The three-day virtual conference, held Aug. 4-6, will explore practical strategies for using emerging technologies to strengthen teaching and learning while keeping students at the center of the classroom experience. This year’s theme, 鈥淧owering Progress,鈥� focuses on helping educators prepare for the rapidly evolving role of technology in education ahead of the new school year.

Highlights include:

• A plenary session 鈥� 鈥淧assive to Powerful: Teaching a Distracted Generation to Create鈥� 鈥� and workshops led by Kristen Brooks, a nationally recognized educational technology consultant.
• Another plenary session 鈥� 鈥淎I Workout: Strengthening Human-Centered Classrooms with AI鈥� 鈥� led by Maine Department of Education Emerging Technology Specialist Nicole Davis.
• Workshops and presentations by technology educators, specialists and coordinators from Maine, California, Texas and Abu Dhabi, as well as faculty, staff and researchers from 91爆料鈥檚 College of Education and Human Development and the Maine Center for Research in STEM Education (RiSE Center).听听

The Summer Technology Institute is held every year for students in two summer courses offered through the Instructional Technology graduate programs, both of which run July 20 to Aug. 14.

Educators who are not registered for the courses can still sign up to attend the institute for professional development credit and earn 25 contact hours or 2.5 continuing education units (CEUs). More information about the different registration options is available online.

Administered by the 91爆料 College of Education and Human Development, the Instructional Technology graduate programs are operated cooperatively by the 91爆料, the 91爆料 at Farmington and the University of Southern Maine. Courses are delivered remotely via 91爆料Online. 

For more information: 91爆料 Senior Lecturer in Instructional Technology Mia Morrison, mia.morrison@maine.edu

Speaking as part of the Maine College of Engineering and Computing Distinguished Lecture Series, co-hosted by the Office of the Vice President for Research, Michael F. Chiang said emerging technologies are making medical care more data-driven, consistent and accessible.

鈥淐linical practice and research are being rapidly reshaped by breakthroughs in artificial intelligence and data science,鈥� said Chiang, director of the National Eye Institute at the National Institutes of Health and elected member of the National Academy of Medicine.

Following the lecture, 91爆料 President Joan Ferrini-Mundy and Giovanna Guidoboni, interim vice president for research and dean of the Maine College of Engineering and Computing, joined Chiang for a panel discussion moderated by Alon Harris, director of the Barry Family Center for Ophthalmic Artificial Intelligence and Human Health and professor at Icahn School of Medicine at Mount Sinai in New York.

Ferrini-Mundy said the rapid pace of innovation is reshaping not only research, but the future of health care.

鈥淲e鈥檙e living in a time when clinical practice and research across fields 鈥� particularly in the medical field 鈥� are being rapidly reshaped by breakthroughs in artificial intelligence and data science,鈥� she said.

Harris, who is also faculty within the Graduate School of Biomedical Science and Engineering at 91爆料, reflected on the breadth of opportunity that exists across Maine and that 91爆料 is uniquely positioned to lead.

鈥淚 had been here before, but during this visit I discovered there is so much more,鈥� he said. 鈥淭his place is so motivating, from the biological and biomedical labs, to the full scale automated vehicles and 3D printed homes with smart health sensors. The level of people we met and the research interests were truly thought-provoking.鈥�

Guidoboni said Chiang鈥檚 work reflects the data-driven, interdisciplinary approach central to research at 91爆料. Over the past 16 years, Guidoboni and Harris have advanced mathematical modeling and data science, including studies on ocular blood flow, eye disease risk and noninvasive health monitoring, with the development of digital twins to help translate the advances of science into personalized medical care.

Their work reflects a broader shift toward using advanced analytics to better understand and treat complex health conditions.

鈥淒r. Chiang鈥檚 work exemplifies the power of combining clinical insight with data science to transform patient care,鈥� Guidoboni said. 鈥淗is leadership at the National Eye Institute is inspiring, especially as these innovations expand access and improve outcomes in rural communities like Maine.鈥�

Chiang said advances in imaging have transformed ophthalmology from a largely descriptive field into one grounded in quantitative data, allowing clinicians to better measure and analyze disease.

He pointed to retinopathy of prematurity 鈥� a condition that can cause blindness in infants 鈥� as an example of how artificial intelligence can improve care. Studies have shown that even experts reviewing the same retinal images often disagree on whether disease is severe.

鈥淭hat discrepancy is real,鈥� Chiang said. 鈥淎nd this is where AI can help doctors make diagnoses that are more accurate and more consistent.鈥�

He also highlighted emerging research suggesting that the eye may offer insights into broader health conditions. Because clinicians can directly observe blood vessels and nerves in the eye, researchers are exploring whether imaging can help predict diseases elsewhere in the body.

鈥淚f that鈥檚 really true and generalizable, then that鈥檚 remarkable,鈥� he said, referring to studies linking eye imaging to neurological disease.

Chiang emphasized that progress in AI depends on access to large, high-quality datasets and collaboration across institutions.

鈥淕arbage in, garbage out,鈥� he said, cautioning that poor-quality data can limit the effectiveness of AI tools.

He also noted that technology could help reduce administrative burdens on physicians, who often spend significant time entering information into electronic health records.

鈥淭he technologies will help automate some of those things,鈥� he said, 鈥渟o doctors can spend more of their focus on the patient.鈥�

Advances in technology are also reshaping how and where care is delivered, particularly in rural areas like Maine.

Chiang pointed to opportunities to expand care beyond traditional clinical settings through telehealth, remote monitoring and home-based tools, reducing the need for patients to travel long distances for care.

鈥淚npatient hospital stays are shorter than they ever used to be,鈥� he said.

Those shifts, he added, raise broader questions about how physicians are trained and how healthcare systems adapt as medicine becomes increasingly data-driven.

As AI continues to evolve, Chiang said its impact will extend beyond diagnosis to reshape research, education and care delivery.

Contact: David Nordman, david.nordman@maine.edu

Chaofan Chen, assistant professor of electrical and computer engineering at the 91爆料, aims to change that by creating AI systems that explain their results and learn from the people who use them.

To accomplish this, Chen is developing tools that show users how an AI model reaches a decision and that allow users to correct decisions when something appears inaccurate. His team will build AI systems that illustrate their reasoning processes when making predictions and adjust those reasonings based on user feedback.

The goal is a two-way conversation between AI that shows its work and people who can help improve it.

鈥淲e live in an exciting era of AI breakthroughs, and my mission is to create systems that don鈥檛 just give answers but reveal their reasoning and can improve themselves based on human feedback,鈥� said Chen, who received a National Science Foundation CAREER Award to support his work. 

The five-year, $584,034 project 鈥� 鈥淐AREER: Opening the Black Box: Advancing Interpretable Machine Learning for Computer Vision鈥� 鈥� aims to bring greater transparency and accountability to AI-powered computer vision systems used in everyday and high-stakes settings.

Modern computer-vision models can detect diseases, identify objects and generate images with remarkable accuracy, but they typically operate as 鈥渂lack boxes,鈥� offering little insight into how decisions are reached. That lack of interpretability prevents users from evaluating whether a choice was sound, identifying flawed assumptions or correcting mistakes. 

In fields such as health care, public safety and scientific research, those blind spots can pose serious risks.

鈥淚n high-stakes settings, black-box AI isn鈥檛 just a mystery 鈥� it鈥檚 a risk. When we can鈥檛 see how decisions are made, we can鈥檛 trust the outcomes that matter most,鈥� Chen said. 鈥淚n healthcare, for example, a black-box model recommending a diagnosis or treatment could leave clinicians guessing at its reasoning 鈥� an uncertainty that patients simply can鈥檛 afford. In this case, interpretability isn鈥檛 a luxury; it鈥檚 a safeguard for real people鈥檚 lives.鈥�

Chen鈥檚 project seeks to replace that opacity with clarity. Chen will develop multimodal models that provide richer, more accessible insights into the decision-making processes of AI systems. He also plans to design generative models that break down how images are created, rather than presenting only a final result.

The research will extend into reinforcement learning, exploring ways to ensure AI decision-making policies remain interpretable.

A major component of the project is strengthening human-AI interaction. Chen aims to create methods that allow users to correct a model鈥檚 reasoning directly and integrate that feedback into the training process so the system becomes more accurate and aligned with human expectations over time.

鈥淒r. Chen鈥檚 CAREER project tackles one of AI鈥檚 most urgent challenges, opening the black box so computer-vision systems explain their decisions in ways people can trust, especially in high-stakes settings,鈥� said Yifend Zhu, professor and chair of 91爆料鈥檚 Department of Electrical and Computer Engineering. 鈥淓qually exciting, he鈥檚 partnering with the Maine Mathematics and Science Alliance to bring interpretable AI into Maine classrooms, empowering teachers and inspiring the next generation of innovators.鈥�

As part of the award, Chen will collaborate with the Maine Mathematics and Science Alliance to develop high school lesson plans introducing responsible and interpretable AI concepts. The effort aims to help students understand not only how AI works but also how to question and guide it.

The project is jointly funded by the NSF Robust Intelligence and EPSCoR programs and will run through June 30, 2030.

Story by William Bickford, graduate student writer
Contact: Marcus Wolf, 207.581.3721; marcus.wolf@maine.edu

The system, named the Context-Guided Segmentation Network (CGS-Net), mimics the way human pathologists study cancer tissue to achieve accuracy in digital cancer diagnosis. Spearheaded by Jeremy Juybari, a doctoral candidate in electrical and computer engineering, and Josh Hamilton, a doctoral candidate in biomedical engineering, the tool introduces a deep learning architecture designed to interpret microscopic images of tissue with greater precision than conventional AI models.

Breast cancer is the second leading cause of cancer-related deaths in women, affecting one in eight over their lifetime. Diagnosis still relies on the microscopic inspection of chemically stained tissue samples, a process that requires expertise and time.

Two-thirds of the world鈥檚 pathologists are concentrated in only 10 countries, leaving large regions facing diagnostic delays that contribute to preventable deaths. In India, for example, roughly 70% of cancer deaths are linked to treatable risk factors compounded by limited access to timely diagnostics.

鈥淭his model integrates both detailed local tissue regions and broader contextual regions to improve the accuracy of cancer predictions in histological slides,鈥� Juybari said. 鈥淏y introducing a unique training algorithm and an innovative initialization strategy, this research demonstrates how incorporating surrounding tissue context can significantly enhance model performance. These findings reinforce the importance of holistic image analysis in medical AI applications.鈥�

Juybari and Hamilton described their work in a paper published in (part of the Nature portfolio), which they co-authored with faculty researchers Andre Khalil, Yifeng Zhu and Chaofan Chen. 

At the core of the system lies a dual-encoder model that mirrors the workflow of a pathologist examining a slide. Traditionally, pathologists gather data themselves by zooming in and out of the images they are examining. CGS-Net would do this simultaneously. One branch of the network processes a high-resolution image patch to capture cell-level details. The other examines a lower-resolution patch encompassing the surrounding tissue, the broader architectural context that helps specialists distinguish normal from malignant structures.

Each patch shares the same center pixel, ensuring that both views of the tissue align precisely. Together, they feed into a system of interconnected encoders and decoders that uses data from both the high and low resolution images for a complete analysis. 

To test the system, the research team at 91爆料 trained CGS-Net to analyze 383 digitized whole-slide images of lymph node tissue and determine which ones showed signs of breast cancer. It was also trained on how to differentiate between healthy and cancerous tissue. The tool consistently outperformed traditional single-input models.

鈥淥ur model, CGS-Net, successfully mimicked how a pathologist looks at histological samples, by using two encoders that simultaneously examine two views of different levels of magnification power,鈥� the researchers wrote in their paper.

While the current study focused on binary cancer segmentation, the team sees wide potential for expansion. Future research aims to incorporate additional resolutions, apply the system to multiclass tissue segmentation and test its adaptability across other cancer types. The architecture could also integrate multimodal data, such as radiology scans or molecular profiles, which are known to further boost diagnostic accuracy.

Beyond technical, the project underscores the interdisciplinary strength of 91爆料鈥檚 research ecosystem, merging engineering, computing and biomedical science to tackle global health disparities.

The study was supported by the National Cancer Institute and the National Science Foundation, and utilized computational resources from the 91爆料 System Advanced Research Computing, Security, and Information Management (ARCSIM). Both the dataset and source code are publicly available, ensuring transparency and collaboration across the scientific community.

As cancer diagnosis becomes increasingly digital, tools like CGS-Net promise to enhance, and not replace, human expertise. By teaching machines to see as doctors do, 91爆料 researchers are helping chart a future where early, accurate detection becomes accessible to all.

Juybari and Hamilton are among the many students who conduct research at Khalil鈥檚 CompuMAINE lab, using image analysis techniques to solve a range of problems at every scale, from cancer cells to galaxies. Learn about more work being done in the CompuMAINE lab when Isabelle Puccio’s episode of Life in the Pines debuts Dec. 10. Visit to watch the unscripted video series exploring what life is really like in the Pine Tree state.

Story by William Bickford, graduate student writer

Contact: Taylor Ward, taylor.ward@maine.edu 

While pursuing his bachelor鈥檚, master鈥檚 and doctoral degrees, Juybari, a San Diego native, worked for Faster Logic LLC, a small defense-focused research and development company in his hometown, providing web and engineering support. Two semesters into his Ph.D. program in 2021, Juybari paused his studies for five months to serve as the company鈥檚 acting CEO after its founder, Raymond Moberly, unexpectedly passed away. Juybari led the company through a government audit and handled operations and personnel.

鈥淪tepping into that role was unexpected, but it was important to me to support the work Raymond had built over seven years,鈥� Juybari said. 鈥淚t was a demanding time, and I learned a great deal about leadership, people and how research moves from concept to real-world development. After working through circumstances beyond my control, the company ultimately dissolved. Once things settled, I returned to 91爆料 to continue my Ph.D., which had always been my long-term plan.鈥�

After completing his undergraduate economics and interdisciplinary studies degree at San Diego State University, he sought to expand his technical knowledge and research capabilities, which ultimately led him to pursue graduate study at 91爆料. Once he completed the math degree, Juybari immediately began his Ph.D. in electrical and computer engineering. 

鈥淲hen you have a good background in math, it makes learning AI much easier,鈥� Juybari said. 鈥淵ou start to realize AI is a bunch of matrix multiplications. Without that strong foundation, it can look like magic.鈥�

While working at the CompuMAINE Lab on coding and AI research, he learned how this technology could help save lives through improved AI for cancer diagnosis and reduce healthcare disparities. 

鈥淚 originally wanted to study economics, but it was math that brought me here,鈥� Juybari said. 鈥淎s I got deeper into research, I realized how many people die from cancer, sometimes simply because they were missed due to healthcare disparities. That really stuck with me.鈥�

Juybari鈥檚 Ph.D. research focuses on AI for medical imaging and cancer detection. He developed the Context-Guided Segmentation Network (CGS-Net), a model that combines detailed tissue features with broader contextual regions to improve the identification of cancerous tissue in microscopic images of biopsied tissue.

Earlier this year, Juybari and his colleagues published their research in the journal (part of the Nature portfolio) in a paper titled 鈥淐ontext-guided Segmentation for Histopathologic Cancer Segmentation.鈥� The paper was featured by the for its innovative approach to improving AI accuracy in medical imaging. The study introduced a novel method in which the model learns how to integrate both local tissue features and broader contextual information, demonstrating how careful model design can enhance predictions in complex histological datasets.

鈥淥ne of the biggest challenges I鈥檝e seen in medical AI is the lack of common benchmarks,鈥� Juybari said. 鈥淚t鈥檚 kind of like the wild west, where researchers use different datasets, and  medical image datasets are often large and complex.鈥�

91爆料鈥檚 mentorship and resources have been central to Juybari鈥檚 success. His co-advisors, Andre Khalil and Yifeng Zhu, offered both guidance and freedom, allowing Juybari to explore ambitious ideas. The Advanced Research Computing, Security, and Information Management (ARCSIM) group provided the computing power and collaborative environment that enabled his research.

Collaboration has defined his graduate journey. Juybari鈥檚 partnership with fellow Ph.D. student Josh Hamilton has been a cornerstone of his research and personal life. They鈥檝e spent long nights tackling complex coding challenges, and have even shared key life moments.

鈥淚 couldn鈥檛 imagine 91爆料 without Josh,鈥� Juybari said. 鈥淲e work together on a majority of our research. Our strengths and weaknesses complement each other. We laugh a lot, it鈥檚 fun.鈥�

Juybari also met his wife, Simona Mitevska, while living in Stodder Hall in 2019. He was studying mathematics then, and she was pursuing master鈥檚 degrees in economics and global policy. Their shared love of numbers and research turned into a lasting relationship. Today, Mitevska works as a senior research analyst in 91爆料鈥檚 Office of Institutional Research and Assessment.

For Juybari, an interdisciplinary background and collaborative mindset are what drive him forward 鈥� whether leading a company or developing AI to fight cancer.

鈥淵ou can鈥檛 know it all,鈥� Juybari said. 鈥淓ven within AI, there are so many different parts to one model. You could be well-versed with one part, have an understanding of another, but not be an expert in everything. You have to work with teams and trust that others will know things you don鈥檛. If you try to do everything yourself, then what鈥檚 the point of working in a team?鈥�

Looking ahead, Juybari remains open to where his path leads next.

鈥淚 like to keep an open mind,鈥� Juybari said. 鈥淢y interdisciplinary background has taught me to see challenges from different angles. I鈥檓 driven more by curiosity and problem-solving than by following a fixed path, and I鈥檓 excited to see where that leads next. 鈥� 

Story by William Bickford, graduate student writer. 

Contact: Marcus Wolf, 207.581.3721; marcus.wolf@maine.edu 

Daigle, who is double majoring in marketing and business information systems and security management, led his team to victory in Sherwin-Williams鈥� Eastern Division Internship Competition with a chatbot service they designed for the company.

Hosted Aug. 1 in West Chester, Pennsylvania, the competition brought together the top intern teams from across the North, Central, Southern, and Eastern regions 鈥� representing more than 24 teams in total. Daigle and his teammates, Paige Martin from Michigan State University and Gracie Coker from Salisbury University, earned first place with their innovative project: Sher-Bot, an AI-powered customer service chatbot designed to enhance digital engagement and streamline support across Sherwin-Williams platforms.

鈥淲e noticed that the current chatbot on the Sherwin-Williams website was difficult to use and lacked the functionality that customers expect,鈥� said Daigle, who interned at the store in Rockland, Maine. 

After analyzing challenges customers experienced and comparing competitor solutions, the team realized there was a huge opportunity to improve the online experience. Sher-Bot was their solution. 

鈥淲e designed a more intuitive, AI-powered chatbot that delivers faster, accurate responses while operating 24/7, breaks down language barriers and generates up to nine times its monthly cost in revenue,鈥� Daigle said.

The project combined customer insight, competitor analysis, Return on Investment (ROI) projections and a five-year implementation roadmap, which students had to present in front of a panel of judges as well as demonstrate their AI tool. Judges 鈥� including Curt Kaucher, president of Sherwin-Williams鈥檚  eastern division 鈥� praised the team鈥檚 professionalism, clear communication and data-driven approach .

鈥淭his was a huge win,鈥� said Annisa Fensin, district manager for Sherwin-Williams in Maine. 鈥淐onnor and his teammates found a blind spot in our customer engagement strategy and brought a timely, AI-driven solution to the table. Their research and presentation were outstanding. To come out on top of the entire Eastern Division speaks volumes about the quality of their work.鈥�

For Daigle, a student-athlete on 91爆料鈥檚 Cross Country and Track and Field teams, the experience offered far more than a summer internship. It helped sharpen his ability to lead, communicate and analyze, skills he鈥檒l carry into his senior year at the 91爆料 and beyond.

鈥淚 decided to intern at Sherwin-Williams because of the company鈥檚 strong reputation for leadership development and commitment to employee growth,鈥� he said. 鈥淭he kind of hands-on experience I had this summer was an incredibly rewarding experience.鈥�

“Sherwin-Williams has been around for more than 150 years, and many of our employees stay with us long-term,” Fensin said. “Our interns bring in a fresh perspective, and we truly value their contributions.鈥�

The Sherwin-Williams internship is a paid, 10-week program that pairs students with store managers, mentors and fellow interns to work on high-impact projects.  It鈥檚 one of many internship opportunities offered to students by employers statewide in partnership with the Maine Business School. Learn more about these opportunities on the school鈥檚 website. 

Contact: Melanie Brooks, melanie.brooks@maine.edu听

The app, developed by Jon Ippolito, a professor of new media at the 91爆料, was unveiled June 25 as a part of an initiative to engage faculty and students in projects that lead to healthier communities and ecosystems. Ippolito created “What Uses More” out of a personal frustration with what he described as the “polarized takes on AI and the environment.”

“The purpose of this app is not to provide definitive measures of AI energy and water use; the industry is notoriously tight-lipped about its footprint,” Ippolito said. “Instead, the goal is to help you visualize the environmental impact of what you do online, as well as to learn how factors like where you live or the type of prompt can dramatically influence the footprint of both AI and non-AI tasks.”

The app characterizes the impact of using AI to generate text, images and videos in relatable units like lightbulb-minutes for energy, which are equivalent to an incandescent bulb running for a minute, and cubic centimeters for water, which are comparable to a raindrop. 

Users can also compare energy and water consumption between using AI and other technology. For example, asking AI to create a three-second video can consume 25 times as much energy as charging a smartphone and twice as much as an hour-long Zoom call with ten people. Additionally, users can manipulate various parameters, such as the data center’s power source and the climate of its location, to observe how these choices impact energy and water usage. 

This tangible representation aims to make abstract environmental impacts more comprehensible to a wider audience. Helping users compare and contrast the impact of online activities amid their unique circumstances can lead to a more informed approach to responsible AI use. 

鈥淲e鈥檝e designed this tool to be less of a calculator and more of a conversation starter.鈥� said Ippolito. 鈥淭eachers can customize scenarios by region and energy source, while students can see how small shifts 鈥� a ChatGPT lookup versus a plain Google search, or an image drawn in Illustrator versus one generated by Midjourney 鈥� can lead to different footprints.鈥�

The app has already attracted more than 2,500 unique visitors from 39 countries, demonstrating growing public interest in the often-hidden environmental costs of technology.  

Looking ahead, Ippolito plans to expand the app to include the environmental footprint of more everyday activities, like driving a car or eating a hamburger, providing a broader context to AI’s impact. He also aims to develop custom versions of the app for iPhone and Android, and is interested in collaborating with 91爆料 students on these future iterations. 

This work aligns with his broader research into the impact of AI on creators 鈥� including writers, programmers and media makers 鈥� and seeks to illuminate the often-overlooked environmental costs in discussions about artificially generated or synthetic media. The project is part of the Stillwater Lab’s Ripple Initiative at 91爆料, which is co-directed by Ippolito and aims to bridge academic knowledge with local economic and ecological needs.

Ippolito鈥檚 research contributes to the university-wide , which connects researchers, educators and industry partners through webinars and the annual Maine AI Conference to advance responsible artificial intelligence innovation that benefits Maine and beyond.

Contact: Taylor Ward, taylor.ward@maine.edu

The 91爆料 Advanced Manufacturing Center (AMC) is helping manufacturers in the state integrate new robotics, such as AI, machine learning and automation technologies, into their operations. These mechanical helpers can take care of the three 鈥淒’s鈥� in manufacturing 鈥� work that is dull, dirty or dangerous. They don鈥檛 get bored, can鈥檛 be overworked and aren鈥檛 subject to injury.

In this episode of 鈥�The Maine Question鈥� podcast, John Belding, director of the AMC, and guests Brad Denholm, associate director of workforce development at the AMC, Ryan Lindsay, operations engineer at Ntension, and Peter Birch, mechanical engineering student and assistant at the AMC, explore Maine鈥檚 future of advanced manufacturing with robotics.

Listen to the podcast on , , , or 鈥淭he Maine Question鈥� website.听

What topics would you like to learn more about? What questions do you have for 91爆料 experts? Email them to mainequestion@maine.edu.

鈥淐hallenges and opportunities in using interpretable AI to develop relationship interventions鈥� was published in , the academic research journal of the National Council on Family Relations, as part of a special issue on AI in family life.

The use of AI in therapy is still in its infancy but has potential to provide families and couples with personalized support to strengthen bonds and overcome relationship problems, according to Daniel Puhlman, assistant professor of family studies in the 91爆料 College of Education and Human Development and the article鈥檚 lead author.

鈥淐ouples going through a separation, for example, where you have high emotions and high conflict, just being in the same space can be difficult, if not dangerous,鈥� Puhlman said. 鈥淚n a situation like that, AI鈥檚 ability to be interpretive and suggest therapeutic interventions or treatment measures could be a very powerful tool.鈥�

Puhlman notes that as more people become familiar with them, AI technologies are already integrating into family life and other types of relationships. However, he and his co-author, assistant professor of computer science Chaofan Chen, note that AI itself is a broad term that encapsulates several different technological systems and processes where machines are programmed to mimic human cognition and perform tasks that require human intelligence. 

鈥淭he specific sub-area of AI that we found most relevant to addressing family science problems is machine learning, which uses algorithms that allow computers to learn from datasets and make predictions or decisions based on that data,鈥� Puhlman said.

Strategies that incorporate human expertise and feedback 鈥� known as the human-in-the-loop technique 鈥� are important for improving the accuracy of machine learning models, he adds.

鈥淚t鈥檚 especially important in fields like family and relationship science, health care and law, where human judgment is critical,鈥� said Puhlman.  

According to the researchers, most of the AI-based technologies currently used in therapeutic contexts offer support to individuals rather than to couples or families. The tools that are most widely used are largely educational and not used for actual treatment or interventions 鈥� therapists can use AI to summarize session notes, for example. Part of the challenge is that while AI is good at interpreting and reporting data when it has a strict structure in place, human behavior is complicated.

鈥淛ust think about why human beings do the things we do, say the things we say, think the things we think, or how we interact with the world around us. Then put two or more of these complex, messy, dynamic creatures together, and you can see the challenges for a system that relies on a strict structure,鈥� Puhlman said.

Another challenge is that there are several different ways to practice therapy, and what works in one situation may not apply to another. 

鈥淵ou can train AI to use a therapeutic mode, but training it to know when it is appropriate to use and why to use it versus a different method gets quite complex,鈥� Puhlman said.

The authors explore these challenges, as well as ethical concerns about AI and privacy; AI providing inaccurate information; and AI and bias.

They also highlight opportunities in four specific areas to further develop AI in the context of relationship therapy: diagnosing relationship problems, providing autonomous treatment to clients, predicting successful treatment outcomes and using biomarkers to monitor client reactions. 

鈥淎lready we鈥檝e seen some researchers using machine learning to study parenting relationships or diagnose relationship problems. Chatbots could provide treatment to underserved populations or help therapists maximize positive outcomes for their clients. All of the challenges we identified still need to be addressed, but these technologies have the potential to make a real difference,鈥� Puhlman said.

Contact: Casey Kelly, casey.kelly@maine.edu

The study published in the in May evaluated more than 7,000 anonymized medical queries from the United States and Australia. The findings outlined where the technology showed promise and what limitations need to be addressed before AI is unleashed on patients 鈥� and may inform the future development of AI tools, clinical procedures and public policy. The study also informs efforts to use AI to support healthcare professionals at a time when workforce shortages are growing and clinician burnout is increasing.

The results showed that the accuracy of most AI-generated responses aligned with expert standards of information, especially with factual and procedural queries, but often struggled with 鈥渨hy鈥� and 鈥渉ow鈥� questions.  

The study also found that while responses were consistent within a given session, inconsistencies appeared when users posed the same questions in later tests. These discrepancies raise concerns, particularly when a patient鈥檚 health is at stake. The findings add to a growing body of evidence that will define AI鈥檚 role in healthcare.

鈥淭his isn鈥檛 about replacing doctors and nurses,鈥� said C. Matt Graham, author of the study and associate professor of information systems and security management at the Maine Business School. 鈥淚t鈥檚 about augmenting their abilities. AI can be a second set of eyes; it can help clinicians sift through mountains of data, recognize patterns and offer evidence-based recommendations in real time.鈥�

The study also compared health metrics, including patient satisfaction, cost and treatment efficacy, across both countries. In Australia, which has a universal healthcare model, patients reported higher satisfaction and one-quarter of cost compared to those in the U.S., where patients also waited twice as long to see providers. Graham notes in the study that health system, regulatory and cultural differences like these will ultimately influence how AI is received and used and that models should be trained to account for these variations. 

Artificial emotional intelligence

While the accuracy of a diagnosis matters, so does the way it is delivered. In the study, AI responses frequently lacked the emotional engagement and empathetic nuance often conveyed by human clinicians. 

The length of AI responses were strikingly consistent, with most varying between 400 and 475 words. Responses by human clinicians showed far more variation, with more concise answers written in response to simpler questions. 

Vocabulary analysis revealed that AI regularly used clinical terms in its responses, which may be hard to understand or feel insensitive to some patients. In situations involving topics such as mental health or terminal illness, AI struggled to convey the compassion that is critical in effective patient-provider relationships. 

鈥淗ealthcare professionals offer healing that is grounded in human connection, through sight, touch, presence and communication 鈥� experiences that AI cannot replicate,鈥� said Kelley Strout, associate professor of 91爆料鈥檚 School of Nursing, who was not involved in the study. 鈥淭he synergy between AI and clinicians鈥� judgment, compassion and application of evidence-based practice has the potential to transform healthcare systems but only if accompanied by rigorous standards, ethical frameworks and safeguards to monitor for errors and unintended consequences.鈥�

A stretched health system

The study arrives amid widespread and growing shortages in the U.S. healthcare workforce. Across the country, patients face long wait times, high costs and a shortage of primary care and specialty providers. These barriers are particularly acute in rural regions, where limited access often leads to delayed diagnoses and worsening health outcomes.

A published by the Health Resources and Services Administration in 2024 stated that Maine鈥檚 primary care doctor-to-patient ratio ranks 47th in the nation, with more than 115 patients for each provider. While a growing number of nurse practitioners and physician assistants are stepping in to fill the gap, demand for care is growing faster. A 2024 Maine Nursing Action Coalition indicated the state will face a shortage of more than 2,800 nurses by 2030. 

Strout said that while AI could help improve patient access and alleviate challenges 鈥� such as burnout, which affects of primary care physicians in the U.S. 鈥� its use must be carefully approached.

Prioritizing providers and patients

AI-powered tools could support round-the-clock virtual assistance and complement provider-to-patient communication through tools like online patient portals, which have skyrocketed in popularity since 2020. The technology, however, also raises fears of job displacement, and experts warn that rapid implementation without ethical guardrails may exacerbate disparities and compromise care quality.

鈥淭echnology is only one part of the solution,鈥� said Graham. 鈥淲e need regulatory standards, human oversight and inclusive datasets. Right now, most AI tools are trained on limited populations. If we鈥檙e not careful, we risk building systems that reflect and even magnify existing inequalities.鈥�

Strout added that as health care systems integrate AI into clinical practice, administrators must ensure that these tools are designed with patients and providers in mind. Lessons from past integration of technology, which at times failed to enhance care delivery, offer valuable guidance for AI developers.

鈥淲e must learn from past missteps. The electronic health record (EHR), for example, was largely developed around billing models rather than patient outcomes or provider workflows,鈥� Strout said. 鈥淎s a result, EHR systems have often contributed to frustration among providers and diminished patient satisfaction. We cannot afford to repeat that history with AI.鈥�

Other factors, such as accountability for mistakes and patient privacy, are top of mind for medical ethicists, policy makers and AI researchers. Solutions to these ethical questions may vary depending on where they are adopted to account for different cultural and regulatory environments.

A growing number of experts call for clearer guidance on AI deployment in clinical settings and beyond, including protocols for transparency, accountability and consent. These issues will take center stage at the on June 13. Organizers encourage anyone with a stake in Maine鈥檚 future, from educators to tech developers, to register by the June 6 deadline to join this pivotal conversation. 

As AI continues to develop, many experts believe it will enhance the service efficiency and decision-making that providers offer to patients. The study鈥檚 findings support the growing consensus that AI鈥檚 limited ethical and emotional adaptability means that human clinicians remain indispensable. Graham says that, in addition to improving the performance of AI tools, future research should focus on managing ethical risks and adapting AI to diverse healthcare contexts to ensure the technology augments rather than undermines human care.

“Technology should enhance the humanity of medicine, not diminish it,” Graham said. “That means designing systems that support clinicians in delivering care, not replacing them altogether.”

Contact: Erin Miller, erin.miller@maine.edu