Editorial, Student's Voice

How Can AI Change the World: Drawing Lessons from Medicine

Author | Christian Rebolledo | Grade 8 | USA
Healthcare and Education share many overlapping sentiments. For example, both are fundamentally service based industries, both seek the betterment of their charges and both are largely viewed as public expenditures. In an evolving world however, questions of efficiency and expediency are inveigling the minds and hearts of policy makers. In healthcare, doctors are increasingly priding themselves on employing such efficiency, and with today’s technology, they are getting much better at it. But to jumpstart a new and improved era of medicine, implementing and developing AI could be the next big step. By improving the accuracy of disease diagnosis, management, and treatment, artificial intelligence will revolutionize the medical world. It is worth evaluating these advances, as they may be laying the groundwork for similar innovations in Education and beyond.

Artificial Intelligence:

Artificial intelligence is the ability of a machine or program to perform tasks often associated with only humans. It is an interdisciplinary concept that revolves around giving systems the gift of intelligence so that they might reason, discover meaning, generalize, or learn from past experience. AI is a tool for computation and interpretation of data, making it beneficial to almost every field.

The tool of artificial intelligence is very broad. Like literature and its many genres, AI has many sub-layers that can be used for different things. This makes this technology very versatile.
Under general AI, there is machine learning which analyzes data and learns from patterns in the data on their own. A subset of machine learning is deep learning, which utilizes neural networks or layers to understand data in a deeper way.

Neural networks filter data and determine its importance as it tries to understand it. These networks mimic how our brains use data and implement the importance of it in a given area.
All of these subsets of AI make this tool applicable to many different fields. Different disciplines can utilize different aspects of artificial intelligence to suit their needs. For example, autonomous vehicles use convolutional neural networks, a form of deep learning, to understand visual imagery in the surrounding area. Utilizing some of these subsets of artificial intelligence, the medical world will be able to advance at a very high speed.

Problems with Healthcare

There are many problems with healthcare that diminish its ultimate goal of efficiently treating patients. These issues include not detecting diseases early enough, malfunction of medicines in patients, the inaccuracy of diagnosis, and sometimes the improper practice of a procedure.

These issues are all caused by one thing: human error. Doctors, of course, know what they are doing. They know the inner workings and functionality of a person’s body. However, even with these beneficial sets of skills, medical practitioners will still inevitably make mistakes. With the help of AI, doctors can find a better course of action relying on a person’s data that they could not do so themselves. Since artificial intelligence can see patterns humans would never be able to see, it would be a great benefit.

For artificial intelligence to diagnose certain diseases like lung cancer, a study by Google shows that a three-dimensional neural network can scan for abnormalities in radiological scans. A convolutional neural network, which runs pixel groups of an image through a filtered system, can help in this particular sense since neural networks are especially beneficial in finding patterns. Understanding atypical patterns can help AI more reliably diagnose certain health conditions.

Early detection of diseases like cancer and Alzheimer’s is something the medical world is trying to progress in. Detecting these diseases early enough would mean a lesser healthcare cost for the patient and faster treatment and recovery. Despite this ambition, medicine is still struggling with this.

According to the National Academy of Medicine, almost 20% of even early diseases are misdiagnosed, which accounts for between 6–17% of adverse hospital events. These numbers may be low, but considering the number of lives that could have been saved, it is still worthwhile. In order to ensure that the results the AI integration produces are accurate and do not contribute to misdiagnoses, it is important for them to train on example data first and understand the importance of inputted data later on.

Given that these diseases start off very unnoticeably, trying to find signs of them is definitely a challenge, especially when the patient exhibits no symptoms. Imagine trying to find a cancer cell when you don’t even know to look for one in the first place. This is where AI plays a role. By inputting different scans of a person’s body through a neural network, artificial intelligence could better predict the possibility of a disease without the patient even coming to the doctor with severe concerns.

With AI guiding procedures, doctors could more easily conduct them with higher accuracy and success, which could all in all increase medical efficiency.

Effects on Doctors

Although it may seem that implementing AI in healthcare would diminish the purpose of the doctor, this simply is not the case. Artificial intelligence is a tool to be utilized, not a means of execution for its ideas. The systems when applied to healthcare would essentially suggest diagnoses of a patient and the best route of action based on its machine learning. Doctors would definitely oversee this and would only utilize AI if it was necessary to the case. However, it would increase efficiency.

In addition to bettering the efficiency of healthcare, AI can provide feedback, guidance, and support to doctors that can assure them of their ideas. Doctors would also have the opportunity to learn from the patterns AI detects since they would not have seen them before.

“AI could also simplify some procedures and give physicians more time to focus on patient infections and possibly reduce unnecessary exams and costs.”
– Dr. Theodoros Zanos

Focusing on patient infections can improve medical practice given that SSI’s (Surgical Site Infections) account for 20% of all hospital-acquired infections and on average, cost $3.5 billion to $10 billion annually. All-in-all, if implemented correctly, AI in healthcare would serve as a beneficial tool and resource for medical practitioners, not a replacement.

Restrictions on Advancement

Artificial intelligence can be the cure to healthcare’s clinical and technological issues. However, the algorithms being utilized today still are not at the point of full optimization. This is because the data used at one clinic cannot be used at another.

Since AI is powered by data input, new algorithms need to be made for different hospitals dealing with different sets of data. For example, if one is a general hospital, the compartmentalized data will not be the same for a cancer research center. This is why people are currently working on advancing the differentiation between algorithms and the versatility of similar programs.
However, with these factors in mind, it will take a while for these changes to be employed given that the application for AI in healthcare is not commonly spread throughout the realm of artificial intelligence.

The Future of AI in Healthcare

Artificial intelligence is a system that can revolutionize the health industry. By utilizing its powers of computation, analyzation, and interpretation, the medical world can advance its efficiency. Currently, AI is being advanced for mass usage.

In the future, AI in healthcare would be able to keep and review medical records, do populational analytics, create diagnoses, and even develop therapeutic drug designs. The potential for artificial intelligence in the medical world is enormous and could impact billions.

Today, technology is advancing at a supersonic rate. The way industries are producing and implementing these advancements is astounding. Artificial intelligence is one of these many technologies that can one-day advance society as a whole, especially in the world of medicine.