How OpenAI Advances are Shaping Healthcare?
Artificial intelligence is rapidly changing the healthcare landscape, from aiding in the development of new drugs to improving patient care. AI algorithms effortlessly automate repetitive tasks, allowing medical professionals and pharmacists to focus on more human-centric issues.
AI systems have been used in healthcare for quite some time, including both public platforms and proprietary developments by large medical corporations.
With OpenAI's support, Moderna launched an internal chatbot called mChat. By the end of 2023, 80% of employees had adopted it, paving the way for a full transition to ChatGPT Enterprise.
The need for rapid processing of vast amounts of information led Healthify to become one of the first healthcare companies to use AI.
Specifically, the platform launched the world's first virtual dietitian, Ria, which by 2020 was handling the majority of their client inquiries. In 2021, Healthify launched the AI application Snap, which calculates calories by recognizing food in photos. For local Indian products, the accuracy of this calculation reached 80%.
There were challenges, however. For instance, Snap struggled with tasks involving photos that contained multiple food items. Ria also had difficulties with long chains of user questions. Additionally, both products were tailored exclusively for the Indian market. It took Healthify two years to localize the platform for the Southeast Asian market.
These scaling issues highlighted the need for more advanced AI solutions.
Through the initiative led by neurosurgeon Rohaid Ali and dermatologist Fatima Mirza, Lifespan began using ChatGPT to train residents for exams, allowing them to quickly develop the necessary skills.
This led to the question: how could OpenAI solutions enhance communication with patients?
They focused on the overly complex language of patient consent forms for surgeries—often requiring medical knowledge to understand.
GPT-4 was tasked with condensing a three-page consent form into simple language without losing any essential meaning. The AI-generated draft was understandable to a sixth-grade level student and was so accurate that doctors made only one addition. The revised form began implementation at Lifespan in fall 2023.
Pediatricians need to write reports on visits and consultations, create care plans for young patients, and more. This reduces the quality of interactions with sick children and leads to rapid burnout among specialists. Additionally, parents face delays in receiving records, and the records themselves are often difficult to understand.
Summer Health hypothesized that generative artificial intelligence could significantly improve medical visit records and focused on GPT-4's capabilities in understanding complex medical language. Moreover, the platform's adaptability to user requirements was highly appealing.
In collaboration with OpenAI, the company enhanced the GPT-4 model, and now, visit notes based on detailed written observations by doctors are created by a bot. These records require only a quick review before being handed over to parents.
Doctors reported a fivefold reduction in the time spent on administrative tasks. This optimization allowed them to devote more attention to their young patients and improved the accuracy and consistency of the records. Parents now receive feedback from doctors about visits four times faster.
Parents also noticed the simplicity of the records, the ease of understanding information about their child's health, and a better grasp of their own care actions.
The latest breakthrough came in 2023 with the launch of a new suite of business-oriented platforms based on ChatGPT-4, including GPT-4 Turbo, ChatGPT Enterprise, and others. These OpenAI products are easily adaptable to the specific needs of individual companies and have already gained wide acceptance in the corporate sector, including healthcare.
Specifically, ChatGPT Enterprise offers unlimited access to GPT-4 and advanced data analysis (formerly known as Code Interpreter). It processes data twice as fast as GPT-4, with four times the input capacity.
Listed below are several success stories from healthcare companies that have adopted these OpenAI products.
Moderna
The company, widely known for its COVID-19 vaccine, began collaborating with OpenAI in early 2023, becoming one of the first in the pharmaceutical industry to integrate ChatGPT Enterprise into all business processes.
Moderna plans to bring up to 15 new products to market over the next five years, ranging from vaccines to personalized cancer treatments. To achieve these ambitious goals, Moderna's management aimed for 100% adoption of generative intelligence by all employees within six months. They identified the top 100 AI users, who then trained the rest of the staff across all company offices.
With OpenAI's support, Moderna launched an internal chatbot called mChat. By the end of 2023, 80% of employees had adopted it, paving the way for a full transition to ChatGPT Enterprise.
90% of companies want to do GenAI, but only 10% of them are successful, and the reason they fail is because they haven’t built the mechanisms of actually transforming the workforce to adopt new technology and new capabilities,said Brad Miller, Chief Information Officer of Moderna.
The company noted that integrating ChatGPT Enterprise allowed the team to focus entirely on issues that truly matter to patients. Additionally, if Moderna were to launch current projects using traditional biopharmaceutical methods, they would need to expand their workforce to nearly 100,000 employees. AI capabilities enable them to achieve results with a much smaller team.
Healthify
One of the largest healthcare platforms, India's Healthify, has assembled a team of hundreds of dietitians and fitness trainers, boasting over 40 million active users. Continuous feedback from patients helps successfully manage various chronic conditions, control weight, develop balanced diets, and create personalized workout plans.
The need for rapid processing of vast amounts of information led Healthify to become one of the first healthcare companies to use AI.
Specifically, the platform launched the world's first virtual dietitian, Ria, which by 2020 was handling the majority of their client inquiries. In 2021, Healthify launched the AI application Snap, which calculates calories by recognizing food in photos. For local Indian products, the accuracy of this calculation reached 80%.
There were challenges, however. For instance, Snap struggled with tasks involving photos that contained multiple food items. Ria also had difficulties with long chains of user questions. Additionally, both products were tailored exclusively for the Indian market. It took Healthify two years to localize the platform for the Southeast Asian market.
These scaling issues highlighted the need for more advanced AI solutions.
We integrated with everybody who’s out there. OpenAI was the best,explains Tushar Vashisht, CEO of Healthify.
During the initial integration of OpenAI products, the GPT-4V(ision) model immediately outperformed Snap by recognizing images of multiple food items from around the world. Upgrading Ria to the GPT-4 Turbo model removed language barriers, provided patients with access to Healthify’s extensive literature, and enabled the creation of personalized user histories on the platform.
In 2024, Healthify plans to implement autonomous modes of operation for the platform. Virtual assistants will proactively form recommendations on nutrition, sleep, and exercise, order food, and schedule gym sessions. With OpenAI’s tools, the company aims to scale its business to 20 more countries.
Lifespan
Sometimes AI helps healthcare professionals solve unexpected problems. For instance, Lifespan, a healthcare system in Rhode Island, USA, used generative AI to develop patient consent forms for surgical procedures, making the documents accessible even to those without a medical background.
Through the initiative led by neurosurgeon Rohaid Ali and dermatologist Fatima Mirza, Lifespan began using ChatGPT to train residents for exams, allowing them to quickly develop the necessary skills.
This led to the question: how could OpenAI solutions enhance communication with patients?
They focused on the overly complex language of patient consent forms for surgeries—often requiring medical knowledge to understand.
GPT-4 was tasked with condensing a three-page consent form into simple language without losing any essential meaning. The AI-generated draft was understandable to a sixth-grade level student and was so accurate that doctors made only one addition. The revised form began implementation at Lifespan in fall 2023.
Surgical consent conversations can be very overwhelming and emotionally charged for patients. Since we’ve implemented the simplified consent forms, countless patients have expressed how meaningful it is for them to have a one-page form they can understand,stated Dr. Fatima Mirza.
Following this success, Lifespan decided to simplify all of its medical documentation using GPT-4.
Summer Health
At Summer Health, a company specializing in round-the-clock pediatric care, it was noted that 50% of specialists' time was consumed by administrative tasks.
Pediatricians need to write reports on visits and consultations, create care plans for young patients, and more. This reduces the quality of interactions with sick children and leads to rapid burnout among specialists. Additionally, parents face delays in receiving records, and the records themselves are often difficult to understand.
Summer Health hypothesized that generative artificial intelligence could significantly improve medical visit records and focused on GPT-4's capabilities in understanding complex medical language. Moreover, the platform's adaptability to user requirements was highly appealing.
In collaboration with OpenAI, the company enhanced the GPT-4 model, and now, visit notes based on detailed written observations by doctors are created by a bot. These records require only a quick review before being handed over to parents.
Doctors reported a fivefold reduction in the time spent on administrative tasks. This optimization allowed them to devote more attention to their young patients and improved the accuracy and consistency of the records. Parents now receive feedback from doctors about visits four times faster.
Parents also noticed the simplicity of the records, the ease of understanding information about their child's health, and a better grasp of their own care actions.
We thought this was something we would get to in 5 years. Seeing that vision pulled forward has been amazing, and we’re excited to continue this journey with OpenAI,expressed Matthew Woo, co-founder of Summer Health.
