Artificial Intelligence in Analyzing Medical Imaging in Detecting Cancers

Challenging to diagnose early on due to its nature of being our own cells; we likely only realize we have it when it gets bad. Artificial Intelligence is a trending field in computer science whose relevance is attributed to its ability to be a solution to almost every problem. Research is trying to implement Artificial Intelligence in diagnosing cancer from image samples of patients. Through the PRISMA protocol, we systematically obtained literature from 2018 until now for a systematic literature review. Various AI architectures and features are presented and discussed. We also examine the overall reception of research into the application of AI in clinical practices. The analysis and results of this paper are meant to serve as a basis for anyone deciding if AI would be reliable for clinical practice. Our results show that the AI models have enough capability to be a feasible solution to assist doctors in diagnosis, and many researchers show confidence in future applications of AI in the clinic. In our statistical analysis, even when their technique is different, the different implementations perform similarly to others, they have accuracies scores above 80 and they deviate similarly, except for SVM. The analysis and results of this paper are meant to serve as a basis for anyone deciding if AI would be reliable for clinical practice. Our results show that the AI models have enough capability to be a feasible solution to assist doctors for diagnosis, and many researchers show confidence in future applications of AI in the clinic.