Artificial intelligence in medicine

Artificial intelligence in medicine

The introduction of artificial intelligence (AI) systems in medicine is one of the most important modern trends in world healthcare. Artificial intelligence technologies are fundamentally changing the global healthcare system, allowing to radically redesign the system of medical diagnostics, the development of new drugs, and generally improve the quality of healthcare services while reducing costs for medical clinics.

2020
 AI systems for predicting risks in pregnant women
In early September 2020, researchers at Carnegie Mellon University unveiled a machine learning algorithm that analyzes placenta samples and calculates a woman's health risk in future pregnancies. The system is designed to help obstetricians-gynecologists, who are very useful in predicting possible complications of future women in labor. According to the authors of the project, their development has already begun to be applied in clinical practice.

Among the most serious prognostic signs, researchers note the damage to the blood vessels of the placenta, which is called decidual vasculopathy. Her presence suggests that the young mother suffered from preeclampsia during pregnancy, a condition that complicates 2-8% of pregnancies and can be fatal to both mother and baby. If preeclampsia is detected early, the patient can be treated before symptoms appear. But because the survey is very time-consuming and requires highly specialized skills, it is rarely done.
The researchers proposed using AI to make assessment more accessible by automatically searching for diseased vessels on placental slides. The team trained their algorithm to identify vascular lesions by providing the AI ??with a vast array of images of placenta samples.
"
Pathologists have been training for years to look for signs of disease in these images, but there are so many pregnant women in hospitals that there is no time to examine each specimen, said researcher Daniel Clymer. - Our AI helps doctors identify images with possible vasculopathy
"
When tested, the algorithm classified lesions more accurately than professional pathologists. However, researchers do not expect the system to replace medical professionals. Instead, AI should increase research bandwidth, lower cost, and make it widely available.

2019
Facial recognition software can recognize a person by an MRI scan of the brain
In late October 2019, it became known that facial recognition software can accurately match photographs of people with MRI scans of the brain.

Conventional facial recognition software correctly identified volunteers in more than 80% of cases, but researchers believe that recruiting more patients will reduce accuracy. It is assumed, however, that such software functions can threaten patient privacy, because laws do not have time to change as quickly as technology. In a NYU report, scientists have already warned of the potential risks of using facial recognition software. The technology is increasingly used by the police to track and trace suspects, but regulations are still seriously lagging behind reality.
It became known that facial recognition software can accurately match photographs of people with MRI scans of the brain
Researchers at the Mayo Clinic found that, thanks to publicly available facial recognition software, they were able to correctly match patient photos to MRI scans in 83% of cases. One of the researchers, Christopher Schwarz, said the team at the Mayo Clinic decided to conduct such a study, noting the high quality of the images used to study the brains of patients with Alzheimer's and dementia. Usually, MRI scans show the outline of the head, including skin and fat, but does not detect bones and hair, but the software was enough for this information.

The study involved 84 volunteers and facial recognition software successfully matched their MRI scans to their photographs. Despite the amazing accuracy of the software, researchers are not ready to trust the results so easily. Radiology professor Eliot Siegel noted that as the sample size increases, the recognition accuracy of photographs is likely to decrease.   

Published by Oksana Kvitka

Comments

No comments yet.

Leave a comment