiration. AI has shown promise in the field of neurology. Artificial intelligence (AI) has led to unexpected learning and perception outcomes when it comes to the data carried by brain neurons due to how challenging it is to comprehend brain neurons. Using this information, abilities like pattern matching, inference, object detection, etc. are developed. Artificially intelligent systems (AIS) would be able to reason, learn on its own, reason, and match patterns, and other cognitive functions if neuroscientists succeed in replicating the human brain in AIS. Pandarinath collaborated with David Sussillo, a computational neuroscientist at the Google Brain Team in San Francisco, California, on his study on latent variables. Sussillo asserts that an artificial neural network is merely a crude analogy of how the brain works. For example, it represents synapses as numbers in a matrix while, in reality, they are intricate bits of biological machinery that interact with their neighbors in dynamic ways and use both chemical and electrical activity to send or terminate impulses. Sussillo claims that “a single integer in a matrix is as far from the truth of what a synapse actually is as you could possibly get” (Quan, 2022).

The artificial intelligence system has developed to the point where it can mimic human thought processes and display what those people are thinking. As a result, AIS are utilized not only in the fields of marketing and data analysis but also in the diagnosis and detection of medical conditions. As a result of these devices' accuracy, operating rooms now use them. A large number of robots are educated to do various surgical procedures. These systems have been trained for great levels of precision, making this conceivable. Artificial intelligence, artificial neural networks, machine learning, and deep learning are just a few of the technologies that the writers focus on in this chapter that have their roots in neuroscience. Understanding the development of these methods, which draw inspiration from the human brain, as well as the various fields in which they are used and have achieved state-of-the-art status, is the main goal.

This chapter focusses on the relationships between neuroscience and AI. It also discusses the AI systems that have been developed to achieve various goals such as visualizing objects, pattern matching, mimicking human knowledge and prediction capabilities. It also analyses the various AI systems to understand the parameters linked to the human brain and the activities performed by them.

The chapter is organized as follows: section (ii) discuss about the background of AI and neuroscience (iii) discusses about the literature survey (iv) discuss about the future scope and conclusion.