Author: lisaweiner

Categories
AI Covid-19

AI detects COVID in chest x rays

DeepCOVID-XR is a Northwestern University developed algorithm that automatically detects the signs of COVID-19 on a basic X-ray of the lungs. The system is able to detect COVID-19 in X-rays 10 times faster than thoracic radiologists and 1% to 6% more accurately.

The developers said the AI could be used to rapidly screen patients at hospital admission, and trigger protocols to help protect healthcare workers.

Accoring to Professor Aggelos Katsaggelos, the alorithm will not replace testing, but will enable the use of cheap, routine, safe x-rays to determine if a patient needs to be isolated.

In 17,000 X-ray images, the algorithm identified lungs which appeared patchy and hazy, as air sacs became inflamed and filled with fluid instead of air, which is common in COVID-19.

When put up against five experienced, fellowship-trained radiologists, DeepCOVID-XR was able to process a set of 300 test X-rays in about 18 minutes, compared to about two and a half to three and a half hours. The AI also delivered an accuracy rate of 82%, about on par with the group’s range of 76% to 81%.

“Radiologists are expensive and not always available,” Katsaggelos said. “X-rays are inexpensive and already a common element of routine care. This could potentially save money and time—especially because timing is so critical when working with COVID-19.”

Categories
BCI

Algorithm isolates specific brain signals, provides feedback

The US Army and USC Prof Maryam Shanechi have developed an algorithm that can determine which specific behaviors—like walking and breathing—belong to specific brain signals.

Segmenting brain signals has been notoriously difficult, as all signals associated with tasks mix together. Shanechi and her team used the algorithm to separate behaviorally relevant brain signals from behaviorally irrelevant brain signals.

Army research office project manager Hamid Krim said that if the algorithm detects behavior indicating a soldier is stressed or overloaded, then a machine could alert that soldier before they recognize their own fatigue. This may enable the development of technology that can interpret signals from the brain and then send signals back, to automatically correct behavior, or to enable soldiers to communicate with out speaking.

In the civilian world, this could allow locked-in patients to communicate, and those with various disabilities and brain diseases to improve function.

Categories
Sensors

“Ambient intelligence” monitoring to prevent medical errors, send alerts

Stanford’s Fei Fei Li, Arnold Milstein and albert Haque have developed AI and sensor based “ambient intelligence” protocols to prevent medical errors and improve outcomes.

Applications include alerting clinicians and visitors when they fail to sanitize their hands before entering a hospital room; monitoring the elderly for behavioral clues of impending health crises; prompting caregivers, remotely clinicians and patients to make life-saving interventions.

Milstein believes that “we are in a foot race with the complexity of bedside care.” He noted that clinicians in a hospital’s neonatal intensive care unit took 600 bedside actions, per patient, per day and hat ambient intelligence is necessary as “perfect execution of this volume of complex actions is well beyond what is reasonable to expect of even the most conscientious clinical teams.”

The alert systems are being tested to see if they can reduce the number of ICU patients who get nosocomial infections.

In one experiment, a tablet near the door shows a solid green screen that transitions to red when a hygiene failure occurs.

A thermal sensor above an ICU bed would enable the detection of twitching or writhing beneath the sheets, and alert clinical team members.

Constant monitoring by ambient intelligence systems at home could detect clues of serious illness or potential accidents, and alert caregivers to make timely interventions, such as when frail seniors start moving more slowly or stop eating regularly.

Categories
Covid-19 Sensors

Electrostatic and electrochemical sensors rapidly detect airborne viruses

Jang Jae-sung and Ulsan colleagues have developed a method to quickly and accurately detect airborne viruses, to inform public health and quarantine efforts.

Electrostatic force captures and condenses viruses in the air, and a paper electrochemical sensor checks samples for antigens and virus antibodies. Liquid particles as small as 1 micrometer have been successfully collected.

Current methods of airborne sample collection typically use vacuums that can damage the samples, and cannot collect very small particles.

A recent study showed that tests on the type-A H1N1 flu virus showed good results. Jang believes that since the coronavirus is similar in structure and size, the technology should be applicable to COVID-19, which he is researching.

Categories
BCI Brain Sensors

Polymer improves medical implants, could enable brain-computer interface

David Martin and University of Delaware colleagues have developed a bio-synthetic coating for electronic components that could avoid the scarring (and signal disruption) caused by traditional microelectric materials. The PEDOT polymer improved the performance of medical implants by reducing their opposition to an electric current.

Pedot film was used with an antibody to stimulate blood vessel growth after injury, and could be used to detect early stages of tumor growth. The polymers could also help sense or treat brain or nervous system disorders, while versions could theoretically attach peptides, antibodies and DNA.

The team believes that materials, when inserted, could connect brains to a computer.

Categories
Covid-19 Sensors Wearables

Presymptomatic COVID detection with wearables

Stanford’s Michael Snyder and colleagues have used smartwatch data to detect early, presymptomatic COVID-19 in 31 individuals out of a cohort of 5,000.

They demonstrated that COVID-19 infections are associated with alterations in heart rate, steps and sleep in 80% of cases. Physiological alterations were detected prior to, or at, symptom onset in 85% of the positive cases, in some cases nine or more days before symptoms.

A method to detect onset of COVID-19 infection in real-time was developed, which detected 67% of infection cases at or before symptom onset.

The study intends to provide a roadmap to a rapid and universal diagnostic method for the large-scale detection of respiratory viral infections in advance of symptoms.

—————————————————————————-

Click to view Professor Snyder’s talk at the 2019 ApplySci conference at Stanford.

Categories
Pharmaceuticals Sensors

Sweat sensor monitors drug levels, informs dosage

Sam Emaminejad, Shuyu Lin, Carlos Milla, Ronald Davis and UCLA and Stanford colleagues have developed a watch which monitors drug levels inside the body by analyzing a wearer’s sweat. The goal is individually tailored drug dosages.

Dosages are currently prescribed based on statistical effectiveness averages driven by weight and age. However, constantly changing body chemistry and an ones genetic makeup affect how fast drugs are absorbed, take effect and are eliminated from the body.

Current efforts to personalize the drug dosage rely heavily on repeated blood draws at the hospital. The samples are then sent out to be analyzed in central labs. These solutions are inconvenient, time-consuming, invasive and expensive. That is why they are only performed on a small subset of patients and on rare occasions.

Emaminjegad wanted to “create a wearable technology that can track the profile of medication inside the body continuously and non-invasively” and he seems to have succeeded, using tiny droplets of sweat.

In a recent study, he tracked the effect of acetaminophen, over a period of hours, by stimulating sweat glands with an electric current, and accurately detecting the drug’s unique electrochemical signal, against the backdrop of signals from many other molecules that may be circulating in the body and in higher concentrations than the drug.

The technology can personalize pharmacotherapy approaches, and, according to Emaminejad also be used to monitor medication adherence and drug abuse.

Categories
Brain Sensors

Sensor platform detects dopamine in sweat; could be used for future treatment

Penn State’s Aida Ebrahimi and Maurico Terrones, RPI’s Humberto Terrones, and colleagues, have developed a highly sensitive, non-invasive wearable Dopamine sensor platform. The goal is the use of the technology to develop wearable sensors able to track and eventually treat conditions caused by too much (ie schizophrenia) or too little (ie Parkinson’s, depression) dopamine.

The low cost, flexible detector was achieved by doping a Molybdenum disulfide with Manganes, embedded in two-dimensional transition metal dichalcogenide.

Current dopamine monitoring methods are invasive and require specialized lab equipment. The researchers described the new method as “very simple and scalable.”

Categories
Covid-19 Sensors

First nutrition monitoring wearable tracks vitamin C; could be useful in COVID treatment

UCSD’s Joe Wang has again disrupted chemical sensing, by creating a wearable vitamin C sensor, which is a departure from now common vital sign and activity sensing wearables. This is the first time a wearable has been used to track nutritional intake, a key component of general health and disease prevention.

Vitamin C cannot be synthesized by the human body and must be obtained through food or supplements. It supports immune health, collagen production, wound healing and may be useful in treating cancer, heart disease, and COVID-19. High doses have been linked to reduced mortality in COVID patients with Sepsis and/or ARDS in studies.

The adhesive patch, applied to skin, stimulates sweating, and quickly detects vitamin c levels using flexible electrodes containing the enzyme ascorbate oxidase. When vitamin C is present, the enzyme converts it to dehydroascrobic acid. The resulting consumption of oxygen generates a current that is measured by the device. 

Click to view Prof Wang’s (brilliant) talk at the recent ApplySci conference at Harvard Medical School.

Categories
Covid-19 Sensors

Single sensor could simultaneously detect, differentiate between flu, coronavirus

University of Texas professor Deji Akinwande is developing a graphene sensor that can tell the difference between flu and coronavirus, and test for both simultaneously. The goal is to save time, medical resources, and cost, and speed appropriate treatment, as a second COVID wave could correspond with the next flu season.

The sensor is the size of a micro USB drive and is infused with antibodies of both COVID-19 and influenza. One part of the device is sensitive only to the flu, while another part will react only to the coronavirus.

The researchers will use inactive samples of COVID-19 and influenza for initial testing, and will measure how he sensor connects with the coronavirus’s spike proteins, which help it enter human cells by binding with them.

The work builds on the team‘s previous iron deficiency detecting graphene sensor work. “It became clear that just by changing the antibody, we could pivot the platform to focus on the coronavirus,” Akinwande said.

Categories
3-D Printing Covid-19 Sensors

Sensors 3D printed directly on lungs, heart could be used with surgical robots to diagnose, monitor disease

Michael McAlpine and University of Minnesota colleagues used 3D printing and motion capture technology to print electronic sensors directly on organs that are expanding and contracting, such as the heart and lungs. This could be used to diagnose and monitor the lungs of patients with COVID-19.

This builds on the team’s technique which enabled the printing of electronics directly on the skin of a hand that moved left to right or rotated.

They used a balloon-like surface and a specialized 3D printer, with motion capture tracking markers to help the 3D printer adapt its printing path to the expansion and contraction movements on the surface. An animal lung in the lab was artificially inflated and a soft hydrogel-based sensor was printed directly on the surface.

According to McAlpine, “the broader idea behind this research, is that this is a big step forward to the goal of combining 3D printing technology with surgical robots. In the future, 3D printing will not be just about printing but instead be part of a larger autonomous robotic system. This could be important for diseases like COVID-19 where health care providers are at risk when treating patients.”

Categories
Covid-19 Sensors

Organic electronic patch + algorithm continuously monitor multiple COVID symptoms

Northwestern and University of Illinois researchers have partnered to combine a COVID symptom-detecting wearable with a method to organize and analyze the massive data sets required to accurately show disease progression. The technology will be used in hospitals and nursing homes, to monitor both patients and healthcare workers, to identify contagion early in an effort to reduce the risk of spread.

Building on his stroke-monitoring wearable, John Rogers at Northwestern uses a patch that sits at the base of the throat and continuously monitors cough, heart and respiratory sounds. The University of Illionois algorithm, developed by Naresh Shanbhag, will allow these parameters to be tracked quantitatively.

The goal is to extract detailed, and sometimes subtle, parameters from the raw data, which provide insights into disease progression. This includes analyzing cough intensity, whether it is wet or dry, and whether a patient swallows afterward,

In addition to hospital patients, the team was able to monitor a nurse before, during, and after she had contracted the virus. The data collected, in combination with the nurse’s detailed notes throughout her illness, allowed the team to pick up heart rate spikes and changes in coughing activity that would have gone undetected in standard non-ICU hospital care.

Data is transmitted from the patch to a phone or a tablet, and then to the cloud, where it is processed. The group is attempting to move the processing directly to the patch and device to save energy consumption, and improve security and privacy.

According to Shanbhag: “We will acquire data for both ill and healthy patients, and learn the characteristics of the data by developing COVID-19 specific, low-complexity machine learning algorithms. We’ll then use the learned models for predicting whether a patient is ill or not and how the disease will progress over time for new patients or individuals.”