Aida Abbasi Marjani - Radiology Department, Paramedical Faculty, Tabriz University of Medical Sciences, Tabriz, Iran

Cancer, as a self-sustaining and adaptive procedure that cooperates dynamically with its microenvironment, remains to foil patients, researchers, and clinicians regardless of significant evolution in consideration of its biological foundations. Artificial intelligence (AI) algorithms, mainly deep learning, have established notable progress in image-recognition tasks. Approaches extending from convolutional neural networks to variational autoencoders have initiated numerous applications in the medical image analysis field, thrusting it onward at a quick step. While at first look AI seem to impend the role of the radiographer, its prevalent approval and operation also offer significant occasions for more independence and self-definition if the occupation effectively makes for, and adjusts to, certain changes to character and principles. AI has been substantially used in the diagnostic procedure of plentiful cancers including head and neck cancer, breast cancer, skin cancer, lung cancer, prostate cancer, etc. The growth and spreading of AI in clinical medicine will enhance our indicative truth and rule-out abilities. Though, unless AI algorithms are accomplished to separate between benign anomalies and clinically expressive lesions, improved imaging sympathy might come at the cost of enlarged wrong positives, as well as confounding states whereby AI discoveries are not allied with conclusions.

cancer, artificial intelligence, imaging, MRI, CT, PET