Artificial intelligence has the potential to improve efficiencies and precision in sleep medicine, resulting in more patient-centered care and better outcomes, according to a new position statement from the American Academy of Sleep Medicine.

Published online as an accepted paper in the Journal of Clinical Sleep Medicine, the position statement was developed by the AASM's Artificial Intelligence in Sleep Medicine Committee. According to the statement, the electrophysiological data collected during polysomnography -- the most comprehensive type of sleep study -- is well-positioned for enhanced analysis through AI and machine-assisted learning.

"When we typically think of AI in sleep medicine, the obvious use case is for the scoring of sleep and associated events," said lead author and committee Chair Dr. Cathy Goldstein, associate professor of sleep medicine and neurology at the University of Michigan. "This would streamline the processes of sleep laboratories and free up sleep technologist time for direct patient care." {module INSIDE STORY}

Because of the vast amounts of data collected by sleep centers, AI and machine learning could advance sleep care, resulting in more accurate diagnoses, prediction of disease and treatment prognosis, characterization of disease subtypes, precision in sleep scoring, and optimization and personalization of sleep treatments. Goldstein noted that AI could be used to automate sleep scoring while identifying additional insights from sleep data.

"AI could allow us to derive more meaningful information from sleep studies, given that our current summary metrics, for example, the apnea-hypopnea index, aren't predictive of the health and quality of life outcomes that are important to patients," she said. "Additionally, AI might help us understand mechanisms underlying obstructive sleep apnea, so we can select the right treatment for the right patient at the right time, as opposed to one-size-fits-all or trial and error approaches."

Important considerations for the integration of AI into the sleep medicine practice include transparency and disclosure, testing on novel data, and laboratory integration. The statement recommends that manufacturers disclose the intended population and goal of any program used in the evaluation of patients; test programs intended for clinical use on independent data; and aid sleep centers in the evaluation of AI-based software performance.

"AI tools hold great promise for medicine in general, but there has also been a great deal of hype, exaggerated claims, and misinformation," explained Goldstein. "We want to interface with industry in a way that will foster safe and efficacious use of AI software to benefit our patients. These tools can only benefit patients if used with careful oversight."

The position statement, and a detailed companion paper on the implications of AI in sleep medicine, are available on the Journal of Clinical Sleep Medicine website.

ITMO Scientists develop an algorithm that makes population history models for people and animals more accurate and easier to generate

Bioinformatics scientists from ITMO University have developed a programming tool that allows for quick and effective analysis of genome data and using it as a basis for building the most probable models of demographic history of populations of plants, animals and people. Operating with complex computational schemes, the software can, with a very high degree of likelihood, predict what history a particular group of living organisms has gone through in the past thousands of years, what periods of mass extinction or mass population growth a population has experienced, and how long it has been in contact with other populations of the same species. The scientists' article dedicated to this methodology has been published in GigaScience.

How to find out when exactly the modern tigers' first ancestors appeared on Earth? When did the two elephant populations split? Is there a difference between the Dama and the Moroccan gazelle? When did the division of the African and the Eurasian homo sapiens occur? The answers to all these questions can be found in the population's demographic history - in other words, the scenario that shows what stages the population went through in the course of its history, whether it underwent any mass extinctions, migrations, or sharp spikes in its numbers. 

Apart from solving fundamental questions, this data can help us in the matters of applied research in the field of ecology and environmental protection. For instance, if some region only has some 800 walruses left, scientists have to understand whether it constitutes a critical decrease or it is a natural population size which has remained constant for several thousand years now, and answer the question of whether valuable resources have to be spent on protecting and saving this species from becoming extinct. {module INSIDE STORY}

The creation of a population's demographic history on the basis of genetic information is a complicated task which requires population geneticists to possess not only knowledge in the field of biology but also programming skills. Such scientists have to garner data and write a code for computing possible models of a population's evolution which could have led to the vast multitude of the genetic information we can witness in this population's representatives today. Up until recently, this was a long process the end result of which relied very heavily on the researcher's initial hypothesis. If it had any defects or the research failed to take some aspect into consideration, the software couldn't correct this initial error and calculated the probability of particular demographic events only within the boundaries predefined by the researcher.

The software developed by a group of ITMO University scientists as part of the Project 5-100 grant programs and with support from JetBrains Research aims to solve this problem. The researchers proposed a programming product which independently and automatically predicts the most probable model of a population's demographic history. At that, it is significantly less dependent on the initial research hypothesis, doesn't require advanced programming skills and produces more accurate results. What is more, the software has the advantage of flexibility, meaning that if the obtained result somehow diverges from archaeological or historical data, you can easily introduce additional limitations into the underlying algorithm to update its hypothesis.

"Using genetic data, our software automatically computes the model it considers optimal," shares Vladimir Ulyantsev. "It looks at the entire volume of the scenarios available. As a scientist, I'll consider the scenarios I deem the most likely, there can be three, five, maybe ten of those. The software, on the other hand, will test all of the models it estimates as probable, this is a much bigger amount. That's why the solutions it comes up with are better than those proposed by people working on the basis of the initial methods. The most beautiful thing here is the method - a genetic algorithm inspired by how evolution happens: species multiply, mutate, with those with the least ability to adapt dying out. In the place of the species we have demographic models and their parameters, and their adaptability is measured on the basis of their similarity with the studied data."

After obtaining this data, the scientists can present it on a map and compare the information indicating that during a particular period a population underwent a migration with archaeological findings and other evidence. These algorithms were used to check a large number of hypotheses and research by evolutionary geneticists. In many cases, the obtained result was much more accurate than that of the initial works.