Some of the researchers have developed a system with the Artificial Intelligence to quickly and easily diagnose and classify brain hemorrhages and to provide with some of the basics of its decisions which is relatively small mage datasets.
According to some of the researchers, such a system could also become an indispensable tool for the hospital emergency departments to evaluate the patients with the symptoms of a potentially life-threatening stroke which even allows the rapid application of the correct treatment.
“Some critics suggest that Machine Learning (ML) algorithms cannot be used in clinical practice because the algorithms do not provide justification for their decisions,” said co-lead author Sehyo June from the Massachusetts General Hospital (MGH) in the US.
To train the system, the research team also began with the 904 head CT scans, each of which consist of around 40 individual images that were also labelled by a team of more than five neuroradiologists as to whether they depicted one of the five hemorrhage subtypes, based on the location just within the brain, or no hemorrhage.
To improve the accuracy of this deep learning system, the team has built in steps mimicking the way radiologists analyze the images, some of the study suggested which is published in the Journal Nature Biomedical Engineering.
Once the model system was created, the team started testing it on the two separate sets of CT scans, a retrospective set taken before the system was developed, which even consisted of 100 scans with and a 100 without intracranial Haemorrhage and a prospective set of 79 scans with and 117 just without the haemorrhage, which is even taken the model was created.
In its analysis of the retrospective set the model system was as accurate in classifying and detecting intracranial hemorrhages as the radiologists that had reviewed the scans had been, the team revealed.
In its analysis of the prospective set, it proved to be even better than non-expert human readers, they added.