Ways ChatGPT (and other generative AI apps) Will Change Healthcare Forever, For the Better

Published by Daniel at April 1, 2023

ChatGPT can help medicine by analyzing large volumes of medical data and providing insights and recommendations for physicians, researchers...

ChatGPT's Benefits for Medicine.

ChatGPT can help medicine in many ways. One of the most significant contributions that ChatGPT can make to the field of medicine is by analyzing large volumes of medical data and providing insights and recommendations for physicians, researchers, and healthcare professionals. With its natural language processing abilities, ChatGPT can help healthcare professionals understand complex medical terminologies, identify patterns in medical records, and provide personalized treatment options for patients. ChatGPT can also assist medical research by synthesizing research papers, identifying trends, and generating hypotheses. Additionally, ChatGPT can improve patient care by providing virtual assistants that can answer basic medical questions, schedule appointments, and even assist in medication management. By leveraging the power of new technology and natural language processing, ChatGPT can help revolutionize the field of medicine and improve patient outcomes.

AI catching up to the Human Brain

The human brain’s unique development allows it to become exponentially capable of solving problems such as predicting arithmetic growth (where numbers constantly increase: 1, 2, 3, 4) and comprehending geometric growth (a pattern that grows at a constant ratio: 1, 3, 9, 27). The increasing complexity of number growth and patterns proves to be progressively more complicated for the human brain to grasp. This is where generative AI comes into play, a technology with a growth rate that roughly doubles every two years (following Moore’s Law). This predicted growth would position ChatGPT with 32 times more power in a decade and over 1,000 times more powerful in two decades.

Looking beyond ChatGPT‘s current abilities is the potential for an incredibly powerful tool that will allow for the processing and analysis of data of future generations. This platform’s future analytical and problem-solving abilities will enable technology to reach healthcare providers’ diagnostic skills. (1)

This is where generative AI comes into play, a technology with a growth rate that roughly doubles every two years (following Moore’s Law). This predicted growth would position ChatGPT with 32 times more power in a decade and over 1,000 times more powerful in two decades. Looking beyond ChatGPT‘s current abilities is the potential for an incredibly powerful tool that will allow for the processing and analysis of data of future generations. This platform’s future analytical and problem-solving abilities will enable technology to reach healthcare providers’ diagnostic skills. (1)

Emulating doctors’ clinical decision-making

Generative AI isn’t a crystal ball. Generative AI solves problems, unlike other AI tools. ChatGPT and other generative AI apps can access thousands of terabytes of data in less than a second or “guess” the next word or idea in a series of terms and concepts.

Begin with an extensive database. For physicians, data comes from classroom lectures, published research, and, more significantly, professional experience. For AI, it’s the totality of digitally printed material.
Extract useful information. A physician will recall the relevant information that applies to a patient’s symptoms or pathology. Generative AI will use billions of parameters to identify the correct information and show a proper pathway to the diagnosis.
Use a predictive process to identify the correct and relevant information. Healthcare providers compare possible diagnoses, whereas today’s ChatGPT tests sentences. Both weigh the options and predict the best and most accurate match.

The most significant difference is that healthcare providers have the common sense to seek clarification of doubts and order and analyze tests to complement the previously collected data and create accurate diagnostic and therapeutic conclusions. Successive generations of generative AI can perform this action of ordering or recommending the appropriate laboratory and radiology tests. The new AI-powered interactive chat created by Microsoft has a feature to ask iterative questions and learn from conversations. Similar to medical residents in a hospital, generative AI will initially make mistakes that will be corrected with progressive feedback from an experienced clinician. With more significant experience and computing power will come increased insight and accuracy, as with physicians. In time, ChatGPT will make fewer errors until it can match or surpass medical professionals’ predictive powers (and clinical quality). (2)

Begin with an extensive database. For physicians, data comes from classroom lectures, published research, and, more significantly, professional experience. For AI, it’s the totality of digitally printed material.
Extract useful information. A physician will recall the relevant information that applies to a patient’s symptoms or pathology. Generative AI will use billions of parameters to identify the correct information and show a proper pathway to the diagnosis.
Use a predictive process to identify the correct and relevant information. Healthcare providers compare possible diagnoses, whereas today’s ChatGPT tests sentences. Both weigh the options and predict the best and most accurate match.

Around-the-clock AI-powered medical assistance

Around 40% of Americans suffer two or more chronic illnesses or diseases which affect their health daily. These patients need continuous and, in some cases, rigorous daily care and monitoring. Unfortunately for them, the required medical attention can only be received in a traditional office-based, in-person medical setting. This is where AI can make the most significant difference. Unlike a solo doctor, the next generations of generative AI can monitor patients 24/7 and provide ongoing medical expertise. Doing so would help patients manage chronic illnesses to avoid continuous deterioration like heart disease, hypertension, and diabetes. This would minimize negative outcomes such as heart attacks, strokes, and some types of cancer.

This service would cost a minimum price, impacting the total expenditure in healthcare positively since chronic diseases contribute to 90% of the overall spending in healthcare. Generative AI could help patients with chronic illnesses by

Syncing with wearable medical devices and supportive consumer technologies like Alexa, Bixby, or Chrome to supply around-the-clock monitoring while giving patients individualized, daily helpful health updates.
Comparing hundred of wearable-device readings against the expected ranges preset by each patient’s doctor assembles patient and physician alerts when something’s wrong.
Reminding patients at home when they are due for preventive screenings, Rx refills, or daily exercise (along with other lifestyle improvements). (3)

Unlike a solo doctor, the next generations of generative AI can monitor patients 24/7 and provide ongoing medical expertise. Doing so would help patients manage chronic illnesses to avoid continuous deterioration like heart disease, hypertension, and diabetes. This would minimize negative outcomes such as heart attacks, strokes, and some types of cancer. This service would cost a minimum price, impacting the total expenditure in healthcare positively since chronic diseases contribute to 90% of the overall spending in healthcare. Generative AI could help patients with chronic illnesses by

Syncing with wearable medical devices and supportive consumer technologies like Alexa, Bixby, or Chrome to supply around-the-clock monitoring while giving patients individualized, daily helpful health updates.
Comparing hundred of wearable-device readings against the expected ranges preset by each patient’s doctor assembles patient and physician alerts when something’s wrong.
Reminding patients at home when they are due for preventive screenings, Rx refills, or daily exercise (along with other lifestyle improvements). (3)

Preventing medical errors

With the success of OpenAI’s image-based AI platform, Dall-E, and promising developments in video-based AI from companies such as Meta, machine-learning capabilities will expand beyond text prediction. In the healthcare industry, video-enabled AI could reduce medical errors, a leading cause of death in the United States. Even though scientists have established the necessary measures to prevent unnecessary fatalities, patient safety is still jeopardized by lapses, particularly in hospitals, resulting in tens of thousands of annual deaths, with some estimates reaching as high as 200,000. Healthcare experts have even recommended that hospitalized individuals bring a family member with them to mitigate the deadly errors caused by human mistakes.

However, this may not be necessary in the future. A recent article published in the New England Journal of Medicine found that almost one out of four individuals admitted to a hospital would experience harm during their stay. Future generations of ChatGPT equipped with video capabilities will be able to monitor healthcare providers, compare their actions to evidence-based guidelines, and alert them when they are about to make a mistake. This progress could prevent almost all errors, as well as most hospital-acquired infections, pneumonia, and pressure ulcers. (4)

Healthcare experts have even recommended that hospitalized individuals bring a family member with them to mitigate the deadly errors caused by human mistakes. However, this may not be necessary in the future. A recent article published in the New England Journal of Medicine found that almost one out of four individuals admitted to a hospital would experience harm during their stay. Future generations of ChatGPT equipped with video capabilities will be able to monitor healthcare providers, compare their actions to evidence-based guidelines, and alert them when they are about to make a mistake. This progress could prevent almost all errors, as well as most hospital-acquired infections, pneumonia, and pressure ulcers. (4)

Helping all doctors performance

Medicine is both an art and a science, with medical students and residents learning both aspects through various methods, including textbooks, journal articles, classroom lectures, and observing experienced clinicians. Future generations of AI will follow a similar approach. When ChatGPT is integrated with bedside patient monitors, laboratory data, and physician-patient interactions, it will be able to predict the best clinical course of action. With each comparison it makes between its decisions and the clinical notes and orders of attending physicians in the electronic health record, ChatGPT will continue to learn and improve. It takes a first-year medical student ten years of education and training to become proficient.

However, future iterations of ChatGPT will be able to accomplish this in mere months or less by learning from the actions of the best clinicians across hundreds of hospitals. As generative AI becomes more skilled at predicting expert behavior, it can make that expertise available to healthcare providers nationwide. (5)

With each comparison it makes between its decisions and the clinical notes and orders of attending physicians in the electronic health record, ChatGPT will continue to learn and improve. It takes a first-year medical student ten years of education and training to become proficient. However, future iterations of ChatGPT will be able to accomplish this in mere months or less by learning from the actions of the best clinicians across hundreds of hospitals. As generative AI becomes more skilled at predicting expert behavior, it can make that expertise available to healthcare providers nationwide. (5)

Limitations: What ChatGPT can’t do

ChatGPT, no matter how advanced and skilled it becomes, will have its limitations. The accuracy of the data that humans input will always be critical to its performance. The doctors’ biases on which the application is trained will also influence it. Nonetheless, ChatGPT will keep advancing and solving increasingly complex medical challenges. It may take ten years (with 32 times the computing power) or twenty years (with 1,000 times the power), but future generations of generative AI will surpass today’s physicians in cognitive and problem-solving capabilities. To prepare the next generation of doctors, educators must challenge healthcare’s implicit norms and incorporate this technology into medical school and residency training. Trainees will benefit by learning how to leverage the clinical capabilities of generative AI rather than considering it a threat. (6)

To prepare the next generation of doctors, educators must challenge healthcare’s implicit norms and incorporate this technology into medical school and residency training. Trainees will benefit by learning how to leverage the clinical capabilities of generative AI rather than considering it a threat. (6)