Unleash

According to FDA, Artificial Intelligence has been broadly defined as the science and engineering of making intelligent machines, especially intelligent computer programs (McCarthy, 2007). Artificial intelligence can use different techniques, including models based on statistical analysis of data, expert systems that primarily rely on if-then statements, and machine learning. Machine Learning is an artificial intelligence technique that can be used to design and train software algorithms to learn from and act on data. It is abbreviated as AI/ML.

There are several ways to categorize types of AI/ML. One perspective considers the level of supervision, distinguishing between supervised ML, unsupervised ML, or semi-supervised ML. Another categorization is based on different models, with well-known examples like Deep Learning or Neural Networks. Additionally, machine learning can be viewed in terms of black-box vs. white-box approaches. At ClearSky, we utilize a variety of AI and ML techniques, including widely known methods like Deep Learning and neural networks. Our specialty lies in Evolutionary Algorithms (abbreviated as EAs).

Inspired by the notion of survival of the fittest from Darwinian evolution, Evolutionary Algorithms (EAs) form a subset of Machine Learning and serve as a powerful method for inducing classifiers. The core idea behind EAs lies in their ability to select the best-performing classifiers through a repeated process of variation and selection, emulating the mechanisms of biological evolution to solve specific, often complex problems, such as the diagnosis of Parkinson’s disease. Evolutionary algorithms offer a distinct advantage in efficiency when handling limited data. Unlike other machine learning methods, such as Deep Learning, that require extensive labeled datasets, EAs demonstrate robust performance even with relatively small amounts of data. This makes them an ideal solution for specialized medical conditions like Parkinson's or Huntington's diseases. Another notable advantage of EAs is their classification as 'white-box' machine learning. This designation signifies that the data processed by evolutionary algorithms is both explainable and consistent. This transparency enhances the interpretability of the model's decision-making process, a crucial aspect in healthcare.

A classifier, in the context of machine learning, is a model or algorithm that is trained to categorize input data into predefined classes or categories. The primary goal of a classifier is to learn patterns and relationships within the data so that it can accurately assign new, unseen instances to the appropriate class. In the case of ClearSky, we leverage Evolutionary Algorithms (EAs) to identify the optimal classifier. This classifier is adept at assigning the collected movement data into categories such as Parkinson's or non-Parkinson's, thereby assisting in the diagnosis of Parkinson's disease.

With the advancement of digital technologies in healthcare, the term 'digital biomarker' has emerged. It is commonly referred to for measurable and quantifiable indicators obtained from digital sources like mobile devices, wearables, and sensors. At ClearSky, we embrace a philosophy centered on leveraging modern technologies for the non-invasive collection of movement data. This approach enables us to provide an affordable alternative to the often expensive and invasive methods traditionally employed in the diagnosis of neurodegenerative diseases. The non-invasive approach to collecting movement data, encompassing hand movements to eye saccades, is particularly relevant to our focus on neurological conditions associated with movement disorders. These include Dystonia, Essential Tremor, Huntington's Disease, Multiple System Atrophy (MSA), and, most notably, Parkinson's—the most common movement disorder.