The future of seeing a computer instead of a physician might be here sooner than anticipated. Through advances in artificial intelligence (AI), it appears possible to make great strides in advances in the healthcare field.
Just like doctors, AI has to be “taught” how to do its “job.” Humans have to tell the computer exactly what to do. Once AI absorbs all the information, it is tested on its performance. In the medical field, AI is most commonly used in diagnosing patients, improving communications between doctors and patients, prescribing medications and even treating patients.
Use of AI in medicine has already expanded to Omaha. Lisa Bazis is the Chief Information Security Officer at Nebraska Medicine (UNMC). “We use a lot of modeling technology that could be considered AI,” Bazis said. UNMC uses a predictive modeling system that alerts when a patient’s vitals are too high or too low. Though AI development does not stop at just assisting doctors.
Recently there have been great advances in artificial intelligence relating to patient diagnosis. According to PubMed Central, a group of scientists has researched the impact of AI in dermatology. Their research demonstrated that AI systems are capable of not only identifying skin cancer but are able to do it in only a fraction of the time it takes doctors.
In 2018, researchers at Seoul National University developed an AI to detect abnormal cell growth in chest radiographs. This program outperformed 17 out of 18 doctors. As of now, these programs are being used to double check doctor’s diagnosis.
However, there are limitations to AI. One of the first steps for a hospital trying to incorporate an AI system is a data collection. Many are apprehensive about letting AI hold data because of patient privacy. Data breaches are becoming more and more common, allowing patients’ data to be leaked.
AI also requires surveillance. Bazis said “ One of the main cons of AI is it requires human surveillance, and doesn’t take into account socioeconomic backgrounds.”
AI also has a reputation of being a black box. Black boxes are fundamentally a closed off system. The operations being demonstrated are not visible to the user. If AI cannot be accessible and interpretable, then it cannot work in a healthcare system.
Despite the limitations, AI will continue to revolutionize the healthcare system. The algorithm has so far shown benefits to both doctors and patients. The key to incorporating AI in the medical field is the proper understanding of the programming by both physicians and patients. Bazis said “putting everything together, AI, will reduce manual tasks, ease the stress on providers, and allow for a higher standard of medicine.” There are many challenges, but many physicians believe that they are worth overcoming to increase accuracy and efficiency in the medical field.