Artificial intelligence has learned to diagnose heart disease
Chinese scientists have developed a new artificial intelligence (AI) algorithm that has learned to diagnose heart diseases such as atherosclerosis. The findings of the work are described in an article published in the journal Biomedical Optics Express.
AI technology is based on the analysis of 3D optical coherence tomography (OCT) images of the arteries of the heart. Specialists are increasingly using OCT to study blood vessels, but due to the large amount of data and the complexity of interpreting the results, it is becoming more difficult to make an accurate diagnosis and start treatment in a timely manner.
The mechanism of work consists of two stages. In the first, the AI uses the received OCT image to predict a possible plaque area. It is then refined using an additional algorithm that is based on clinical characteristics that mimic the experience of professional doctors.
The researchers tested AI on 5553 OCTs with atherosclerosis and 3224 images without plaque in the arteries. The computer was able to correctly identify approximately 80 percent of the cases. In the future, experts want to optimize the device: improve the processing of 3D information, as well as add more data to improve the performance of artificial intelligence. In addition, they want to introduce parameters such as the population of the planet and other sets of indirectly affecting criteria into the algorithm.