Macnica / InnerEye

Training AI algorithms with BrainTech

17. Dezember 2020, 18:31 Uhr   |  Andreas Knoll

Training AI algorithms with BrainTech
© InnerEye / Macnica

Das Logo von InnerEye

Macnica is getting into BrainTech: The Japanese technology supplier with DACH office in Ingolstadt has signed a distribution agreement with Israeli BrainTech company InnerEye for Japan, Germany and Brazil.

The term BrainTech is a combination of brain (Brain) and technology (Tech) and describes techniques and processes based on brain research. In 2013, Barack Obama, the former president of the United States, and Shimon Peres, the former president of Israel, launched a BrainTech initiative that has quickly produced promising results, according to Macnica. Complementing this, various brain research projects have been carried out in Japan, and the results will now be incorporated into services and products. BrainTech is the connecting element here, coupling technical facilities and devices with the human brain, and is expected to play an important role in promoting social innovation and digital transformation.

With InnerEye's technology, Macnica now offers a solution for the rapid training of AI algorithms by capturing EEG signals (electroencephalogram) of human experts. The brain activities in problem solving are recorded using the EEG signals and transmitted to the AI algorithms in the form of connotations. The aim is to transfer the relevant experience of experts to the systems of AI, e.g. when looking at medical X-Ray images.

When looking at objects, processing activities are automatically initiated in the human brain for assessment and classification, with the neurons involved generating and distributing electrical signals. These signals can be scanned without invasive techniques on the surface of the scalp and transmitted to computer units for evaluation. The signal patterns differ significantly between known and unknown objects or, for example, whether suspicious objects are detected in the X-Ray image of a piece of luggage or whether it is inconspicuous.

The Solution of InnerEye now displays images on a display to people in quick succession and determines a key figure from the captured EEG signals, i.e. from the cognitive reactions of the viewers, and appends them to the image files as so-called "soft labels".

With this method, AI algorithms can be trained very quickly and charged with the expertise of the human viewers, so that a transfer of the expertise from person to machine takes place, thus the professional competence is efficiently passed on. The EEG patterns can also be used to show concentration, fatigue or distraction and thus evaluate the results with a safety factor.

Seite 1 von 2

1. Training AI algorithms with BrainTech
2. Learning AI Systems

Auf Facebook teilen Auf Twitter teilen Auf Linkedin teilen Via Mail teilen

Verwandte Artikel