Our goal: Discover cognitive outcomes prior to clinical onset

We have validated a methodology about active digital biomarkers after decades of research. Our first publication appeared at the Medicine Meets Virtual Reality conference in 2000 and pioneered the usage of virtual reality based complex activities of daily living and hand micromovements, as a new methodology to measure cognitive function. Several R&D years later, financed by public grants (≅USD 35 million), we validated a novel digital biomarker platform that combines data streams from: hands micromovements & micro-errors, gait micro-errors, posture changes, eye tracking, eye pupil dilation, dual-task micro-errors, visuospatial navigation micro-errors and recently voice parameters. A total of 250 features generated from sensor data recorded at 100Hz.

A novel Digital Biomarkers platform

Artificial intelligence (AI)- and machine learning (ML)-based technologies have the potential to transform healthcare by deriving new and important insights from the vast amount of data generated during the delivery of healthcare every day. One of the greatest benefits of AI/ML in software resides in its ability to learn from real-world use and its capability to improve its performance.

After each usage, our (ML)-based Digital Biomarkers platform delivers a detailed report of our proprietary 250 features from everyday function, organized per cognitive domain for easy translation into everyday practice, disease related outcomes and clinical trial endpoints. Most importantly the clinical management has the ability to track any combination of those individual features in time, thus creating a very sensitive, ecologically valid and continuous assessment of everyday function, in the most realistic and scalable way possible today.

At Altoida we have developed an instrumental activities of daily living (iADL) methodology based on longitudinal clinical studies with 5000+ patients at risk over a period of 8+ years. Our findings, later verified by independent research, suggest that midlife persons with increased risk for later life dementia already show some subtle cognitive changes principally in navigation, micro-movements & visuospatial functions rather than the functions most commonly currently used to investigate preclinical and prodromal AD [1].

NMI: A low-risk (ML)-based algorithm

Leveraging on the continuous stream of data above, our Machine Learning classifier, called Neuro Motor Index (NMI), is a “Software as a Medical Device” (SaMD), defined by FDA and International Medical Device Regulators Forum (IMDRF) as “software intended to be used for one or more medical purposes that perform these purposes without being part of a hardware medical device”. NMI is driving clinical management about very early signs of cognitive dysfunction related to Alzheimer’s type neurodegeneration. This is positioning our NMI at the lowest (I) risk associated with the clinical situation and device use. Use here the FDA table to assess the risk.

NMI: A continuous learning algorithm

While AI/ML-based SaMD exist on a spectrum categorized by risk to patients, they also exist on a spectrum from locked to continuously learning. “Locked” algorithms are those that provide the same result each time the same input is provided. In contrast to a locked algorithm, an adaptive algorithm (e.g., a continuous learning algorithm) changes its behavior using a defined learning process.

Our NMI digital biomarkers platform is allowing continuous algorithm adaptation for a given set of inputs controlled by the clinical management. These algorithm changes are typically implemented and validated through a well-defined and possibly fully automated process that aims at improving NMI performance based on analysis of new or additional data. The adaptation process can be intended to address several different clinical aspects, such as optimizing performance within a specific environment (e.g., based on the local patient population and new biomarkers collected).

The NMI adaptation process follows two stages: learning and updating. NMI “learns” how to change its behavior, for example, from the addition of new cases to the already existing database. The “update” then occurs when the new version of NMI is deployed.

The ability of NMI to learn from real-world feedback (training) and improve its performance (adaptation) makes it uniquely situated among software as a medical device (SaMD) and a rapidly expanding area of research and development.

Digital Biomarkers Based Individualized Prognosis for People at Risk of Dementia: the AltoidaML Multi-site External Validation Study

PubMed.gov, 2020

See the study

Five-year biomarker progression variability for Alzheimer’s disease dementia prediction: Can a complex instrumental activities of daily living marker fill in the gaps?

Alzheimers Dement, Nov 2014

See the study

Digital biomarker‐based individualized prognosis for people at risk of dementia

Alzheimers Association, Aug 2020

See the study

More Information

Altoida Inc., Houston, USA
Altoida AG, Lucerne, Switzerland

contact@altoida.com

Our Terms of Service and our Privacy Policy