A new Artificial Intelligent (AI) system reads prostate cancer scans as well as radiologists with over 10 years of experience. According a new study conducted by UCLA researchers, the system evaluates Magnetic Resonance Imaging (MRI) with 80.5% accuracy, compared to 83.9% accuracy of seasoned radiologists.
So, yes the radiologists are still better at their job than their counterpart AI. However, there is a reason for all prostate cancer patients to cheer due to this news. To specialize in radiology, experts are required to be trained with over 1000 scans. These experts correctly need to determine whether a certain cancer is tumorous or benign. However, many hospitals do not have the necessary resources to provide specialized training to radiologists.
Hence, this new AI system called FocalNet promises a boon for radiology field using a neural network. The network consists of an algorithm which is a compilation of over a million variables. During the study, the system looked at 417 scans of men and women alike. These helped the AI learn about the assessment and classification of tumors. Later, the research team compared the results of its diagnosis with radiologists over 10 years of experience.
The study is a glimpse into the future, where artificial intelligence system save time and potentially offer diagnostic guidance to inexperienced radiologists.
AI A Bright New Horizon for Commerce
AI systems like FocalNet are a bright horizon for commercial advancements. Recently, Google showcased its AI system, which can communicate naturally with restaurants to book a reservation. The system was able to understand absolutely minor nuisances in human communication from a complete stranger to further its message.
Similarly, AI systems are taking on customer service jobs to guide irate and hopeful customers alike. Systems like FocalNet are at this point, just the beginning of a more reliable and standard benchmark for customer service around the world.