In the full academic year since the launch of ChatGPT, Yale administrators and instructors have altered their guidelines and teaching styles to accommodate for the new technology. Though many remain concerned about plagiarism with respect to AI, some professors have embraced using AI in the classroom. 

Ben Glaser, a professor of English, initially became interested in applying AI to the humanities after he took a summer course for faculty members on natural language processing in 2022. He recalled thinking that AI was going to “transform the writing landscape.”

“I thought, ‘I’ll teach a writing course, because then we’ll be thinking about how to write well, which I do all the time,’” Glaser said. “‘We’ll be thinking about these tools that might hinder or maybe help.’” 

In the fall of 2023, Glaser taught an introductory English seminar titled “Writing Essays with AI,” in which he and his students discussed how artificial intelligence could be applied to writing. They also explored the relationship between authorship, creativity and AI. 

His ultimate goal, he said, was to help his students become better writers.

“It’s easy to say, ‘Oh, the AI can never be creative,’” Glaser said. “And I’m like, ‘Can’t it? Let’s interrogate these distinctions.’” 

Students in the course read AI-generated stories, analyzed the differences between poetry written by humans and written by AI and even practiced using AI to plan essays. For their final projects, they researched how various industries are using AI.

In November 2022, shortly after Glaser designed and proposed the course – roughly a year before he first taught it — the company OpenAI launched ChatGPT, its popular generative AI program. According to Glaser, the release of the chatbot program prompted students and teachers to pay more attention to how AI could transform learning. 

Poorvu Center responds

For the Poorvu Center, the University’s center for teaching and learning support, the release of ChatGPT was a catalyst to develop academic AI guidelines, said Alfred Guy. The deputy director of the Poorvu Center and director of the Poorvu Center’s writing and tutoring programs, Guy has helped conduct workshops on AI and facilitate professor education programs about the technology.

Guy learned of the imminent release of ChatGPT through Facebook, and he immediately felt that the Poorvu Center ought to respond. After the chatbot went public, the Poorvu Center released its first guidelines on AI usage in Yale classrooms in January 2023.

“The very first thing we said is, ‘These tools are powerful, and people are going to use them,’” Guy said. “Everything that comes after this should be thought about in terms of how people are going to use these tools.” 

As AI software like ChatGPT became more mainstream, Guy noted that Yale instructors did not generally react with panic or fear. Still, they wondered how AI would affect them, and whether they needed to take concrete steps of action in their own classrooms. Guy said that the Poorvu Center’s teaching guidelines, which are now overseen by a five-person committee, are aimed at answering these questions.

The current guidelines include suggestions for how instructors can address AI in their syllabi, remind students to correctly cite AI and offer precautions when using AI technologies. Beyond providing links to dozens of articles and webinars on AI, they also encourage instructors to try AI tools for themselves and share their feedback with the center.

Guy said the Poorvu Center’s approach to AI over the past year has changed. Rather than striking a cautionary tone, the Center now encourages hands-on exploration of AI technology. 

“We have shifted very slightly toward AI in our tone and in our attitude,” Guy said. “Even in our specific advice to faculty, we are saying, ‘You really should engage.’” 

Glaser also pointed out that Yale’s open approach to teaching and learning with AI does not necessarily mirror that of other colleges and universities. He said that because of broader institutional support and awareness of the technology, he thinks that AI is not as contentious at Yale as it may be at other universities.

“If we get out of the Yale bubble, the writing landscape looks really different, and AI tools are gonna behave differently,” Glaser said.

AI in the classroom

As one of their class projects, Glaser required the class to revise Poorvu’s suggested AI-use guidelines.

Jared Wyetzner ’27 was a member of Glaser’s class and previously co-founded Myndful-AI, a machine learning chatbot that provides high school students with mental health resources.

Wyetzner said the class generally agreed that AI tools had a role in Yale education.

“What we moved toward is that AI can belong in your classroom,” he said. “There’s just a certain way that should be used to facilitate work.” 

To Wyetzner, AI is best seen as a tool that he compared to the “calculator” of writing. 

“You learn how to do math, addition, multiplication all that, by yourself, and then it gets to the point where you use a calculator and that becomes the standard,” he said. “How can we still learn from our writing and also have AI tools in the process?”

One Yale course quickly added AI to its classroom: CPSC 100, Yale’s introductory programming class, which is co-taught with Harvard’s CS50 course.

According to Ozan Erat, a Yale computer science professor who helps run CPSC 100, the course used two different AI technologies. The first, a chatbot called Duck Debugger, allowed students to ask questions about the course and helped them debug code. The second, CS50 Duck Bot, was integrated with the online forum Ed Discussion and replied to students’ questions.

Because of Duck Debugger, Erat says attendance at CS50 office hours dropped by roughly 30 percent. He said that this was a positive development, as students with easier questions could ask Duck Debugger at home while those with more in-depth questions could receive more attention during office hours. 

Though Erat was initially worried about academic dishonesty, he said that by the end of the semester there was not an “excessive use of cheating.” He noted that the percentage of students referred to the University’s Executive Committee did not drastically change, and the CS50 instructors plan to continue to use Duck Debugger and Duck Bot as part of the course. 

Plagiarism worries

The News spoke with Mick Hunter, the Chair of the Yale College Executive Committee, about students using AI to commit plagiarism and academic dishonesty. He said the Committee began to see cases related to AI shortly after the launch of ChatGPT. In response, the University added a section on AI use to its academic integrity guidelines. 

While the use of AI was “clumsy” in these first cases — students generating false citations for a paper, for example — now the cases are “less blatant,” Hunter said.

“We’re still seeing students who are breaking the rules and using AI in ways that are not allowed, but it seems students are either learning to cover their tracks or use AI more responsibly,” Hunter added. 

However, Hunter estimated that the Executive Committee has received around 10 cases related to AI since November 2022, placing them “in the minority” of academic dishonesty cases. 

By virtue of his job in the Poorvu Center, Guy considers himself “just short of an expert” on pre-AI plagiarism.

Though Guy acknowledged that AI could contribute to plagiarism in writing, he expressed optimism about different ways instructors could help limit it. 

Guy referred to multiple research studies which demonstrate that rates of plagiarism decrease when instructors require students to write low-stakes responses to class material, set interim deadlines for larger assignments through the semester and foster conversations where students describe their ideas. 

Changing assignments to include requirements beyond the capacity of a language-generating AI model, he proposed, could also reduce instances of plagiarism. 

Glaser and his students also discovered that using chatbots for writing comes with its own challenges. Due to its capacity for error and “idiosyncratic” responses, students need a type of “literacy” to interpret AI outputs and screen for errors. 

“We quickly realized in the classes that to get anything good out of them, you actually have quite a bit of a dialogue,” he said. 

By the time the students finish writing ChatGPT a good prompt, assessing the quality of its response and incorporating the response into their writing — not to mention citing it — the process might be more effort than it is worth, Glaser pointed out. 

“By the end of that process, I’m not worried about plagiarism,” Glaser said. “I’m just wondering, ‘Was that actually efficient or useful? Did it make you a better writer?’ I think the answer is sometimes yes.”

Yale established the Poorvu Center in 2014.

The post AI and the classroom appeared first on Yale Daily News.

With the geographical closeness of a small campus and proximity to people who are of similar ages with common interests, dating inside the Yale bubble makes sense. Some of us are looking to experiment with sexuality, some of us are searching for romance and companionship, and some of us are crippled by the fear that once we graduate, move to a big city and begin our careers, our odds of finding love all but vanish and we are therefore determined to leverage our 17 percent chance of meeting our future spouse in college to its fullest potential. 

Some Yalies indeed meet their future spouses on campus. Take, for example, Tara Falcone ’11, who shared the story of meeting her husband over email. 

“I met John at a Kappa-ADPhi get-together in February of our freshman year,” Falcone wrote to the News. “We locked eyes across the table during a round of flip cup (non-alcoholic of course), then spent the rest of the night chatting. He sent me a private Facebook message the next day (the early days of DMs – scandalous).”

After a lunch shared in Commons, the couple started dating and they’ve been together ever since. Assignment aced. 

But not everyone can be Tara and John. 

Matchmaking services like Marriage Pact seek to alleviate the fear of being forever single — albeit in a lighthearted manner — by using an algorithm to pair undergraduates with potential future spouses, but the worry of being a love-lacking graduate is only strengthened by a quick peek at the classifieds of any Ivy League alumni-magazine. There, you’ll find dozens of personal ads for people who are professionally successful, but still searching for love. 

In those same classifieds, you’ll also find a handful of dating and matchmaking services, all targeted at single graduates of prestigious schools. There’s Elegant Introductions, a matchmaking service that describes itself as “elite matchmaking for select singles,” The Right Time Consultants, a curated matchmaking service for “top shelf clients,” or the Ivy Plus Society, an alumni networking event which the New York Times called a “meet market for the pedigreed.”

I have to wonder, are these services really about companionship? Or are they selling a chance at power, wealth and prestige? 

Dating in the Ivy League is implicitly swayed by the fact that we’re dating potentially famous or powerful people: lawyers, doctors, actors, authors, politicians. Beyond academic or professional success, the carefully curated pool of Ivy League admits are also far more likely to be in the top wealthiest 1 percent. When options are purposely limited to this small, elitist pool, perhaps dating becomes less about falling in love and more about securing a future. 

So, would Yalies use these Ivy-only dating services?

Friends and peers who weighed in on the idea had mixed feelings. 

“I wouldn’t date any of those pretentious hoes,” said one anonymous senior, who met his long-term girlfriend at Yale. 

Others were uncomfortable with the elitism of these services and reasoned that going to the same school doesn’t guarantee you’ll have anything in common. 

But some suggested that, after graduation, a similar schooling background might provide common interests and a source for shared ambition with a potential partner. 

“I don’t mind dating at Yale,” said another senior who requested to remain anonymous. “I met my girlfriend at Yale. And my other girlfriend. And my other other girlfriend.”

Dating (and not dating) at Yale is, in short, whatever you make of it. And if you don’t find a special someone in your bright college years? Well, you can always place your own ad in the classifieds. 

The post Bulldogs and Puppy Love appeared first on Yale Daily News.

In 2021, Sofia Noori, an instructor at the School of Medicine’s Department of Psychiatry and a graduate of Yale’s psychiatry residency program, founded the platform Nema Health. The program uses intensive therapy, delivered virtually, to treat post-traumatic stress disorder, or PTSD.

Noori said she hopes that the virtual model, coupled with intensive, standard-of-care therapy options that few mental health providers offer, could make headway in making trauma care more accessible.

“By doing it remotely, we’re trying to take away one more barrier for a patient to come and do trauma therapy,” Noori said. 

Post-traumatic stress disorder, or PTSD, is a debilitating psychiatric illness that millions of Americans suffer from each year, including veterans and survivors of sexual violence. The condition often goes undiagnosed and untreated, even though many patients experience psychological distress, nightmares and intense flashbacks, according to the U.S. Department of Veterans Affairs.

Individuals experiencing PTSD might try to avoid reminders of their trauma which also makes it difficult for them to seek out help, said Noori.

As a daughter of refugees who survived the Vietnam War, and as a survivor of sexual assault in college, Noori knows firsthand how trauma can impact health. Based on her own experiences, she acknowledged how difficult seeking treatment can be for survivors. 

During her psychiatry residency at Yale, Noori looked for ways to combine healthcare innovation with evidence-based trauma treatment. However, while there were effective treatments for PTSD, it was nearly impossible to find providers who offered them, she said.

Then, as the pandemic forced mental health care to shift online, Noori realized a virtual model could be an answer to delivering accessible, effective care to community members. As the first chief resident of digital psychiatry at Yale, she helped found the Center for Digital Psychiatry at the Connecticut Mental Health Center, which focuses on integrating digital health into treatment for patients with serious mental illness. 

Shortly after she graduated from residency, she started Nema Health. 

Noori said that Nema Health’s clients experience relief from PTSD symptoms in as little as two weeks. By program completion, she pointed out that nearly 80 percent of patients no longer meet the criteria for a PTSD diagnosis.

But these impressive results, Noori pointed out, are not unique to Nema’s program: PTSD is a highly treatable condition with long-term results.

“Treatments for PTSD are actually super effective,” Noori said. “They generally take 10 to 12 sessions to complete. And they lead to generally permanent recovery.”

Nema Health uses what Noori refers to as one of the “gold standard” PTSD treatments: a method called Cognitive Processing Therapy, or CPT, in which therapists guide patients to challenge and reframe harmful thought processes linked to their trauma. But according to Noori, many providers don’t offer CPT, creating a scarcity of therapy options for patients. 

As an evidenced-based treatment, Noori said, CPT has been shown to be highly effective when delivered virtually — just as Nema Health intends to do. 

The program uses an “intensive model:” instead of meeting once a week, the program advertises that Nema clients are matched to a dedicated therapist with whom they meet for three to five sessions per week over three to four weeks. 

After the initial course of treatment, Nema continues to provide regular check-ins, medication management and peer mentorship, said Noori.

“If you actually look at the research, Nema’s outcomes are consistent with what people see,” she added.

Right now, Nema Health operates in three states: New York, Connecticut and New Jersey, though Noori’s goal is to expand its teletherapy coverage to all 50 states.

The cost of Nema’s virtual therapy can be prohibitive, admitted Mariam Malik, co-founder and COO of Nema. Without insurance coverage, the program costs $400 for an initial evaluation and $250 for each individual session according to Nema’s website, though Malik clarified in an email to the News that these services are offered on a sliding fee scale for those who cannot afford to pay full price.

Nema’s treatment is currently covered by a few health plans in the Trisdate area, including United Healthcare, Optum, ConnectiCare and Oxford. According to Malik, the company is working to expand coverage to other insurance providers, including Medicaid, though negotiating pricing with insurance providers has proved time-consuming.

“A lot of people who need our care wouldn’t be able to afford it out of pocket,” said Malik. “Our care needs to be accessible, especially with people who struggle from complex trauma. Being covered by their health benefits is part of Nema’s mission.” 

Noori acknowledged that virtual therapy can’t fix every gap in mental health access, and it might not be the appropriate approach for every patient facing mental health challenges. 

For Walter “Stan” Mathis, a former colleague of Noori’s and a clinical psychiatrist who directs the Connecticut Mental Health Center’s Assertive Community Treatment team, telehealth options are incompatible with the needs of clients he treats each day.

Mathis works primarily with clients who have psychotic disorders — visiting them in their homes or out in their communities. For people who are experiencing psychosis and have trouble differentiating what is reality, he said, virtual care is seldom a good option. 

“There’s a real fantasy of treating mental illness homogeneously,” Mathis said. “But person to person is different, and there are big picture diagnostic differences.” 

Beyond differing treatment needs, there are also technical and structural barriers that make accessing virtual care difficult. At the Center for Digital Psychiatry in 2019, Noori and Mathis helped develop an initiative to teach clients to use their smartphones for telehealth. 

But instead, the program ended up teaching clients basic technology literacy skills.

“Our training sessions were like, how to charge your phone, and that’s what they wanted to talk about,” Mathis said. 

Still, Mathis said he believes that programs like Nema Health are a great way to increase the availability of mental health care. 

“I think we’ll end up using telehealth as a way to make the spectrum of care have more options,” he said. 

The Connecticut Mental Health Center was founded in 1966. 

Update, Feb. 12: This article has been updated to reflect the correct definition for avoidance in PTSD, offer more information about Nema Health’s sliding payment system and clarify that Nema Health offers virtual therapy — not psychiatry.

The post Yale faculty member’s online therapy program seeks to break down barriers to PTSD care  appeared first on Yale Daily News.

In this episode of Footnotes, we explore the controversial career and personal life of Yale graduate Lee de Forest, who is best known for inventing the part that made early radio and television possible. Was he really the “father of radio” or was he just a great storyteller?

Hosted by Hannah Mark ’25. Produced by Hannah Mark ’25 and Grace Ellis ’25.

Music:
Delicates by Blue Dot Sessions
Tall Harvey by Blue Dot Sessions
Etude 9 Stefan by Blue Dot Sessions
Coulis Coulis by Blue Dot Sessions

The post Footnotes E2: Mythmaking and the Father of Radio appeared first on Yale Daily News.

Led by researchers at the Yale Center for Natural Carbon Capture and the Yale School of the Environment, the project also hopes to develop tools to measure the effectiveness of natural carbon capture methods, which researchers believe is crucial for addressing the issue of climate change. 

The project’s funding comes from a $5 million grant as part of the DOE’S Earthshot Initiative, an agency-led effort to increase decarbonization and help the United States reach its goal of a net-zero carbon economy by 2050. 

“There’s no time to sit around and twiddle our thumbs,” said David Bercovici, the Frederick William Beinecke Professor of Earth & Planetary Sciences and co-director of the YNCC. “This project is really connecting the models for soils and enhanced weathering to rivers and estuaries and all the way to oceans.”

In 2010, the United Nations agreed that global warming caused by greenhouse gasses needs to be contained to below a 2 degrees Celsius increase from temperature levels before the industrial revolution. If the global temperature rises to 2 degrees Celsius above pre-industrial levels, scientists warn that heatwaves, drought, extreme weather and rising sea levels could cause mass-extinction events. 

However, according to Matthew Eisaman, a professor of Earth and Planetary Studies and a researcher at the Yale Center for Natural Carbon Capture, efforts to mitigate climate change by reducing carbon dioxide emissions are not enough.

“We need billions of tons of direct removal of carbon dioxide from the air in the next coming decades,” he said. 

Eisaman, who is on the Yale research team for the study, said the project’s goal is to create models to examine how carbon flows through natural systems: from the soil, to rivers and eventually to the ocean. The modeling could inform how natural methods of carbon removal, such as enhanced mineral weathering and ocean alkalinity enhancement, can be used on a larger scale.

Enhanced mineral weathering relies on using natural rock weathering to draw carbon out of the atmosphere. When rain falls, raindrops mix with carbon dioxide and form an acid that weathers silicate rocks — which make up the majority of the Earth’s crust — and create bicarbonate ions. These bicarbonate ions then wash into soil and rivers. 

The chemically-basic bicarbonate runoff from this process can also help de-acidify the ocean, which is a large sink for carbon dioxide on Earth. Acidic oceans absorb less CO2 and are harmful for animal and plant life, the researchers said. Bicarbonate entering the oceans can decrease its acidity and increase the amount of CO2 that they can absorb from the atmosphere. 

Projects like that of the Yale team can help scientists understand the cycles through which carbon dioxide is drawn out of the atmosphere: how it moves through the soil into waterways and what chemical reactions occur during the process. 

“There is the need to track the flow of this carbon to its long term storage in the ocean as there are processes that can lead to ‘leakage’ or the loss of the captured CO2 during transport,” wrote study-leader Peter Raymond, senior associate dean of research at YSE, in an email to the News. 

Models play a “key role” in verifying how much carbon dioxide has been removed from the atmosphere through methods such as enhanced rock weathering or ocean alkalinity enhancement, said Eisaman.

They can also help scientists find new ways to accelerate the process of natural carbon capture.

“[These models] are something that’s needed if these approaches are going to scale to the billions of tons per year that we need,” Eisaman said.

Created in 2021, and co-directed by YSE professors David Bercovici and Liza Comita, the Yale Center for Natural Carbon Capture supports research projects that tackle climate change — solutions that range from the ocean floor to local farms. 

Calling for collaboration between researchers, Bercovici emphasized the urgency of the climate crisis. The YNCC must create multi-pronged solutions, to seek ways to capture carbon on land, in farms and the ocean, he said.

According to Bercovici, carbon capture solutions are only one facet of climate change, and other measures must be pursued, including eliminating dependence on fossil fuels and becoming a renewable economy. 

“Carbon capture is not the panacea that’s gonna solve it. There’s no way, the problem is way too big. We have to be doing everything,” Bercovici said. 

The DOE’s Carbon Negative Shot was announced on Nov. 5, 2021.

The post Yale team awarded $5 million from Department of Energy to combat climate change appeared first on Yale Daily News.

Loved ones, medical staff, and the patients themselves doubt the reality of their symptoms. Doctors struggle to offer them explanations.  

At the Yale Multidisciplinary Long Covid Clinic, however, medical workers try to unravel the mystery of long COVID. The center is directed by Lisa Sanders, associate professor of general medicine, who is no stranger to medical mysteries like long COVID. She currently writes the “Diagnosis” and “Think Like a Doctor” columns for the New York Times, and she previously published two medical books,“Every Patient Tells a Story: Medical Mysteries and the Art of Diagnosis” and “Diagnosis: Solving the Most Baffling Medical Mysteries.” 

In her role as director, Sanders is essentially a detective in a white coat. When a patient is referred to the clinic, Sanders first investigates whether the culprit is long COVID. This process involves ruling out other factors that may cause similar symptoms. For example, if a patient reports feeling fatigued, she checks that the patient does not have sleep apnea. According to Sanders, by repeating this process, the researchers remove some uncertainty from a long COVID diagnosis. 

From there, Sanders told the News, she tackles the downstream effects of the disease. Because the symptoms of long COVID are known and treatable, Sanders said she can treat the effects, despite not understanding the root cause.

“If you can’t treat the actual disease, just treat the symptoms that are tormenting people and reassure them that at least some people get better,” Sanders said. “Depending on who you read, 80-plus percent of people don’t have symptoms after a year.”

The Yale Multidisciplinary Long Covid Clinic is on the third floor of Yale New Haven Hospital’s St. Raphael Campus. Patients enter a waiting room adorned with long COVID support group flyers and prints of beach paintings. A narrow hallway connects Sanders’ office, an examination room and a physical therapy room.

Sanders said that next month the Center will move to a larger space.

“Our patients say ‘oh my god, this is such a small room, this is such a small space.’ Like really?” Sanders said. “It’s gonna be so luxurious.”

Despite its small size, the Center provides each patient with help from social workers, physical therapists and Sanders. 

Early in the pandemic, this level of treatment was not available to long-haulers, a colloquial term for people with long COVID. Instead, patients in New Haven with long COVID symptoms had to seek out specialists in separate departments that best aligned with their symptoms, Sanders said. Sanders said that eventually, specialists realized that the best way to help people with long COVID was not with specialty care, rather, it was with a specialized long COVID Center. Sanders was immediately interested in leading the center.

“The thought that I could learn about this, and maybe help people with it seemed appealing,” Sanders told the News.

Part of this learning is trying to define long COVID. 

Characterized by persistent symptoms long after initial infection, long COVID remains a puzzling challenge for patients, healthcare professionals and researchers. However, common symptoms include fatigue, brain fog, cough and shortness of breath after their initial COVID-19 infection, Sanders told the News. For many patients, these symptoms can become physically and emotionally debilitating, Sanders said.

“You had COVID, and then something bad happened,” Sanders said. “But that’s all we can say now because we have no idea what the pathophysiology is.”

Parallels to polio

COVID-19 is not the first epidemic to leave a wake of people experiencing lingering symptoms, said Naomi Rogers, a history professor at Yale who studies 20th and 21st-century medicine and disability. 

In the middle of the 20th century, the United States weathered a series of polio outbreaks, a virus that primarily affects children and can, in rare cases, cause long-term paralysis. One 1955 polio epidemic infected 3,950 people in Massachusetts. Of those infected, 2,771 were paralyzed, some for a few weeks or months and some permanently.

“Polio was seen as a really scary disease because of its disabling effects,” Rogers said. “Everybody feared that if their child got it, they would never be able to work independently or live independently.” 

According to Rogers, some patients with polio also developed “post-polio” syndrome. These people had polio as children and fully recovered, but then, decades after their illness, became paralyzed.

Rogers pointed to similarities between post-polio syndrome and long COVID. Like patients with long COVID, people with post-polio syndrome continued to have symptoms of the disease long after the actual infection had passed, Rogers said. Some patients with post-polio were dismissed by their doctors because the doctors had never heard of the disease, according to Rogers. People with post-polio syndrome, like those with long COVID, Rogers said, had to advocate for themselves in medical settings and seek ways to alleviate their symptoms outside of traditional medicine. 

Rogers also highlighted similarities between long COVID and diseases such as chronic Lyme and Chronic Fatigue Syndrome. For people with these conditions, she said, it is difficult to make friends, family or doctors aware that there is an actual disease behind what she said are relatively vague symptoms. 

“You can’t see pain. You can’t see weakness. There are a lot of chronic responses that are not visible,” Rogers said.

In addition to vague symptoms, Rogers said that long COVID is complicated because nobody knows exactly what it is or how to think about it. 

Unlike other chronic diseases such as cancer or AIDS, she explained, long COVID is too recent for the public to have a clear picture of what it is like. 

“When we say long COVID, we don’t have in our minds what long COVID looks like,” Rogers said. 

Studying Long Covid

A group of interdisciplinary researchers at Yale are working to uncover what long COVID is and how it can be treated. 

During the pandemic, researchers at Yale, including Akiko Iwasaki, Sterling Professor of immunobiology at the School of Medicine, noticed that some patients who had COVID-19 developed chronic symptoms that lasted for a long time after their infection. Iwasaki and her lab also noted that some people who got the COVID-19 vaccine had an immune response.

As the pandemic progressed, Iwasaki established a network of researchers studying COVID-19 who also wanted to understand the unusual immune responses caused by COVID-19. This collaboration became the Center for Infection and Immunity, or CII, which opened in August 2023. 

The center, which Iwasaki directs, seeks to understand how chronic diseases develop and uncover links between infection and chronic disease. It also tries to connect immune responses to factors such as sex, genetics, age and environment, and develop therapies for treating chronic disease.

Iwasaki said that, because post-acute syndromes, such as long COVID, are difficult to understand and quantify, studying them requires an interdisciplinary approach.

Patient-centered approach

In studying chronic disease, the CII seeks to include patients as full participants in their research. CII partners with the Center for Outcomes Research and Evaluation, or CORE, to enhance their patient-centered research approach.

“We’re trying to create a different spirit around the research,” Harlan Krumholz, a professor at the School of Medicine and the director of CORE told the News. Krumholz is also on the CII leadership team. 

Krumholz said patients with post-acute syndromes are “desperate for answers.” 

But because chronic syndromes like long COVID are poorly understood, he said many patients feel “abandoned” by doctors, caregivers and the medical system. 

“There’s not one evidence-based therapy,” Krumholz said. “We even have trouble defining the condition, because there aren’t any diagnostics.”

To combat this feeling of neglect, CORE and CII developed a new model for doing patient research which Krumholz called “highly participant-centric.”

One way the researchers study long COVID is through the Yale LISTEN study, which stands for Listen to Immune, Symptom and Treatment Experiences Now. The LISTEN study follows a pool of participants with long COVID who share health data and specimens like blood or saliva with researchers. 

People participating in the LISTEN study can opt to participate in different research projects. For example, the CII is also conducting a trial on the effect of Paxlovid — an antiviral drug used to treat COVID-19 — on long COVID. 

The trial is being conducted virtually, to make it easy for people to participate in research. Patients involved in the study receive drugs or placebo pills by mail, then fill out surveys online and electronically transmit their data to researchers. 

The trial organizers also hold regular town hall meetings, where study participants can ask questions and suggest changes. When researchers learn new information through their research, they share it with the participants. 

New long COVID study gives hope to researchers, patients 

In September, researchers from the CII and Mount Sinai Hospital published a paper in Nature that identified biological markers associated with long COVID. By analyzing blood samples, the researchers found that hormonal dysfunction, exhaustion of certain immune cells and the reactivation of latent viruses are key factors in long COVID. 

“To us, that means that there is an immunological underpinning to this disease,” Iwasaki said. 

According to Iwasaki, the evidence that long COVID is a biological disease is an important discovery not only to understand the disease but also to validate patients’ struggles. People suffering from long COVID and other chronic diseases are often perceived as inventing or exaggerating their illness, Iwasaki said. 

The idea that patients are making up their symptoms, Iwasaki said, is one of the major misconceptions that the CII seeks to debunk. Iwasaki added that the identification of biological markers helps dispel the belief that long COVID is only a psychiatric illness. 

Sanders said she believes there should have never been a need to prove long COVID patients had an actual disease. 

“I never thought they needed any validation. I thought that people who saw it as psychiatric alone were nuts,” Sanders told the News. “Eleven million people have long COVID in the United States alone, and there’s so many more than that in the world. Over a million people don’t just come up with the same problem.” 

Long COVID is not the CII’s only focus. Researchers at the Center also study diseases such as chronic Lyme disease, which is a disease spread by ticks, and myalgic encephalomyelitis chronic fatigue syndrome, or ME/CFS.

While post-acute syndromes are complex and tricky to study, long COVID provides a unique research opportunity, Iwasaki said. People with the disease became ill around the same time and from the same virus, which allows researchers to simultaneously study a large number of people who have a similar chronic illness. 

Iwasaki said that it is difficult to find this kind of continuity with other post-acute viral syndromes because most chronic diseases occur in small numbers and at different times. 

Research developments often come up in conversation during visits to the Long COVID Clinic. Patients frequently tell Sanders about new research they have encountered, and Sanders will follow up on it. Sanders even has a Google alert that sends her every online publication that has to do with long COVID.

Like Iwasaki, Sanders views long COVID as a unique opportunity to develop our understanding of chronic illness.

“It would be terrible to waste this and not learn anything from it, like we have done with so many other infections,” Sanders said. “We have the technology and the curiosity. God only knows what other factors allowed this to be an active area of research, but I’m so glad it is.”

Adults aged between 35 and 49 exhibited the highest prevalence of experiencing long COVID.

The post ‘A different spirit’ of research: Yale clinic seeks to unravel the mystery of long Covid appeared first on Yale Daily News.

Last month, the Chan Zuckerberg Initiative, an organization co-founded and co-led by Facebook founder Mark Zuckerberg and Priscilla Chan, announced the launch of a new biomedical research hub in New York City that connects researchers from Columbia, Rockefeller University and Yale. Called the Chan Zuckerberg Biohub New York, or CZ Biohub NY, the center will be located in Manhattan, with satellite sites on the campuses of Columbia and Yale. 

According to researchers involved, the Biohub hopes to invest resources into long-term biomedical research. 

“The idea of these Biohubs is to bring people together from multiple disciplines, and tackle challenges that will take 10 to 15 years of risky work,” said John Tsang, the director of Yale Center for Systems and Engineering Immunobiology, who is on the leadership committee for the CZ Biohub NY and wrote the original research proposal for the project.

CZ Biohub NY is the latest research institute in the Chan Zuckerberg Biohub Network. 

While each of the four institutes in the CZ Biohub Network focuses on different projects, the general goal of the research is to understand the mechanisms of disease and develop technology to help diagnose and treat diseases. 

The CZ Biohub NY will focus on studying and bioengineering immune cells. As a systems immunologist, Tsang told the News he is interested in using the Biohub to explore how immune system cells detect and monitor disease in the body, even before disease symptoms are visible through outward observation. 

“The immune system has been implicated in basically all diseases, because it’s the system where you have cells that go around your whole body, and the cells are monitoring what’s going on,” Tsang said. 

Part of the Biohub’s work will be to understand how immune cells travel through the body and how they respond to new biological information.

The Biohub team will also work to engineer immune cells to detect abnormalities in tissues and repair damage before it leads to disease, according to Tsang. Tsang told the News that this kind of technology could be applied to predict and treat diseases including cancer and autoimmune conditions such as type one diabetes. 

“It’s really like getting the Magic School Bus to work,” Tsang said. “That’s really super exhilarating in my mind.” 

According to Andrea Califano, the president of the Chan Zuckerberg Biohub New York and the former chair of the Department of Systems Biology at Columbia University Irving Medical Center, this kind of research is not being done anywhere else. 

Califano pointed to two major characteristics that make the Biohub’s goals achievable: its funding structure and collaborative model.

“It’s an entire new level of science that has to be achieved,” Califano said. 

Funded through philanthropy, the Biohub operates on a predetermined budget, said Califano, allowing it to conduct high-risk research that would not typically be funded through public research grants such as the ones awarded through the National Institutes of Health. This funding allows the Biohub to do research far beyond the scale of a university lab, Califano said. 

Califano said the Biohub draws together three complementary research institutions that have strengths in different areas of biology, immunology and data science to create a multi-institutional, interdisciplinary collaboration. 

“When I initially was brought into the idea, I thought it was a little bit sci-fi, but then when I started working with my colleagues, we realized that all the main ingredients were actually there,” Califano told the News. “This is science that no university in isolation could perform.”

This scale of research is ambitious because there is still a lot about the immune system that scientists do not understand, said Andrew Wang, a physician and professor of rheumatology at the School of Medicine, who is not affiliated with the CZ Biohub New York. 

The same immune cell can behave differently depending on the situation and location in the body, Wang added — a phenomenon that still baffles researchers. As a result, Wang said that the kind of research the Biohub will do will need to be a “simultaneous dive” into all aspects of the immune system. But Wang said he is optimistic about what the initiative could accomplish. 

“It’s efforts like this that end up making big impacts,” he added. “I’m excited to see what comes out of it.” 

The Chan Zuckerberg Biohub Network program was announced by Priscilla Chan and Mark Zuckerberg on Dec. 7, 2021. 

Update, Nov. 11: This article was updated to reflect Andrea Califano’s current title. It was also updated to clarify language regarding the CZ Biohub Network and its leadership.

The post Yale joins new Zuckerberg-funded biomedical research collaboration  appeared first on Yale Daily News.

But according to a Yale study published this summer, patients who were Black, Hispanic or Latino, Spanish-speaking or insured by Medicaid were more likely to be skipped in emergency room lines. Those patients also had a higher likelihood of being treated in hallways and leaving before treatment was complete.

The study looked at over 90,000 cases of queue jumping at the Yale New Haven Hospital’s emergency department between 2017 and 2020 and adds to a growing body of research pointing out disparities in the timing and quality of care received in the emergency room. 

According to Hazar Khidir, instructor of emergency medicine at the School of Medicine and a co-author of the study, these disparities are often rooted in systemic problems like individual bias, structural racism and economic inequality.

“It’s not just a triage nurse being racist toward a patient,” Khidir said. “There are issues at the structural level.” 

The queue

When a patient first enters the emergency room, they go through a screening process to ensure that they will receive the appropriate level of care.  

According to Rohit Sangal, lead author of the study and the associate medical director of Yale New Haven Hospital’s adult emergency department, emergency rooms typically use a metric called the Emergency Severity Index, or ESI, to triage patients. 

The ESI scale is ranked from one to five, with the most ill patients, such as those experiencing a stroke or heart attack, categorized as ESI 1. The least sick patients, such as those requiring a medication refill, are categorized as ESI 5.  If there is no space in the emergency room — which can happen when there is a shortage of staff, a shortage of beds or when many patients arrive at once — a line forms. This line, said Sangal, is known as a “queue.” 

Patients in a queue are seen by a doctor according to how sick they are and when they arrive. But still, unexplained queue jumps — the term the researchers used to describe people being skipped in line — might occur. The data used in the study do not specify why someone was jumped. 

But queue jumps can happen for a variety of reasons, said Lesley Meng, a professor at the School of Management and co-author of the study.  One legitimate reason for a queue-jump occurs when a waiting patient’s condition worsens. For example, if a patient experiences a seizure or another emergency situation in the waiting room, according to Sangal, they would be jumped in line. 

“The waiting process is dynamic,” Sangal said. “A patient’s clinical status may change, which changes where they need to be in the queue.” 

Even without the possibility of being skipped in line, patients frequently avoid the emergency room if they are worried they might have to wait for several hours, said Caitlin Donovan, a spokesperson for the National Patient Advocate Foundation. 

If a patient experiences long wait times, they might be reluctant to return to the hospital to treat their health problems.

Ignoring health problems can have serious consequences for both the patient and the system, Donovan said, including “worse health outcomes and ultimately higher costs.” 

“Ideally, there’s a treatment space for everyone when they come in,” Sangal told the News.

However, Meng also noted that it is often difficult for patients to tell if they are being skipped over for medical reasons or for other ones — including socioeconomic privilege.

For example, Khidir explained, a privileged patient might have their primary care doctor call the emergency room ahead of their visit to make sure they are seen quickly. Someone without a primary care doctor might have to wait longer. 

She also pointed to one of the study’s findings: patients insured by Medicaid are more likely to be queue-jumped, even though emergency room providers are typically unaware of a patient’s insurance status. 

According to Khidir, this is because Medicaid and insurance status can affect the triage process through indirect means. 

For instance, patients on Medicaid might have a more difficult time finding a doctor who accepts their insurance. If they don’t have a regular doctor, they might not receive a diagnosis for conditions they have, Khidir explained. By the time they are seen in the emergency room, they would appear to be less sick than they actually are.

“A triage nurse doesn’t need to see if a patient has Medicaid for that to have an effect on the triage process,” Khidir said. 

Addressing inequalities 

The Yale queue-jumping study is far from the first research to point out disparities in the emergency room. For example, a study published last September found that non-white patients received less urgent ESI triage scores than white patients, while another study from Boston University found that Black and Hispanic patients experience longer emergency-room wait times than white patients.

Once emergency room inequalities are documented through studies like the queue-jumping study, hospital administrators then decide how to address disparities. 

Howard Forman, a professor of radiology at the School of Medicine and the director of Yale’s healthcare management program, said that there is one obvious solution to the problem.

“Queue jumping wouldn’t matter at all if there were no queues,” Forman told the News.

In addition to getting rid of wait times, Forman said the healthcare industry should prioritize repairing structural inequalities that disadvantage patients from low-income and marginalized backgrounds. Though he acknowledged that those issues are difficult to correct immediately, removing queues and structural inequality should remain on “our short list of problems we’re working on,” Forman said. 

In the short term, other solutions could include changing the triage system, said Chris Chmura, the current manager of clinical projects and education at YNHH, who previously worked as a triage and trauma nurse in the emergency department. 

The ESI system, developed in 1999, is now over 20 years old. Chmura believes that it is not well equipped to handle today’s patients with more complex health needs.

“We’re using tools that were designed in a totally different healthcare setting,” said Chmura. 

Over the past year, YNHH has shifted to a machine-learning tool that uses electronic health records to make more informed triage decisions. 

Chmura said that this tool can help patients receive triage rankings called acuity scores, which are based on the algorithm’s predictions of their health outcomes and the risk of an adverse event. 

“We’ve optimized our front-end flow and our triage process and physician processes,” said Beth Liebhardt, the executive director of emergency services at YNHH, and a co-author on the study. 

Liebhardt said that the data used in the queue-jumping study were collected before 2020. Since then, YNHH has begun to shift away from ESI rankings to a hybrid triage system that uses machine learning to rank patients. 

Liebhardt believes that the study’s findings might change if repeated today.

“With a different timeframe, I think those numbers would look different,” Liebhardt said. 

In addition to machine learning, other triage strategies, such as a split-flow model where patients are separated by the severity of their health needs, might help reduce bias in the emergency department, said Sangal. 

Another solution, according to Forman, is for hospitals to hire multilingual patient advocates to staff emergency department waiting rooms. These advocates could translate for patients who are uncomfortable with English and alert the triage nurse if a patient’s condition worsens.

For Forman, it’s also important that any new solution for addressing inequalities in the emergency room is thoroughly researched. He said it “would be best” to test each idea one by one to see which are meaningful and which are less so.  

Meng confirmed that she and other researchers from the medical school, the School of Management and YNHH are investigating changes made in the YNHH emergency room and their effect on patients. 

“We’re essentially studying every little piece of a patient’s journey through the ED to try to understand whether there are inefficiencies, whether patients are harmed during their waiting, and to come up with ways to make things better,” Meng said.

The study was published in July in the scientific journal JAMA Network Open and

conducted by researchers from the School of Medicine, the School of Management and Yale New Haven Hospital.

The post Black, Hispanic and low-income patients deprioritized in YNHH emergency room lines, study finds appeared first on Yale Daily News.

The report, which was released Sept. 14, documented the dumping of light-toned objects — likely body bags containing human remains — at a site reported to be a mass grave. Through data from open-source and publicly available sources including social media posts and satellite images, the report confirms humanitarian volunteer eye-witness accounts of bodies being dumped into water surrounding El-Geneia. 

“The situation in Sudan deserves a lot of attention and needs to be documented,” said Kaveh Khoshnood, the faculty director of the HRL at Yale. “This [report] is essential for policymakers to make correct decisions.” 

Nataniel Raymond, the executive director of the HRL and overseer of the project, explained that the violence in Sudan is partly rooted in an ongoing power struggle between the Sudanese Armed Forces and the Rapid Support Forces, a paramilitary organization in Sudan. He added that the violence also stems from the RSF committing a series of ethnically-based killings of non-Arab people, targeting the Masalit and Burgo tribes in West Darfur.

The report investigated alleged grave sites the RSF used for mass burials of people from the Masalait and Burgo tribes. Though the report does not explicitly confirm who was responsible, the researchers note that the timing of their findings corresponds with credible accounts suggesting that the RSF massacred Sudanese citizens and committed ethnic violence against non-Arab communities, particularly Masalit people.  

According to Khoshnood, the HRL, which operates as part of the State Department’s Conflict Observatory program, primarily monitored the Russo-Ukrainian war until April 2023, when conflict broke out in Sudan. Since then, the HRL has released a series of reports recording human rights violations during the ongoing conflict. 

Another HRL report, published Aug. 31, confirmed on-the-ground reporting of a dire humanitarian crisis in Nyala, the capital city of South Darfur. According to Raymond, civilians are caught between SAF and RSF clashes, and aid cannot enter the city. Satellite images included in the report revealed giant holes in the roofs of Nyala’s mosques, homes and hospitals — the result of artillery fire. 

“Civilians can’t flee, and humanitarian aid can’t enter Nyala,” Raymond said. “The civilians are being killed and wounded with only one functioning hospital left in operation.”

Raymond told the News that the HRL uses multiple data sources to produce their reports. Open-source information — such as social media and local news stories — constitutes one portion of the data. However, Raymond noted that open-source information is especially difficult to access in Sudan because of widespread telecommunications and internet outages.  

The researchers also gather information from publicly available sources, such as NASA’s thermal sensors, which detect fires. They use commercial imagery from satellites, including those by Planet and Maxar Technologies. The imagery varies from high to middle or low resolution, Raymond said. The researchers can also choose to use different sensors depending on what they want to target. 

“It’s like golfing,” he said. “We use different types of clubs to hit different types of holes.” 

Some of the data collected by the HRL seeks to verify open-source information. Other times, Raymond said, the researchers notice something in an image and try to find open-source information to confirm or explain what they found. They also look for evidence of an event they suspect might be happening. The image collecting happens in real time, often in six to 12 hour cycles, but weather conditions, especially clouds present during the rainy season, can make it impossible to see. 

David Simon, the director of the Genocide Studies program at Yale, who was not involved in the report, told the News that the HRL’s research may help the U.S. State Department make a determination of genocide in Sudan.

“There’s every indication that there’s an attempt to destroy the Masalit tribe and perhaps others as well, which is a core qualification of genocide,” Simon said. “It’s happening now, and there’s this almost inexplicable lack of reaction by global powers.”  

Simon said that, though the findings from the HRL reports are “absolutely harrowing” and describe “international crimes,” the lab cannot make legal judgments or determine policy. That, he said, is the role of the State Department and other institutions.

The State Department uses a framework, which is published on its website, to assess atrocity risk, distinguishing between war crimes, crimes against humanity and genocide. The atrocity risk assessment framework guides government officials to examine at key actors and targeted groups case by case to help develop atrocity prevention policy making. 

“It feels to me that the HRL passes on report after report to folks that have the capacity to make a policy or legal judgment, and it’s just not happening,” Simon said. “The reports deserve much more traction than they’re getting. 

Raymond told the News that the State Department has issued “multiple statements” based on the August report. Additionally, he said that the HRL’s reports have been used as part of ceasefire negotiations, investigations by the International Criminal Court into events in Darfur and for efforts delivering humanitarian aid in Sudan. 

The creation of the Conflict Observatory, a non-governmental organization that analyzes evidence of war crimes and other atrocities, was announced by the State Department on May 17, 2022. 

The post “Absolutely harrowing”: Yale public health researchers use social media posts and satellite images to corroborate human rights atrocities in Sudan appeared first on Yale Daily News.

The watch on your wrist ticks steadily throughout the day, collecting and transmitting information about your heart rate, oxygen saturation and the levels of sugar in your blood. Sensors scan your face and body, making inferences about your state of health.

By the time you see a doctor, algorithms have already synthesized this data and organized it in ways that fit a diagnosis, detecting health problems before symptoms arise. 

We aren’t there yet, but, according to Harlan Krumholz, a professor of medicine at the School of Medicine, this could be the future of healthcare powered by artificial intelligence.

“This is an entirely historic juncture in the history of medicine,” Krumholz said. “What we’re going to be able to do in the next decades, compared to what we have been able to do, is going to be fundamentally different and much better.” 

Over the past months, Yale researchers have published a variety of papers on machine learning in medicine, from wearable devices that can detect heart defects to algorithms that can triage COVID-19 patients. Though much of this technology is still in development, the rapid surge of AI innovation has prompted experts to consider how it will impact healthcare in the near future. 

Questions remain about the reliability of AI conclusions, the ethics of using AI to treat patients and how this technology might transform the healthcare landscape. 

Synergy: human and artificial intelligence at Yale

Two recent Yale studies highlight what the future of AI-assisted health care could look like. 

In August, researchers at the School of Medicine developed an algorithm to diagnose aortic stenosis, a narrowing of a valve in the body’s largest blood vessel. Currently, diagnosis usually entails a preliminary screening by the patient’s primary care provider and then a visit to the radiologist, where the patient must undergo a diagnostic doppler exam.

The new Yale algorithm, however, can diagnose a patient from just an echocardiogram performed by a primary care doctor.

“We are at the cusp of doing transformative work in diagnosing a lot of conditions that otherwise we were missing in our clinical care,” said Dr. Rohan Khera, senior author of the study and clinical director of the Yale Center for Outcomes Research & Evaluation, CORE. “All this work is powered by patients and their data, and how we intend to use it is to give back to the most underserved communities. That’s our big focus area.”

The algorithm was also designed to be compatible with cheap and accessible handheld ultrasound machines, said lead author Evangelos Oikonomou, a clinical fellow at the School of Medicine. This would bring first-stage aortic stenosis testing to the community, instead of being limited to those that are referred to a skilled and potentially expensive radiologist. It could also allow the disease to be diagnosed before symptoms arise. 

In a second study, researchers used AI to support physicians in hospitals by predicting COVID-19 outcomes for emergency room patients — all within 12 hours. 

According to first author Georgia Charkoftaki, an associate research scientist at the Yale School of Public Health, hospitals often run out of beds during COVID-19 outbreaks. AI-powered predictions could help determine which patients need inpatient care and which patients can safely recover at home.

The algorithm is also designed to be adaptable to other diseases. 

“When [Respiratory Syncytial Virus] babies come to the ICU, they are given the standard of care, but not all of them respond,” Charkoftaki said. “Some are intubated, others are out in a week. The symptoms [of RSV] are similar to COVID and so we are working on a study for clinical metabolomics there as well.”

However, AI isn’t always accurate, Charkoftaki admitted.

As such, Charkoftaki said that medical professionals need to use AI “in a smart way.” 

“Don’t take it blindly, but use it to benefit patients and the discovery of new drugs,” Charkoftaki told the News. “You always need a brain behind it.” 

Machines in medicine

Though the concept of artificial intelligence has existed since mathematician Alan Turing’s work in the 1950s, the release of ChatGPT in November 2022 brought AI into public conversation. The chatbot garnered widespread attention, reaching over 100 million users in two months.

According to Lawrence Staib ENG ’90, a professor of radiology and biomedical engineering, AI-powered healthcare does not yet consist of asking a sentient chatbot medical questions. Staib, who regularly uses machine learning models in his research with medical imaging, says AI interfaces are more similar to a calculator: users input data, an algorithm runs and it generates an output, like a number, image, or cancer stage. The use of these algorithms is still relatively uncommon in most medical fields.

While the recent public conversation on AI has centered around large language models — programs like ChatGPT which are trained to understand text in context rather than as isolated words — these algorithms are not the focus of most AI innovation in healthcare, Staib said. 

Instead, researchers are using machine learning in healthcare to recognize patterns humans would not detect. When trained on large databases, machine learning models often identify “hidden signals,” said David van Dijk, an assistant professor of medicine and computer science. In his research, van Dijk works to develop novel algorithms for discovering these hidden signals, which include biomarkers and disease mechanisms, to diagnose patients and determine prognosis. 

“You’re looking for something that’s hidden in the data,” van Dijk said. “You’re looking for signatures that may be important for studying that disease.” 

Staib added that these hidden signals are also found in medical imaging. 

In a computerized tomography — or CT — scan, for example, a machine learning algorithm can identify subtle elements of the image that even a trained radiologist might miss. 

While these pattern recognition algorithms could be helpful in analyzing patient data, it is sometimes unclear how they arrive at conclusions and how reliable those conclusions are. 

“It may be picking up something, and it may be pretty accurate, but it may not be clear what it’s actually detecting,” Staib cautions.

One famous example of that ambiguity occurred at the University of Washington, where researchers designed a machine learning model to distinguish between wolves and huskies. Since all the images of wolves were taken in snowy forests and all the images of huskies were taken in Arizona, the model learned to identify the species based on their environment. When the algorithm was given an image of a husky in the snow, it was always classified as a wolf. 

To address this issue, researchers are working on explainable artificial intelligence: the kind of program, Staib said, that “not only makes a judgment, but also tells you how it made that judgment or how confident it is in that judgment.”

Experts say that the goal of a partnership between human experts and AI is to reduce human error and clarify AI’s judgment process. 

“In medicine, well-intended practitioners still sometimes miss key pieces of information,” Krumholtz said.

Algorithms, Krumholtz said, can make sure that nothing “falls through the cracks.” 

But he added the need for human oversight will not go away. 

“Ultimately, medicine still requires intense human judgements,” he said. 

Big data and its pitfalls

The key to training a successful machine-learning model is data — and lots of it. But where this data comes from and how it is used can raise ethical questions, said Bonnie Kaplan, a professor of biostatistics and faculty affiliate at the Solomon Center for Health Law and Policy at Yale Law School.

The Health Insurance Portability and Accountability Act, or HIPPA, regulates patient data collected in healthcare institutions, such as hospitals, clinics, nursing homes and dentists offices, Kaplan said. If this data is scrubbed of identifying details, though, health institutions can sell it without patient consent. 

This kind of scrubbed patient information constitutes much of the data with which health-related machine learning models are trained. 

Still, health data is collected in places beyond healthcare institutions, like on period tracking apps, genetics websites and social media. Depending on the agreements that users sign — knowingly or not — to access these services, related health data can be sold with identifying information and without consent, experts say. And if scrubbed patient data is combined with this unregulated health data, it becomes relatively easy to identify people, which in turn poses a serious privacy risk.

“Healthcare data can be stigmatizing,” Kaplan told the News. “It can be used to deny insurance or credit or employment.”

For researchers, AI in healthcare raises other questions as well: who is responsible for regulating it, what privacy protections should be in place and who is liable if something goes wrong.

Kaplan said that while there’s a “general sense” of what constitutes ethical AI usage, “how to achieve [it], or even define the words, is not clear.”

While some, like Krumholz, are optimistic about the future of AI in healthcare, others like Kaplan point out that much of the current discourse remains speculative. 

“We’ve got all these promises that AI is going to revolutionize healthcare,” Kaplan said. “I think that’s overblown, but still very motivating. We don’t get those utopian dreams, but we do get a lot of great stuff.”

Sixty million people use ChatGPT every day.

The post Yale researchers investigate the future of AI in healthcare appeared first on Yale Daily News.