Sokat: AI News of the Week
A number of recent studies have shown that artificial intelligence does in fact have a place/belong in healthcare. Not only can new technologies streamline processes (which has been done within breast cancer imaging), they can minimize workload on doctors and free up time for more complex work.
SoKat has compiled six articles that each detail a new AI development, and how it has positively impacted different concentrations in healthcare. The future is bright for partnerships between artificial intelligence companies and healthcare providers.
“Putting artificial intelligence at the heart of health care — with help from MIT,” MIT Professional Education
November 14, 2021
A recent Mayo Clinic study found that AI-enhanced electrocardiograms (ECGs) have the potential to save lives by speeding diagnosis and treatment in patients with heart failure who are seen in the emergency room.
A research team at MIT successfully completed and published their ECG project in August 2020, with promising results. In analyzing the ECGs of about 1,600 patients, the AI-enhanced method was both faster and more effective — outperforming the standard blood tests with a performance measure (AUC) of 0.89 versus 0.80. This improvement could enhance health outcomes by improving diagnostic accuracy and increasing the speed with which patients receive appropriate care.
Looking forward, more research will be conducted to determine the potential for AI to improve care and outcomes for cardiac patients.
“Artificial Intelligence Advances Breast Cancer Detection,” Erin McNemar, MPA
November 14, 2021
A new artificial intelligence technology has the potential to significantly reduce radiologists’ workload when distinguishing breasts with and without lesions in MRI results.
Study lead author Erik Verburg states that “the approach can first be used to assist radiologists to reduce overall reading time. Consequently, more time could become available to focus on the really complex breast MRI examinations.”
This AI-based triaging system will continue to be validated in other datasets and hopefully implemented in hospitals around the U.S. over the next few years.
AI in the Workplace
April 13, 2022
As the automation of labor through machinery and artificial intelligence continues to spark debate regarding the future of jobs, workers are beginning to consider how AI could, in fact, complement their jobs.
Contrary to concerns that AI will replace human labor, data suggests that 70% of US workers “would like to utilize AI for their jobs to some degree”. The top task that workers think can be completed by AI include data processing (defined simply as the process of converting raw data to a more usable form). If not completely replacing the task in its entirety, more than 50% of workers think that AI can at least perform some functions of digital tasks. It is important to note that many workers are starting to see AI as a tool to enhance their efficiency, rather than a threat to their jobs.
Growing confidence in the role of AI in the workplace is encouraging IT leaders to invest more in developing AI solutions and implementing technology within the workplace. However, given the fact that more and more workers are starting to consider AI as an alternate to performing low-value tasks, it is worthwhile to maintain a sense of optimism towards the future of jobs.
March 24, 2022
A non-profit AI learning service, the AIClub, presents statistics that suggest AI participation for under-represented groups such as women is much higher than other forms of computer science.
The AIClub’s study suggests that the format of AI of directly seeing real-world problems being solved in their own community specifically appeals to girls. It also points to the fact that AI can be built almost anywhere with some sort of internet access, which enables girls around the world to get engaged even where there is limited availability of education. AI also offers a large selection of genres ranging from music to healthcare giving it an advantage over Robotics.
Their data suggests that AI is a viable vehicle to drive more girls to STEM-related fields like computer science and engineering.
“Using Artificial Intelligence to Detect Anti-cancer Immunity,” Erin McNemar, MPA
November 14, 2021
UT Southwestern Medical Center and the University of Texas MD Anderson Cancer Center researchers developed a method to determine anti-cancer immunity using artificial intelligence.
Mutations in the cancer cells’ genomes cause the revelation of different neoantigens on their surfaces. Some of these neoantigens can be identified by immune T cells in our bodies, therefore allowing our immune system to take down the unknown bodies. Due to the sheer volume of antigens and outcomes to detect, researchers are delegating the task to machine learning.
The team built an algorithm named pMTnet from known binding or nonbinding combinations or three different components: neoantigens, major histocompatibility complexes (MHCs), and T cell receptors (TCRs). The tool has been used to gather information on neoantigens cataloged in The Cancer Genome Atlas. Additionally, it could predict patients’ immune response to different therapies.
Lower Hospital Infections
November 3, 2021
As the COVID-19 pandemic continues to spread, hospitals are working hard to protect patients from being infected while checked into their facilities. According to Eric Eskioglu, MD, the Chief Medical and Scientific Officer for Novant Health, about 1.7 million hospital-acquired infections (HAIs) occur each year, and that 100,000 result in death.
Eskioglu explained that hygiene monitoring, while definitely crucial, was not enough at this time. Adding artificial intelligence into the mix, Novant partnered with SwipeSense. SwipeSense’s role is to provide automated electronic tracking down to the staff member to help with hand hygiene compliance. This eliminates the need for practitioners to observe each other and therefore observer bias, and allows them to focus on patient care.
November 3, 2021
University of Central Florida researchers are exploring whether a combination of artificial intelligence and telehealth services can improve provider training and patient outcomes.
Throughout the COVID-19 pandemic, telehealth has been a crucial tool linking patients to providers; however, it does not come without limitations, which could potentially lead to misdiagnoses and unaddressed issues. Moreover, medical tests and measurements are difficult to obtain through a screen.
To combat this, UCF researchers will study both doctor and patient behaviors during telehealth visits to determine areas where AI can be incorporated. They will observe eye movement, heart rate and communication throughout the sessions. They will consult physicians and engineers to optimize solutions. The study is set to begin in January 2022.
April 12, 2022
The Massachusetts Institute of Technology is offering a six week AI literacy course called the “FutureMakers” program. This interactive experience is targeted towards middle school and high school students who will be immersed in the world of AI and Machine Learning.
During this six week program, students will participate in a “Create-a-Thon,” where admitants take a deep dive into the fundamentals of AI and coding for their first four weeks, and then transition to a two week period of designing an app for social good. These students will then present their work to an expert panel of MIT judges, who will pick a winner of the competition.
Their purpose of the program is not simply to teach adolescents to code, rather to create a future society of digital citizens in the rapidly increasing world of AI. The competition itself is a great way for students to use their newly acquired knowledge of machine learning to create applications for a more inclusive and progressive society.