Google Brain, Intel AI Lab & UC Berkeley Reveal the Power of Motion2Vec

The researchers hope that Motion2Vec will one day assist physicians by sealing relatively simple incisions.

It has come to light that, Motion2Vec, an AI model developed by Google Brain, Intel AI Lab, and UC Berkeley has the ability to learn how to do tasks related with robotic surgery such as suturing, needle-passing, needle inserting, and knot binding by watching surgery clips. The model was tested in a lab with a two-armed da Vinci robot moving a needle through cloth.

What really is Motion2Vec?

According to Infotech Report, Motion2Vec is a representation learning algorithm trained using semi-supervised learning, and it follows in the tradition of similarly named models like Word2Vec and Grasp2Vec, trained with knowledge found in an embedding space. UC Berkeley researchers previously used YouTube videos to train agents to dance, do backflips, and perform a range of acrobatics, and Google has used video to train algorithms to do things like generate realistic video or predict depth using mannequin challenge videos from YouTube.

They also added that the researchers say their work shows that video robotics used in surgery can be improved by feeding them expert demonstration videos to teach new robotic manipulation skills. "Results suggest performance improvement in segmentation over state-of-the-art baselines, while introducing pose imitation on this dataset with cm error 0:94 in position per observation respectively," the paper reads.

How did the researchers get Motion2Vec to learn stitches?

While viewing videos of surgeons doing the job, artificial intelligence learns to resew. The robot is then able to replicate the movements it has observed, which range from insertion and extraction to needles transfer, thanks to the visual observation sessions.

A study showed, to evaluate the system's accuracy, the project's team used surgical videos from the JIGSAWS database. Motion2Vec had an 85.5% performance rate for 78 films watched, according to the information site 20minutes. The findings, however, revealed an average margin of error of 0.94 centimeters.

The robot is currently under experimentation in phase D, but the latter will obviously be of considerable use in an operating theatre once it has indeed been created. However, Dr. Ajay Tanwani, one of the researchers involved in the system development, underlined that the device won't fully replace a human surgeon. Despite the fact that the technology is still in its early stages, the researchers hope that Motion2Vec will one day assist physicians by sealing relatively simple incisions.