Will AI and Human Brain Cells Power Future Biocomputers?

Source: Alexandra_Koch/Pixabay

A new study published in Frontiers in Science introduces biocomputing with “organoid intelligence” (OI), an emerging multidisciplinary field at the intersection of artificial intelligence (AI), biology, stem cells, computing, engineering, and neuroscience.

“The adaptation of OI research models to neurodegenerative diseases would offer the first human-based preclinical model to help us understand and develop effective treatments for these devastating diseases,” wrote lead author Thomas Hartung, professor of environmental health sciences at the Johns Hopkins Bloomberg School of Public Health and Whiting School of Engineering in collaboration with researchers affiliated with Johns Hopkins University, Cortical Labs, the University of California San Diego, and the University of Luxembourg.

“We have coined the term “organoid intelligence” (OI) to describe an emerging field aiming to expand the definition of biocomputing toward brain-directed OI computing, i.e., to leverage the self-assembled machinery of 3D human brain cell cultures (brain organoids) to memorize and compute inputs,” the researchers wrote.

Intelligence-in-a-Dish

Organoids are three-dimensional miniature human organs used for scientific research that were grown in a lab dish from pluripotent stem cells derived from human tissue. One of the challenges in neuroscience and brain disease research is having the ability to conduct research on actual functioning human brain tissue. Many research studies use rodents as a surrogate, which has its drawbacks. An Allen Institute for Brain Science study published in Nature in 2019 found widespread differences between homologous human and mouse cell types such as gene expression, morphology, laminar distributions, and alterations in proportions. The species-specific differences underscore the importance of conducting research using human cells and the value of using organoids as a proxy.

“Brain organoids recapitulate organ histoarchitecture and functionality far more closely than traditional 2D cultures,” wrote the researchers. “They can contain myelinated axons and not only show spontaneous electrophysiological activity but also demonstrate complex oscillatory behavior, and exhibit high cell density and layering patterns, all of which make brain organoids superior to traditional monolayer cultures.”

Organoids are developed using induced pluripotent stem cells, or iPSCs. Pluripotent stem cells have the ability to develop into other types of cells. Stem cells are unspecialized cells that give rise to differentiated cells. In 2018, a team of Tufts-led scientists published their study that showed how they grew a 3D model of human brain tissue that exhibited spontaneous neural activity for at least nine months.

According to the researchers, induced pluripotent stem cells and the creation of 3D brain organoids using the iPSC method will make their organoid intelligence approach feasible.

“The past decade has seen a revolution in brain cell cultures, moving from traditional monolayer cultures to more organ-like, organized 3D cultures—i.e. brain organoids,” the researchers wrote. “These can be generated either from embryonic stem cells or from the less ethically problematic iPSC typically derived from skin samples. The Johns Hopkins Center for Alternatives to Animal Testing, among others, has produced such brain organoids with high levels of standardization and scalability.”

Organoids offer a way to conduct biomedical, neuroscience, pharmaceutical, and life sciences research that may help accelerate the discovery and development of novel treatments for epilepsy, amyotrophic lateral sclerosis (also known as ALS or Lou Gehrig's disease), Alzheimer’s, Parkinson’s, Huntington’s, muscular dystrophy, and other diseases.

AI Machine Learning

Artificial intelligence machine learning will be a critical part of organoid intelligence systems and applied towards decoding the immense datasets of neural activity recordings captured by the novel organoid microelectrode array. The researchers envision using the pattern-recognition capabilities of AI machine learning to help quantify organoid function and architecture changes which involves processes such as autoencoding, signal detection using sequencing and time series models, deconvoluting signal patterns from background noises, and sensor integration to speed up processing based on unsupervised machine learning and dimension reduction. The objective is to map the input and outputs from the organoids’ neurological connections.

Synthetic Biological Intelligence via Biocomputing

Biocomputing, (or biological computing), is an alternative to silicon-based computing. In the fall of last year, Brett Kagan, Chief Scientific Officer at Cortical Labs, and his research colleagues unveiled the first synthetic biological intelligence to show real-time adaptive behavior called “DishBrain,” a brain cell system grown in a laboratory dish that learned to play in a computer game-world inspired by the classic arcade game of “Pong.” Kagan is a co-author of the current study.

Artificial Intelligence Essential Reads

Is OpenAI's Latest LLM Advancement Also a Step Back?

How We Talk About AI Is Important: A Case Study

“We anticipate OI-based biocomputing systems to allow faster decision-making, continuous learning during tasks, and greater energy and data efficiency,” the scientists wrote. “Furthermore, the development of “intelligence-in-a-dish” could help elucidate the pathophysiology of devastating developmental and degenerative diseases (such as dementia), potentially aiding the identification of novel therapeutic approaches to address major global unmet needs.”

Next Generation Brain-Computer Interface

One of the key factors to creating organoid intelligence systems is to have the ability to record the electrophysiological outputs for bi-directional processing between complex neural assemblies. For OI to be successful, requires a new type of brain-computer interface (BCI), also known as brain-machine interface (BMI) that can record neural activity with 3D brain organoids.

“Brain-machine interface technologies have been in progress for at least two decades but remain primitive,” wrote the researchers. “Microelectrode arrays (MEAs) form the heart of many such interfaces since they can be used to both stimulate and record, and offer unprecedented parallelism and individual addressability. However, most are predominantly in a 2D chip-based format, being designed for use with monolayer cell cultures. This represents a likely problem as brain organoids are spherical 3D structures that make limited contact with a 2D MEA chip.”

The researchers also point out that a majority of 2D electrode chips are inflexible, which reduces the quality of the recording due to the improper fit between the rigid electrode chip and the brain.

To enable organoid intelligence systems, the scientists plan to create a new type of flexible 3D microelectrode array interface inspired by electroencephalogram (EEG) caps that resemble swimming caps that record brain activity noninvasively using small electrodes placed on the scalp’s surface. The scientists plan to grow the organoids inside a shell covered with multielectrode nanostructures and probes. This super soft coated shell is designed to be flexible, self-folding and buckling to enable high-resolution 3D spatiotemporal stimulation and recording of the activity across the whole surface of the organoid.

“Ultimately, we aim toward a revolution in biological computing that could overcome many of the limitations of silicon-based computing and AI and have significant implications worldwide,” the scientists wrote.

Copyright © 2023 Cami Rosso All rights reserved.