AI Daily - Artificial Intelligence News

View Original

AI Can Now Identify Brain Injuries

Artificial Intelligence (AI) is a large, popular and promising sector of computer science that develops machines and software program that can intelligently and efficiently care out tasks and make decisions at incredible speeds. However, AI in the healthcare sector is not a recent development, in fact the 1980s intelligent clinical tools were become more widely accepted.

In 1985, the first recognised use of a AI/robot-assisted surgical procedure when the PUMA 560 was used in a neurosurgical biopsy that with greater precision than humans the PUMA 560 successfully completed the procedure, as we know that wasn’t the end for AI in healthcare, today AI is being use to process and interpret large quantities of data in research and in clinical settings, including reading imaging.

Head injury burdens up to 60 million people on average yearly and is one of the most prevalent cause of death in young adults, common procedure is to send these victims in for a CT scan to check for blood in or around the brain and assess whether surgery is needed or not, however only 10 to 15 percent of these patients have lesions that can be seen by the CT scan. AI technology being researched at top research universities could aid to recognise patients who need additional treatment, though any clinical use of the tool would need to be methodically validated.

Talented researchers, from the University of Cambridge and Imperial College London, have clinically validated and verified the AI on a sizeable amount of sets of CT scans and discovered that it was effectively able to identify, section, measure and distinguish dissimilar kinds of brain lesions letting researchers discover more.

The researchers advanced an artificial intelligent device built on an artificial neural network. They exercised the tool on over 600 dissimilar CT scans, screening brain lesions of dissimilar magnitudes and types. The AI was able to categorise separate parts of each image and express whether it was ordinary or not, seeing that the AI may be more reliable and dependable than a human at noticing the smallest of changes with time this would be valuable for forthcoming studies in how head injuries develop and evolve progressively.

Even though the researchers are presently planning to utilise the AI for strictly research, they say with appropriate authentication, it could also be utilised in certain clinical situations, for instance in resource-restricted areas where there are less radiologists than desired. Moreover, researchers suggest that the AI may have a possible utilisation in the E.R, assisting to get patients home as soon as possible.