As we are becoming more aware of artificial intelligence artificial intelligence (AI) and its range of areas it could help in, more advanced AI systems are being created to help in the multiple business sectors. One of the sectors in which AI would help the most is the medical sector. As human mistakes can have big influences in this sector, AI is being created to minimize the risks for patients.
The newest AI system in the medical world is created to help find small and upcoming tumours. The average human success rate for finding smaller tumours is about 65%, so in 35% of the cases a tumour is not found. The new system has been tested and the results show an accuracy rate of 95%. This would be a significant improvement that could save many lives.
The system had to learn how a tumour looked in specific parts of the body to be useful in multiple situations. This has been achieved by creating a model of a brain. Modelling the human way of learning in the system, the opportunity was created for the AI to recognize the tumours in a CT scan. Using electronic parts made it possible to find matches between the CT scan and the studied material to determine the presence of a tumour.
Creating enough of these systems would require many different electronic components and could be quite costly. Before creating the possibility of this AI becoming an industrial device, more research needs to be done to try and improve the accuracy rate even further. As this innovation has the possibility to save lives, it would potentially be interesting for governments to invest in this technology. One way to do this would be by arranging the necessary electronic supplies. Cutting the costs for production, leading to a higher budget in research and development.
As this system has only recently been discovered, it is likely that there is room for improvement. Even though the success rate is currently at 95% already, the system can be upgraded to be able to find 99% of the tumours. If this could be realised, we could say that AI helped us battle diseases.