Blood clots form naturally as a way to stop bleeding when someone is injured. But blood clots in patients with medical issues, such as mechanical heart valves or other heart conditions, can lead to a stroke or heart attack. That’s why millions of Americans take blood-thinning medications, such as warfarin, that make it harder for their blood to clot.

Warfarin isn’t perfect, however, and requires patients to be tested frequently to make sure their blood is in the correct range 鈥� blood that clots too easily could still lead to a stroke or a heart attack while blood that doesn’t clot can lead to extended bleeding after an injury. To be tested, patients either have to go to a clinic laboratory or use a costly at-home testing system.

Researchers at the 天美影视传媒 have developed a new blood-clotting test that uses only a single drop of blood and a smartphone vibration motor and camera. The system includes a plastic attachment that holds a tiny cup beneath the phone’s camera.

A person adds a drop of blood to the cup, which contains a small copper particle and a chemical that starts the blood-clotting process. Then the phone’s vibration motor shakes the cup while the camera monitors the movement of the particle, which slows down and then stops moving as the clot forms. The researchers showed that this method falls within the accuracy range of the standard instruments of the field.

The team Feb. 11 in Nature Communications.

“Back in the day, doctors used to manually rock tubes of blood back and forth to monitor how long it took a clot to form. This, however, requires a lot of blood, making it infeasible to use in home settings,” said senior author , UW professor in the Paul G. Allen School of Computer Science & Engineering. “The creative leap we make here is that we’re showing that by using the vibration motor on a smartphone, our algorithms can do the same thing, except with a single drop of blood. And we get accuracy similar to the best commercially available techniques.”

Doctors can rank blood-clotting ability using two numbers:

• the time it takes for the clot to form, what’s known as the “prothrombin time” or PT
• a ratio calculated from the PT that allows doctors to more easily compare results between different tests or laboratories, called the “international normalized ratio” or INR

“Most people taking this medication are taking it for life. But this is not a set-and-forget type of thing 鈥� in the U.S., most people are only in what we call the ‘desirable range’ of PT/INR levels about 64% of the time,” said co-author , assistant professor of anesthesiology and pain medicine in the UW School of Medicine. “This number is even lower 鈥� only about 40% of the time 鈥� in countries such as India or Uganda where there is less frequent testing. How can we make this better? We need to make it easier for people to test more frequently and take ownership of their health care.”

Patients who can monitor their PT/INR levels from home would only need to go to see a clinician if the test suggested they were outside of that desirable range, Michaelsen said.

The researchers wanted an inexpensive device that could work similarly to how at-home blood sugar monitors work for people with diabetes: A person can prick their finger and test a drop of blood.

“We started by vibrating a single drop of blood and trying to monitor waves on the surface,” said lead author , a UW doctoral student in the Allen School. “But that was really challenging with such a small amount of blood.”

The team added a small copper particle because its motion was so much more reliable to track.

“As the blood clots, it forms a network that tightens. And in that process, the particle goes from happily bouncing around to no longer moving,” Michaelsen said.

To calculate PT and INR, the phone collects two time stamps: first when the user inserts the blood and second when the particle stops moving.

“For the first time stamp, we’re looking for when the user inserts a capillary tube containing the sample in the frame,” Chan said. “For the end of the measurement, we look directly at the interior of the cup so that the only movement within those frames is the copper particle. The particle stops moving abruptly because blood clots very quickly, and you can observe that difference between frames. From there we can calculate the PT, and this can be mapped to INR.”

The researchers tested this method on three different types of blood samples. As a proof of concept, the team started with plasma, a component of blood that is transparent and therefore easier to test. The researchers tested plasma from 140 anonymized patients at the 天美影视传媒 Medical Center. The team also examined plasma from 79 patients with known blood-clotting issues. For both these conditions, the test had results that were similar to commercially available tests.

To mimic what a patient at home would experience, the team then tested whole blood from 80 anonymized patients at both Harborview and the 天美影视传媒 medical centers. This test also yielded results that were in the accuracy range of commercial tests.

This device is still in a proof-of-concept stage. The researchers have publicly and are exploring commercialization opportunities as well as further testing. For example, currently all these tests have been done in the lab. The next step is to work with patients to test this system at home. The researchers also want to see how the system fares in more resource-limited areas and countries.

“Almost every smartphone from the past decade has a vibration motor and a camera. This means that almost everyone who has a phone can use this. All you need is a simple plastic attachment, no additional electronics of any kind,” Gollakota said. “This is the best of all worlds 鈥� it’s basically the holy grail of PT/INR testing. It makes it frugal and accessible to millions of people, even where resources are very limited.”

Additional co-authors on this paper are Joanne Estergreen, clinical laboratory supervisor in the UW School of Medicine’s laboratory medicine and pathology department, and , professor of laboratory medicine and pathology in the UW School of Medicine. This research was funded by a Moore Foundation fellowship.

For more information, contact bloodclot@cs.washington.edu.

Twenty scientists and engineers at the 天美影视传媒 are among the 38 new members elected to the Washington State Academy of Sciences for 2021, according to a July 15 . New members were chosen for 鈥渢heir outstanding record of scientific and technical achievement, and their willingness to work on behalf of the Academy to bring the best available science to bear on issues within the state of Washington.鈥�

Current academy members selected 29 of the new members. An additional nine were elected by virtue of joining one of the National Academies.

New UW members who were elected by current academy members are:

• , professor and Port of Tacoma Chair in Environmental Science at UW Tacoma, director of the and science director of the , 鈥渇or foundational work on the environmental fate, behavior and toxicity of PCBs.鈥�
• , professor of psychology, 鈥渇or contributions in research on racial and gender inequality that has influenced practices in education, government, and business鈥� and 鈥渇or shifting the explanation for inequality away from individual deficiencies and examining how societal stereotypes and structures cause inequalities.鈥�
• , professor of chemistry and member faculty at the , 鈥渇or leadership in the innovative synthesis and chemical modification of nanoscale materials for application in light emission and catalysis.鈥�
• , professor of global health and of environmental and occupational health sciences, and founding director of the , 鈥渇or work on the health impacts of climate change, on climate impact forecasting, on adaptation to climate change and on climate policy to protect health.鈥�
• , professor of environmental and forest sciences and dean emeritus of the College of the Environment, 鈥渇or foundational studies of regional paleoenvironmental history and sustained excellence in academic leadership to catalyze and sustain transformative research and educational initiatives.鈥� Graumlich is also president-elect of the American Geophysical Union.
• Dr. , the Joseph W. Eschbach Endowed Chair in Kidney Research and co-director of the , 鈥渇or pioneering contributions and outstanding achievements in the development of the novel wearable artificial kidney, as well as numerous investigator-initiated clinical trials and multi-center collaborative studies.鈥�
• , professor of environmental chemistry and chair of the Physical Sciences Division at UW Bothell, 鈥渇or leadership in monitoring and understanding the global transport of atmospheric pollutants from energy production, wildfire, and other sources, as well as science communication and service that has informed citizens and enhanced public policy.鈥�
• , professor and chair of psychology, 鈥渇or contributions demonstrating how psychological science can inform long-standing issues about racial and gender discrimination鈥� and 鈥渇or research that has deep and penetrating implications for the law and societal efforts to remedy social inequities with evidence-based programs and actions.鈥�
• , the Leon C. Johnson Professor of Chemistry, member faculty at the and chair of the Department of Chemistry, 鈥渇or developing new spectroscopy tools for measuring energy flow in molecules and materials with high spatial and temporal resolution.鈥�
• , professor of astronomy, 鈥渇or founding the and leading the decades-long development of the interdisciplinary modeling framework and community needed to establish the science of exoplanet astrobiology鈥� and 鈥渇or training the next generation of interdisciplinary scientists who will search for life beyond Earth.鈥�
• , professor and chair of aeronautics and astronautics, 鈥渇or leadership and significant advances in nonlinear methods for integrated sensing and control in engineered, bioinspired and biological flight systems鈥� and 鈥渇or leadership in cross-disciplinary aerospace workforce development.鈥�
• , associate professor of chemistry and member faculty with the Molecular Engineering and Sciences Institute, 鈥渇or exceptional contributions to the development of synthetic polymers and nanomaterials for self-assembly and advanced manufacturing with application in sustainability, medicine and microelectronics.鈥�
• Dr. , Associate Dean of Medical Technology Innovation in the College of Engineering and the School of Medicine, the Graham and Brenda Siddall Endowed Chair in Cornea Research, and medical director of the UW Eye Institute, 鈥渇or developing and providing first class clinical treatment of severe corneal blindness to hundreds of people, for establishing the world premier artificial cornea program in Washington, and for leading collaborative research to translate innovative engineering technologies into creative clinical solution.鈥�
• Dr. , professor of medicine and director of the , 鈥渇or global recognition as an authority on drug and vaccine development for viral and parasitic diseases through work as an infectious disease physician and immunologist.鈥�
• Dr. , professor of pediatrics and of anesthesiology and pain medicine, and director of the , 鈥渇or outstanding leadership in pediatric anesthesiology and in the care of children with traumatic brain injury鈥� and 鈥渇or internationally recognized expertise in traumatic brain injury and direction of the Harborview Injury Prevention and Research Center for the last decade as an exceptional mentor and visionary leader.鈥�

UW members who will join the Washington State Academy of Sciences by virtue of their election to one of the National Academies are:

• , professor of biostatistics, 鈥渇or the development of novel statistical models for longitudinal data to better diagnose disease, track its trajectory, and predict its outcomes鈥� and 鈥渇or revolutionizing how dynamic predictors are judged by their discrimination and calibration and has significantly advanced methods for randomized controlled trials.鈥� Heagerty was elected to the National Academy of Medicine in 2021.
• , the Bill and Melinda Gates Chair in Computer Science and Engineering, 鈥渇or foundational contributions to the mathematics of computer systems and of the internet, as well as to the design and probabilistic analysis of algorithms, especially on-line algorithms, and algorithmic mechanism design and game theory.鈥� Karlin was elected to the National Academy of Sciences in 2021.
• , professor emeritus of applied mathematics and data science fellow at the , 鈥渇or inventing key algorithms for hyperbolic conservation laws and transforming them into powerful numerical technologies鈥� and 鈥渇or creating the Clawpack package, which underpins a wide range of application codes in everyday use, such as for hazard assessment due to tsunamis and other geophysical phenomena.鈥� LeVeque was elected to the National Academy of Sciences in 2021.
• , the Benjamin D. Hall Endowed Chair in Basic Life Sciences and an investigator with the Howard Hughes Medical Institute, 鈥渇or advancing our physical understanding of cell motility and growth in animals and bacteria鈥� and 鈥渇or discovering how the pathogen Listeria uses actin polymerization to move inside human cells, how crawling animal cells coordinate actomyosin dynamics and the mechanical basis of size control and daughter cell separation in bacteria.鈥� Theriot was elected to the National Academy of Sciences in 2021.
• , professor and chair of biological structure, 鈥渇or elucidating cellular transformations through which neurons pattern their dendrites, and the interplay of activity-dependent and -independent mechanisms leading to assembly of stereotyped circuits鈥� and 鈥渇or revelations regarding the fundamental principles of neuronal development through the application of an elegant combination of in vivo imaging, physiology, ultrastructure and genetics to the vertebrate retina.鈥� Wong was elected to the National Academy of Sciences in 2021.

New members to the Washington State Academy of Sciences are scheduled to be inducted at a meeting in September.

The four new AAAS fellows among the UW faculty are:

, professor emeritus in the Paul G. Allen School of Computer Science & Engineering, is honored for contributions to artificial intelligence and machine learning. Domingos is particularly known for his introduction of Markov logic networks, which presented a simple yet efficient approach to unifying first-order logic and probabilistic reasoning to support inference learning. He also helped pioneer the field of adversarial learning, producing the first algorithm to automate the process of adversarial classification to enable data mining systems to adapt rapidly against evolving adversarial attacks. Domingos subsequently contributed the first unsupervised approach to semantic parsing, which enables machines to extract knowledge from text and speech, a process that underpins machine learning and natural language processing. In 2015, he published 鈥�,鈥� a book that examines how machine learning increasingly influences every aspect of people鈥檚 lives. Domingos joined the UW faculty in 1999 and remains active in research after attaining emeritus status earlier this year.

, professor in the Department of Physiology and Biophysics, is a pioneer in brain-machine interfaces. His earlier work was on the brain鈥檚 direction of arm and leg movements. Fetz later showed that the brain could volitionally control certain nerve cells, called cortical neurons, in various patterns. This became the foundation for research on the unexpected ability of neural activity to drive external devices. Fetz also conducted studies of interneurons in the spine, and demonstrated that they had many properties of cells in the cortex, including their preparation to carry out instructed movements. Fetz also developed dynamic network models to simulate neural interactions that target tracking and short-term memory. In an historical achievement, his lab designed and tested an implantable neurochip that can record activity of cortical cells and convert this in real-time to stimulate the cortex, spinal cord or muscles. The brain can learn to incorporate this artificial feedback loop into behaviors. The neurochip holds future promise for clinical applications, such as moving paralyzed muscles.

is a professor in the Department of Anesthesiology and Pain Medicine, as well as a professor in the Public Health Sciences Division at the Fred Hutchinson Cancer Research Center. Raftery studies the small molecules at work during metabolism in cells, animals and people. He has developed analytical and statistical methods to profile metabolites in complex biological samples. Metabolites are the end products of many biochemical functions in living systems. Raftery鈥檚 research is working to discover sensitive biomarkers indicating the presence of disease and its progression. He has applied his advances in metabolomics to detect very early stages of cancer, as well as in his research on diabetes and heart disease. He is a scientist at the UW Mitochondrial and Metabolism Center, which, among its goals, is investigating the roles of cell metabolism dysfunction in common diseases and is also seeking related diagnostic and therapeutic tools. Raftery also directs the interdisciplinary Northwest Metabolomics Research Center, which fosters collaborations among scientists from several institutions. The lab uses some of the latest technologies and capabilities to improve the metabolic understanding of a variety of serious disorders.

, a professor in the Paul G. Allen School of Computer Science & Engineering, was honored for his contributions to artificial intelligence spanning automated planning, software agents, crowdsourcing and internet information extraction, as well as his efforts to commercialize AI technologies. Weld leads the UW鈥檚 , where he focuses on advancing explainable AI to allow people to better understand and control AI-powered tools, assistants and systems and combine human and machine intelligence to accomplish more together than alone. Weld has co-founded multiple startup companies, including Netbot, Inc., which produced the first online comparison shopping engine that was subsequently acquired by Excite, and AdRelevance, an early provider of tools for monitoring online advertising data acquired by Nielsen Netratings. A member of the UW faculty since 1988, Weld is a venture partner and member of the Technology Advisory Board of Madrona Venture Group and Allen Institute for Artificial Intelligence, where he also leads the focused on the development of AI-powered tools to help scientists extract useful knowledge from scholarly literature.

In addition, , a professor in the Vaccine and Infectious Disease Division of the Fred Hutchinson Cancer Research Center, was selected 鈥渇or distinguished contributions to the field of HIV prevention research, particularly for design and analysis of clinical trials of pre-exposure prophylaxis and treatment as prevention.鈥� Donnell is also a UW affiliate of global health and of health services.

Gone are the days when people use smart speakers 鈥� like Amazon Echo or Google Home 鈥� only as kitchen timers or dinner party music players. These devices have started helping people track their own health, and can even monitor for cardiac arrest.

Now researchers at the 天美影视传媒 have developed a new smart speaker skill that lets a device use white noise to both soothe sleeping babies and monitor their breathing and movement.

With this skill, called BreathJunior, the smart speaker plays white noise and records how the noise is reflected back to detect breathing motions of infants’ tiny chests. When the researchers tested BreathJunior with five babies in a local hospital’s neonatal intensive care unit, it detected respiratory rates that closely matched the rates detected by standard vital sign monitors. The team will present its findings October 22 at the conference in Los Cabos, Mexico.

“One of the biggest challenges new parents face is making sure their babies get enough sleep. They also want to monitor their children while they鈥檙e sleeping. With this in mind, we sought to develop a system that combines soothing white noise with the ability to unobtrusively measure an infant鈥檚 motion and breathing,” said co-author, an assistant professor of anesthesiology and pain medicine at the UW School of Medicine.

To make things easy for new parents, the team made a system that could run on a smart speaker that replicates the hardware in an Amazon Echo.

“Smart speakers are becoming more and more prevalent, and these devices already have the ability to play white noise,” said co-author, an associate professor in the UW’s Paul G. Allen School of Computer Science & Engineering and the director of the . “If we could use this white noise feature as a contactless way to monitor infants鈥� hand and leg movements, breathing and crying, then the smart speaker becomes a device that can do it all, which is really exciting.”

White noise is a, which makes a seemingly random soothing sound that can help cover up other noises that might wake a sleeping baby. To use white noise as a breathing monitor, the team needed to develop a method to detect tiny changes between the white noise a smart speaker plays and the white noise that gets reflected back from the infant鈥檚 body into the speaker’s array of microphones.

“We start out by transmitting a random white noise signal. But we are generating this random signal, so we know exactly what the randomness is,” said first author, a doctoral student in the Allen School. “That signal goes out and reflects off the baby. Then the smart speaker’s microphones get a random signal back. Because we know the original signal, we can cancel out any randomness from that and then we’re left with only information about the motion from the baby.”

Detecting breathing in babies has an extra wrinkle: The movement of their chests is so tiny that the smart speaker needs to know exactly where the babies are to be able to “see” them breathing.

“The breathing signal is so weak that we can’t just look for a change in the overall signal we get back,” Wang said. “We needed a way to scan the room and pinpoint where the baby is to maximize changes in the white noise signal. Our algorithm takes advantage of the fact that smart speakers have an array of microphones that can be used to focus in the direction of the infant鈥檚 chest. It starts listening for changes in a bunch of potential directions, and then continues the search toward the direction that gives the clearest signal.”

BreathJunior tracks both small motions 鈥� such as the chest movement involved in breathing 鈥� and large motions 鈥� such as babies moving around in their cribs. It can also pick up the sound of a baby crying.

The team created a prototype smart speaker to test BreathJunior on an infant simulator. The researchers could set the simulator to breathe at specific rates, which allowed them to test how well BreathJunior detected a variety of respiratory rates 鈥� from a slow 20 breaths per minute to聽60 breaths per minute. The infant simulator also allowed the team to test if BreathJunior could detect abnormal breathing patterns, such as apnea, that are common in babies who are born early and may not have developed respiratory centers in their brains. The system performed well for both tests.

Then the team tested how well their prototype tracked real babies’ breathing in the neonatal intensive care unit or NICU. These babies are connected to wired, hospital-grade respiratory monitors, so the team could compare their readouts to BreathJunior’s. The system was able to accurately identify respiratory rates up to 65 breaths per minute.

“Infants in the NICU are more likely to have either quite high or very slow breathing rates, which is why the NICU monitors their breathing so closely,” Sunshine said. “BreathJunior holds potential for parents who want to use white noise to help their child sleep and who also want a way to monitor their child鈥檚 breathing and motion. It also has appeal as a tool for monitoring breathing in the subset of infants in whom home respiratory monitoring is clinically indicated, as well as in hospital environments where doctors want to use unwired respiratory monitoring.

“However, it is very important to note that the American Academy of Pediatrics recommends not using a monitor that markets itself as reducing the risk of sudden infant death syndrome, and this research and the team makes no such claim.”

While BreathJunior currently uses white noise to track breathing and motion, the researchers would like to expand its capabilities so that it could also use other soothing sounds like lullabies.

The team plans to commercialize this technology through a UW spinout,.

“In just a few years, we have come a long way from monitoring large motions in adults to extracting the tiny motion of a newborn infant’s breathing,” Gollakota said. “This has been possible because of algorithmic innovations as well as advances in smart speaker hardware. Looking ahead, one can envision transforming a smart speaker into a that can contactlessly monitor a variety of vital signs beyond just breathing.”

This research was funded by the National Science Foundation.

###

For more information, contact whitenoise@cs.washington.edu.

Grant numbers: CNS 1812559, 1914873

Almost 500,000 Americans die each year from , when the heart suddenly stops beating.

People experiencing cardiac arrest will suddenly become unresponsive and either stop breathing or gasp for air, a sign known as agonal breathing. Immediate CPR can double or triple someone’s chance of survival, but that requires a bystander to be present.

Cardiac arrests often occur outside of the hospital and in the privacy of someone’s home. suggests that one of the most common locations for an out-of-hospital cardiac arrest is in a patient’s bedroom, where no one is likely around or awake to respond and provide care.

Researchers at the 天美影视传媒 have developed a new tool to monitor people for cardiac arrest while they’re asleep without touching them. A new skill for a smart speaker 鈥� like Google Home and Amazon Alexa 鈥� or smartphone lets the device detect the gasping sound of agonal breathing and call for help. On average, the proof-of-concept tool, which was developed using real agonal breathing instances captured from 911 calls, detected agonal breathing events 97% of the time from up to 20 feet (or 6 meters) away. The findings June 19 in the Nature journal .

“A lot of people have smart speakers in their homes, and these devices have amazing capabilities that we can take advantage of,” said co-corresponding author , an associate professor in the UW’s Paul G. Allen School of Computer Science & Engineering. “We envision a contactless system that works by continuously and passively monitoring the bedroom for an agonal breathing event, and alerts anyone nearby to come provide CPR. And then if there’s no response, the device can automatically call 911.”

Agonal breathing is present for about 50% of people who experience cardiac arrests, according to 911 call data, and patients who take agonal breaths often have a better chance of surviving.

“This kind of breathing happens when a patient experiences really low oxygen levels,” said co-corresponding author , an assistant professor of anesthesiology and pain medicine at the UW School of Medicine. “It’s sort of a guttural gasping noise, and its uniqueness makes it a good audio biomarker to use to identify if someone is experiencing a cardiac arrest.”

The researchers gathered sounds of agonal breathing from real 911 calls to Seattle’s Emergency Medical Services. Because cardiac arrest patients are often unconscious, bystanders recorded the agonal breathing sounds by putting their phones up to the patient’s mouth so that the dispatcher could determine whether the patient needed immediate CPR. The team collected 162 calls between 2009 and 2017 and extracted 2.5 seconds of audio at the start of each agonal breath to come up with a total of 236 clips. The team captured the recordings on different smart devices 鈥� an Amazon Alexa, an iPhone 5s and a Samsung Galaxy S4 鈥� and used various machine learning techniques to boost the dataset to 7,316 positive clips.

“We played these examples at different distances to simulate what it would sound like if it the patient was at different places in the bedroom,” said first author , a doctoral student in the Allen School. “We also added different interfering sounds such as sounds of cats and dogs, cars honking, air conditioning, things that you might normally hear in a home.”

For the negative dataset, the team used 83 hours of audio data collected during sleep studies, yielding 7,305 sound samples. These clips contained typical sounds that people make in their sleep, such as snoring or obstructive sleep apnea.

From these datasets, the team used machine learning to create a tool that could detect agonal breathing 97% of the time when the smart device was placed up to 6 meters away from a speaker generating the sounds.

Next the team tested the algorithm to make sure that it wouldn’t accidentally classify a different type of breathing, like snoring, as agonal breathing.

“We don’t want to alert either emergency services or loved ones unnecessarily, so it’s important that we reduce our false positive rate,” Chan said.

For the sleep lab data, the algorithm incorrectly categorized a breathing sound as agonal breathing 0.14% of the time. The false positive rate was about 0.22% for separate audio clips, in which volunteers had recorded themselves while sleeping in their own homes. But when the team had the tool classify something as agonal breathing only when it detected two distinct events at least 10 seconds apart, the false positive rate fell to 0% for both tests.

The team envisions this algorithm could function like an app, or a skill for Alexa that runs passively on a smart speaker or smartphone while people sleep.

“This could run locally on the processors contained in the Alexa. It’s running in real time, so you don’t need to store anything or send anything to the cloud,” Gollakota said.

“Right now, this is a good proof of concept using the 911 calls in the Seattle metropolitan area,” he said. “But we need to get access to more 911 calls related to cardiac arrest so that we can improve the accuracy of the algorithm further and ensure that it generalizes across a larger population.”

The researchers plan to commercialize this technology through a UW spinout, .

“Cardiac arrests are a very common way for people to die, and right now many of them can go unwitnessed,” Sunshine said. “Part of what makes this technology so compelling is that it could help us catch more patients in time for them to be treated.”

, a professor of general internal medicine at the UW School of Medicine and the medical director of was also a co-author on this paper. This research was funded by the National Science Foundation.

###

For more information, contact the research team at cardiacalert@cs.washington.edu.

At least 115 people die every day in the U.S. after overdosing on opioids, .

And in 2016, illegal injectable opioids involved in overdose-related deaths. This spike has led to a national public health crisis and epidemic.

During an overdose, a person breathes slower or stops breathing altogether. These symptoms are reversible with the drug naloxone if caught in time.

But people who use opioids by themselves have no way of asking for help in the event of an overdose.

Researchers at the 天美影视传媒 have developed a cellphone app, called Second Chance, that uses sonar to monitor someone’s breathing rate and sense when an opioid overdose has occurred. The app accurately detects overdose-related symptoms about 90 percent of the time and can track someone’s breathing from up to 3 feet away. The team Jan. 9 in Science Translational Medicine.

“The idea is that people can use the app during opioid use so that if they overdose, the phone can potentially connect them to a friend or emergency services to provide naloxone,” said co-corresponding author , an associate professor in the UW’s Paul G. Allen School of Computer Science & Engineering. “Here we show that we have created an algorithm for a smartphone that is capable of detecting overdoses by monitoring how someone’s breathing changes before and after opioid use.”

The Second Chance app sends inaudible sound waves from the phone to people’s chests and then monitors the way the sound waves return to the phone to look for specific breathing patterns.

“We’re looking for two main precursors to opioid overdose: when a person stops breathing, or when a person’s breathing rate is seven breaths per minute or lower,” said co-corresponding author , an assistant professor of anesthesiology and pain medicine at the UW School of Medicine. “Less than eight breaths per minute is a common cutoff point in a hospital that would trigger people to go to the bedside and make sure a patient is OK.”

In addition to watching breathing, Second Chance also monitors how people move.

“People aren’t always perfectly still while they’re injecting drugs, so we want to still be able to track their breathing as they’re moving around,” said lead author , a doctoral student in the Allen School. “We can also look for characteristic motions during opioid overdose, like if someone’s head slumps or nods off.”

To be able to use real-world data to design and test the algorithm behind the app, the researchers partnered with the in Vancouver, Canada. Insite is the first legal supervised consumption site in North America. As part of the study, participants at Insite wore monitors on their chests that also track breathing rates.

“We asked participants to prepare their drugs like they normally would, but then we monitored them for a minute pre-injection so the algorithm could get a baseline value for their breathing rate,” said Nandakumar. “After we got a baseline, we continued monitoring during the injection and then for five minutes afterward, because that’s the window when overdose symptoms occur.”

Of the 94 participants who tested the algorithm, 47 had a breathing rate of seven breaths per minute or slower, 49 stopped breathing for a significant period, and two people experienced an overdose event that required oxygen, ventilation and/or naloxone treatment. On average, the algorithm correctly identified breathing problems that foreshadow overdose 90 percent of the time.

The researchers also wanted to make sure the algorithm could detect actual overdose events, because these occur infrequently at Insite. The researchers worked with anesthesiology teams at UW Medical Center to 鈥渟imulate鈥� overdoses in an operating room, allowing the app to monitor people and detect when they stop breathing.

“When patients undergo anesthesia, they experience much of the same physiology that people experience when they’re having an overdose,” Sunshine said. “Nothing happens when people experience this event in the operating room because they’re receiving oxygen and they are under the care of an anesthesiology team. But this is a unique environment to capture difficult-to-reproduce data to help further refine the algorithms for what it looks like when someone has an acute overdose.”

For the simulation, the team recruited healthy participants undergoing previously scheduled elective surgeries. After providing informed consent, the patients then received standard anesthetic medications that led to 30 seconds of slower or absent breathing, and these events were captured by the device. The algorithm correctly predicted 19 out of the 20 simulated overdoses. For the one case it was wrong, the patient’s breathing rate was just above the algorithm’s threshold.

Right now, Second Chance is only monitoring the people who use it. The team would eventually like the app to interact with them.

“When the app detects decreased or absent breathing, we’d like it to send an alarm asking the person to interact with it,” Gollakota said. “Then if the person fails to interact with it, that’s when we say: ‘OK this is a stage where we need to alert someone,’ and the phone can contact someone with naloxone.”

The researchers are applying for FDA approval and have plans to commercialize this technology through a UW spinout called . While this app could be used for all forms of opioid use, the team cautions that right now they have only tested it on illegal injectable opioid use because deaths from those overdoses are the most common.

“We’re experiencing an unprecedented epidemic of deaths from opioid use, and it’s unfortunate because these overdoses are completely reversible phenomena if they’re detected in time,” Sunshine said. “The goal of this project is to try to connect people who are often experiencing overdoses alone to known therapies that can save their lives. We hope that by keeping people safer, they can eventually access long-term treatment.”

This research was funded by the UW Alcohol and Drug Abuse Institute and the National Science Foundation.

###

For more information, contact the research team at secondchance@cs.washington.edu.

Now researchers at the 天美影视传媒 have developed a new machine-learning system, called Prescience, which uses input from patient charts and standard operating room sensors to predict the likelihood that a patient will develop hypoxemia 鈥� a condition when blood oxygen levels dip slightly below normal. Hypoxemia can lead to serious consequences, such as infections and abnormal heart behavior.

Prescience also provides real-world explanations behind its predictions. With this information, anesthesiologists can better understand why a patient is at risk for hypoxemia and prevent it before it happens. The team, which Oct. 10 in Nature Biomedical Engineering, estimates that Prescience could improve the ability of anesthesiologists to anticipate and prevent 2.4 million more hypoxemia cases in the United States every year.

“Modern machine-learning methods often just spit out a prediction result. They don’t explain to you what patient features contributed to that prediction,” said , an associate professor in the UW’s Paul G. Allen School of Computer Science & Engineering and senior author of the paper. “Our new method opens this black box and actually enables us to understand why two different patients might develop hypoxemia. That’s the power.”

Lee and , a doctoral student in the Allen School, started the project by meeting with collaborators from UW Medicine to find out what they needed in the operating room.

“One of the things the anesthesiologists said was: ‘We are not really satisfied with just a prediction. We want to know why,'” Lee said. “So that got us thinking.”

Lee and Lundberg set out to create a machine-learning system that could both make predictions and explain them. First, they acquired a dataset of 50,000 real surgeries from 天美影视传媒 Medical Center and Harborview Medical Center in Seattle. These data include patient intake information like age and weight as well as real-time, minute-by-minute information 鈥� heart rate, blood oxygen levels and more 鈥� throughout the surgeries. The scientists used all of these data to teach Prescience to make predictions.

The team wanted Prescience to solve two different kinds of problems. Prescience needed to look at pre-surgery information and predict whether any given patient would have hypoxemia while under anesthesia. Prescience also had to predict hypoxemia at any point throughout surgery by looking at real-time information. Finally, Lee and Lundberg to train Prescience to generate understandable explanations behind its predictions.

For the pre-surgery data, Prescience found that body mass index was one important feature that contributed to a prediction that a patient would experience hypoxemia during surgery. But during surgery, the blood oxygen levels themselves contributed the most to a prediction.

With this information in mind, it was time to put Prescience to the test.

Lee and Lundberg created a web interface that ran anesthesiologists through pre-surgery and real-time cases from surgeries in the dataset that were not used to train Prescience. For the real-time test, the researchers specifically chose cases that would be hard to predict, such as when a patient’s blood oxygen level is stable for 10 minutes and then drops.

“We wanted to know if this was going to be informative to anesthesiologists,” said Lundberg, who is the first author on the paper. “So for some of their cases, they got a bar of additional information from Prescience.”

Prescience improved the doctors’ ability to correctly predict a patient’s hypoxemia risk by 16 percent before a surgery and by 12 percent in real time during a surgery. Overall, with the help of Prescience, the anesthesiologists were able to correctly distinguish between the two scenarios nearly 80 percent of the time both before and during surgery.

“This research will allow us to better anticipate complications and target our treatment to each patient,” said co-author Dr. Monica Vavilala, professor of anesthesiology and pain medicine at the UW School of Medicine and director of the Harborview Injury Prevention & Research Center. “If we know there’s one aspect that’s causing the problem, then we can approach that first and more quickly. This could really change the way we practice, so this is a really big deal.”

Prescience isn’t quite ready to be in operating rooms yet. Lee and Lundberg plan to continue working with anesthesiologists to improve Prescience and give it an interface that’s both intuitive and useful. In addition, the team hopes that later versions of Prescience will be able to predict other harmful conditions, such as low blood pressure, and recommend treatment plans.

Regardless of Prescience’s future, one point is clear: This technology is meant , Lundberg said.

“Prescience doesn’t treat anyone,” he said. “Instead it tells you why it’s concerned, which then enables the doctor to make better treatment decisions.”

Co-author Dr. Bala Nair, research associate professor of anesthesiology and pain medicine at the UW School of Medicine, helped Lee and Lundberg acquire the dataset. Another co-author, Dr. Jerry Kim, assistant professor of anesthesiology and pain medicine at Seattle Children’s Hospital, proposed the initial project with Lee. Other co-authors include Shu-Fang Newman, UW Medicine Anesthesiology & Pain Medicine; Dr. Mayumi Horibe, Veterans Affairs Puget Sound Health Care System; and Dr. Michael Eisses, Dr. Trevor Adams, Dr. David Liston and Dr. Daniel Low, Seattle Children’s Hospital. This research was funded by the National Science Foundation (DBI-1355899), NSF Graduate Research Fellowship (DGE-1256082), and a UW eScience/ITHS seed grant, “Machine Learning in Operating Rooms.”

###

For more information, contact Lee at suinlee@cs.washington.edu.