fbpx
Connect with us

The Conversation

Tracking vampire worms with machine learning − using AI to diagnose schistosomiasis before the parasites causing it hatch in your blood

Published

on

theconversation.com – Trirupa Chakraborty, Ph.D. Candidate in Integrative Biology, of Pittsburgh – 2024-10-21 07:44:00

Trirupa Chakraborty, University of Pittsburgh; Aniruddh Sarkar, Georgia Institute of Technology, and Jishnu Das, University of Pittsburgh

Blood samples of patients infected with a parasitic worm that causes schistosomiasis contain hidden information that marks different stages of the disease. In our recently published research, our team used machine learning to uncover that hidden information and improve early detection and diagnosis of infection.

The parasite that causes schistosomiasis completes its cycle in two – first in snails and then in mammals such as people, dogs and mice. Freshwater worm eggs enter human hosts through the skin and circulate throughout the body, damaging multiple organs, the liver, intestine, bladder and urethra. When these larvae reach blood vessels connecting the intestines to the liver, they mature into adult worms. They then release eggs that are excreted when the infected person defecates, continuing the transmission cycle.

Since diagnosis currently relies on detecting eggs in feces, usually miss the early stages of infection. By the time eggs are detected, patients have already reached an advanced stage of the disease. Because diagnosis rates are poor, public typically mass-administer the drug praziquantel to populations in affected regions. However, praziquantel cannot clear juvenile worms in early stages of infection, nor can it prevent reinfection.

Diagram of schistosomiasis infection cycle

Schistosomiasis isn’t usually diagnosed until the late stages of the disease.

DPDx/CDC

Our study provides a clear path forward to improving early detection and diagnosis by identifying the hidden information in blood that signals active, early stage infection.

Your body responds to a schistosomiasis infection by mounting an immune response involving several types of immune cells, as well as antibodies specifically targeting molecules secreted by or present on the worm and eggs. Our study introduces two ways to screen for certain characteristics of antibodies that signal early infection.

The first is an assay that captures a quantitative and qualitative profile of immune response, including various classes of antibodies and characteristics that dictate how they communicate with other immune cells. This us to identify specific facets of the immune response that distinguish uninfected patients from patients with early and late-stage disease.

Second, we developed a new machine learning approach that analyzes antibodies to identify latent characteristics of the immune response linked to disease stage and severity. We trained the model on immune profile data from infected and uninfected patients and tested the model on data that wasn’t used for training and data from a different geographical location. We identified not only biomarkers for the disease but also the potential mechanism that underlies infection.

Why it matters

Schistosomiasis is a neglected tropical disease that affects over 200 million people worldwide, causing 280,000 deaths annually. Early diagnosis can improve treatment effectiveness and prevent severe disease.

In addition, unlike many machine learning methods that are black boxes, our approach is also interpretable. This means it can provide insights into why and how the disease develops beyond simply identifying markers of disease, guiding future strategies for early diagnosis and treatment.

Microscopy image of large white oval-shaped structures enclosing magenta oval-shaped structures, surrounded by smaller cells

Clusters of Schistosoma haematobium eggs surrounded by immune cells in bladder tissue.

CDC/Dr. Edwin P. Ewing Jr.

What still isn’t known

The schistosomiasis infection signatures we identified remain stable across two geographical regions across two continents. Future research could explore how well these biomarkers apply to additional populations.

Further, our work identifies a potential mechanism behind disease progression. We found that a particular immune response against a specific protein on the surface of the worm signals an intermediate stage of infection. Understanding how the immune system responds to this understudied antigen could improve diagnosis and treatment.

What’s next

Besides improving our understanding of how the immune system responds to different stages of infection, our findings identify key antigens that could pave the way for designing cost-effective and efficient approaches to diagnosis and treatments. Our next steps will include actually deploying these strategies in the field for early detection and management of disease.

The Research Brief is a short take about interesting academic work.The Conversation

Trirupa Chakraborty, Ph.D. Candidate in Integrative Systems Biology, University of Pittsburgh; Aniruddh Sarkar, Assistant Professor of Biomedical Engineering, Georgia Institute of Technology, and Jishnu DasUniversity of Pittsburgh

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post Tracking vampire worms with machine learning − using AI to diagnose schistosomiasis before the parasites causing it hatch in your blood appeared first on .com

The Conversation

More kids than ever need special education, but burnout has caused a teacher shortage

Published

on

theconversation.com – Kimber Wilkerson, Professor of Special Education, University of Wisconsin- – 2024-10-21 07:44:00

Many special education teachers quit after less than five years on the job.

10’000 Hours/Digital Vision via Getty Images

Kimber Wilkerson, University of Wisconsin-Madison

A growing number of students in public schools – right now, about 15% of them – are eligible for special education services. These services include specially designed instruction for students with autism, learning or physical disabilities, or traumatic brain injuries. But going into the current school year, more than half of U.S. public schools anticipate being short-staffed in special education. Dr. Kimber Wilkerson, a professor of special education and department chair at the University of Wisconsin-Madison, explains why there’s a shortage and what needs to be done to close the gap.

Dr. Kimber Wilkerson discusses the special education teacher shortage.

has collaborated with SciLine to bring you highlights from the discussion, which have been edited for brevity and clarity.

Which students receive special education services?

Kimber Wilkerson: Students with a disability label receive special education services. They need these additional services and sometimes instruction in school so they can access the curriculum and thrive like their peers.

What is with staffing for special education?

Wilkerson: Since special education became a thing in the ’70s, there have always been challenges in filling all the special education positions.

In the past 10 years preceding the pandemic, those challenges started to increase. There were more open positions in special education at the beginning of each school year than in previous decades. In the 2023-24 school year, 42 states plus the District of Columbia reported teacher shortages in special education.

What is causing these shortages?

Wilkerson: One, there are fewer young people choosing teaching as a major in college and as a profession. And special education is affected by these lower rates more than other forms of education.

Also, there’s more attrition – people leaving their teaching job sooner than you might expect – not because they’re retiring, but because they are tired of the job.

They want to do something different. They want to go back to school. Sometimes it’s life circumstances, but the number of people leaving the job before retirement age has increased. And in our , Wisconsin, about 35% of all educators leave the field before they hit their fifth year.

That number is even higher for special educators. About half of special educators are out of the profession within five years.

Why do special education teachers the profession?

Wilkerson: There’s not a national study that speaks to that reason. There are some localized studies, and people report things like too much paperwork or too many administrative tasks associated with the job. Sometimes they report the students’ behavioral challenges. Sometimes it’s a feeling of isolation, or a lack of support from the school.

How are students with disabilities affected when their school does not have enough special educators?

Wilkerson: In a school that’s one special educator short, the other special educators have to take over that caseload. Instead of having 12 students on their caseload, maybe now they have 20. So, the amount of individual attention given to each student with a disability decreases.

Also, when teachers with experience leave the profession, they leave behind a less experienced group of teachers. This means the students are losing out on the benefit of those years of wisdom and experience.

What are some strategies to recruit and retain more special education teachers?

Wilkerson: There’s a range of strategies that different universities, states and school districts have taken, like residency programs.

In these programs, the person who is learning to be a teacher, and who is referred to as a teaching resident, works alongside a mentor teacher for an entire year in a school, and they get paid to do so. They’re not the teacher of record, but they’re learning and getting paid, and they’re in that school community.

Can you tell us about your recent study on supporting new special education teachers?

Wilkerson: One thing that made a big difference is when the teachers in our study, which is now under review, had access to a mentor and a group of their peers. We called this facilitated peer-to-peer group of teachers a “community of practice.” Every other week, on Zoom, we’d get these new special education teachers from different school districts together, along with experienced teachers. And they would do some sort of work on problems, bringing in the things that were challenging, and work on possible solutions as a group.

We also used Zoom to do one-on-one mentoring. And what people liked about it was that they could to someone who wasn’t right in their building and right in their district who they could be open and vulnerable with.

Sometimes, special educators can be isolated because they’re not necessarily a part of a grade-level team. They work with kids across a lot of classrooms. This gave them an to have their own kind of community, and that made a difference.

We also surveyed their level of burnout and how good they felt about the job they did. And then we surveyed special education teachers who weren’t participating in our community of practice.

At the end of the year, those people who had that mentoring and the community of practice felt less burnt out, and they also felt more effective in the area of classroom management. And that’s critical, because burnout is one of the primary reasons people leave the profession.

So if we can make people feel like they’re better equipped to handle this challenging position, then that’s one strategy to increase the number of people wanting to stay in their job year after year.

Watch the full interview to hear more.

SciLine is a service based at the American Association for the Advancement of Science, a nonprofit that helps journalists include scientific evidence and experts in their stories.The Conversation

Kimber Wilkerson, Professor of Special Education, University of Wisconsin-Madison

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post More kids than ever need special education, but burnout has caused a teacher shortage appeared first on theconversation.com

Continue Reading

The Conversation

Wild animals can experience trauma and adversity too − as ecologists, we came up with an index to track how it affects them

Published

on

theconversation.com – Xochitl Ortiz Ross, Ph.D. Candidate in Ecology & Evolutionary Biology, of California, Los Angeles – 2024-10-21 07:44:00

Marmots were the perfect test species for a wildlife adversity index.

Xochitl Ortiz Ross

Xochitl Ortiz Ross, University of California, Los Angeles

Psychologists know that childhood trauma, or the experience of harmful or adverse , can have lasting repercussions on the and well-being of people well into adulthood. But while the consequences of early adversity have been well researched in humans, people aren’t the only ones who can experience adversity.

If you have a rescue dog, you probably have witnessed how the abuse or neglect it may have experienced earlier in now influence its behavior – these pets tend to be more skittish or reactive. Wild animals also experience adversity. Although their negative experiences are easy to dismiss as part of life in the wild, they still have lifelong repercussions – just like traumatic events in people and pets.

As behavioral ecologists, we are interested in how adverse experiences early in life can affect animals’ behavior, the kinds of decisions they make and the way they interact with the world around them. In other words, we want to see how these experience affect the way they behave and survive in the wild.

Many studies in humans and other animals have shown the importance of early life experiences in shaping how individuals develop. But researchers know less about how multiple, different instances of adversity or stressors can accumulate within the body and what their overall impact is on an animal’s well-being.

Wild populations face many kinds of stressors. They compete for food, risk getting eaten by a predator, suffer illness and must contend with extreme weather conditions. And as if life in the wild wasn’t hard enough, humans are now adding additional stressors such as chemical, light and sound pollution, as well as habitat destruction.

Given the widespread loss of biodiversity, understanding how animals react to and are harmed by these stressors can help conservation groups better protect them. But accounting for such a diversity of stressors is no easy feat. To address this need and demonstrate the cumulative impact of multiple stressors, our research team decided to develop an index for wild animals based on psychological research on human childhood trauma.

A cumulative adversity index

Developmental psychologists began to develop what psychologists now call the adverse childhood experiences score, which the amount of adversity a person experienced as a child. Briefly, this index adds up all the adverse events – including forms of neglect, abuse or other household dysfunction – an individual experienced during childhood into a single cumulative score.

This score can then be used to predict later-life health risks such as chronic health conditions, mental illness or even economic status. This approach has revolutionized many human health intervention programs by identifying at-risk children and adults, which allows for more targeted interventions and preventive efforts.

So, what about wild animals? Can we use a similar type of score or index to predict negative survival outcomes and identify at-risk individuals and populations?

These are the questions we were interested in answering in our latest research paper. We developed a framework on how to create a cumulative adversity index – similar to the adverse childhood experiences score, but for populations of wild animals. We then used this index to gain insights about the survival and longevity of yellow-bellied marmots. In other words, we wanted to see whether we could use this index to estimate how long a marmot would .

A marmot case study

Yellow-bellied marmots are a large ground squirrel closely related to groundhogs. Our research group has been studying these marmots in Colorado at the Rocky Mountain Biological Laboratory since 1962.

A marmot with a small device clipped to its ear, looking upwards.

A marmot wearing an ear tag.

Xochitl Ortiz Ross

Yellow-bellied marmots are an excellent study system because they are diurnal, or active during the day, and they have an address. They live in burrows scattered across a small, defined geographical area called a colony. The size of the colony and the number of individuals that reside within varies greatly from year to year, but they are normally composed of matrilines, which means related females tend to remain within the natal colony, while male relatives move away to find a new colony.

Yellow-bellied marmots hibernate for most of the year, but they become active between April and September. During this active period, we observe each colony and regularly trap each individual in the population – that’s over 200 unique individuals just in 2023. We then mark their backs with a distinct symbol and give them uniquely numbered ear tags so they can be later identified.

Although they can live up to 15 years, we have detailed information about the life experiences of individual marmots spanning almost 30 generations. They were the perfect test population for our cumulative adversity index.

Among the sources of adversity, we included ecological measures such as a late spring, a summer drought and high predator presence. We also included parental measures such as an underweight or stressed mother, being born or weaned late, and losing their mother. The model also included demographic measures such as being born in a large litter or having many male siblings.

Importantly, we looked only at females, since they are the ones who tend to stay home. Therefore, some of the adversities listed are only applicable to females. For example, females born in litters with many males become masculinized, likely from the high testosterone levels in the mother’s uterus. The females behave more like males, but this also reduces their life span and reproductive output. Therefore, having many male siblings is harmful to females, but maybe not to males.

A yellow-bellied marmot shown on a trail camera in Montana.

So, does our index, or the number of adverse events a marmot experienced early on, explain differences in marmot survival? We found that, yes, it does.

Experiencing even just one adversity before age 2 nearly halved an adult marmot’s odds of survival, regardless of the type of adversity they experienced. This is the first record of lasting negative consequences from losing a mother in this species.

So what?

Our study isn’t the only one of its kind. A few other studies have used an index similar to the human adverse childhood experiences score with wild primates and hyenas, with largely similar results. We are interested in broadening this framework so that other researchers can adopt it for the species they study.

A better understanding of how animals can or cannot cope with multiple sources of adversity can inform wildlife conservation and management practices. For example, an index like ours could help identify at-risk populations that require a more immediate conservation action.

Instead of tackling the one stressor that seems to have the greatest effect on a species, this approach could help managers consider how best to reduce the total number of stressors a species experiences.

For example, changing weather patterns driven by global heating trends may create new stressors that a wildlife manager can’t address. But it might be possible to reduce how many times these animals have to interact with people during key times of the year by closing trails, or providing extra food to replace the food they lose from harsh weather.

While this index is still in early development, it could one day help researchers ask new questions about how animals adapt to stress in the wild.The Conversation

Xochitl Ortiz RossUniversity of California, Los Angeles

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post Wild animals can experience trauma and adversity too − as ecologists, we came up with an index to track how it affects them appeared first on theconversation.com

Continue Reading

The Conversation

Comparing the Trump and Harris records on technology regulation

Published

on

theconversation.com – Anjana Susarla, Professor of Information , Michigan – 2024-10-18 07:22:00

The Federal Trade Commission is one of the main venues for regulation of big tech and its wares.

Alpha Photo/Flickr, CC BY-NC

Anjana Susarla, Michigan State University

It’s not surprising that technology regulation is an important issue in the 2024 U.S. presidential campaign.

The past decade has seen advanced technologies, from social algorithms to large language model artificial intelligence systems, profoundly affect society. These changes, which spanned the Trump and Biden-Harris administrations, spurred calls for the federal government to regulate the technologies and the powerful corporations that wield them.

As a researcher of information systems and AI, I examined both candidates’ records on technology regulation. Here are the important differences.

Algorithmic harms

With artificial intelligence now widespread, governments worldwide are grappling with how to regulate various aspects of the technology. The candidates offer different visions for U.S. AI policy. One area where there is a stark difference is in recognizing and addressing algorithmic harms from the widespread use of AI technology.

AI affects your in ways that might escape your notice. Biases in algorithms used for lending and hiring decisions could end up reinforcing a vicious cycle of discrimination. For example, a student who can’t get a loan for college would then be less likely to get the education needed to pull herself out of poverty.

At the AI Safety Summit in the U.K. in November 2023, Harris spoke of the promise of AI but also the perils from algorithmic bias, deepfakes and wrongful arrests. Biden signed an executive order on AI on Oct. 30, 2023, that recognized AI systems can pose unacceptable risks of harm to civil and human rights and individual well-being. In parallel, federal agencies such as the Federal Trade Commission have carried out enforcement actions to guard against algorithmic harms.

a man sits at a desk writing on a piece of paper as a woman looks on

signs an executive order addressing the risks of artificial intelligence on Oct. 30, 2023, with Vice President Kamala Harris at his side.

AP Photo/Evan Vucci

By contrast, the Trump administration did not take a public stance on mitigation of algorithmic harms. Trump has said he wants to repeal President Biden’s AI executive order. In recent interviews, however, Trump noted the dangers from technologies such as deepfakes and challenges posed to security from AI systems, suggesting a willingness to engage with the growing risks from AI.

Technological standards

The Trump administration signed the American AI Initiative executive order on Feb. 11, 2019. The order pledged to double AI research investment and established the first set of national AI research institutes. The order also included a plan for AI technical standards and established guidance for the federal government’s use of AI. Trump also signed an executive order on Dec. 3, 2020, promoting the use of trustworthy AI in the federal government.

The Biden-Harris administration has tried to go further. Harris convened the heads of Google, Microsoft and other tech companies at the White House on May 4, 2023, to undertake a set of voluntary commitments to safeguard individual rights. The Biden administration’s executive order contains an important initiative to probe the vulnerablity of very large-scale, general-purpose AI models trained on massive amounts of data. The goal is to determine the risks hackers pose to these models, including the ones that power OpenAI’s popular ChatGPT and DALL-E.

a man in a business suit waves from in front of the door to an airplane

Donald Trump departs from Washington D.C., on Feb. 11, 2019, shortly after signing an executive order on artificial intelligence that included setting technical standards.

Nicholas Kamm/AFP via Getty Images

Antitrust

Antitrust law enforcement – restricting or conditioning mergers and acquisitions – is another way the federal government regulates the technology industry.

The Trump administration’s antitrust dossier includes its attempt to block AT&T’s acquisition of Time Warner. The merger was eventually allowed by a federal judge after the FTC under the Trump administration filed a suit to block the deal. The Trump administration also filed an antitrust case against Google focused on its dominance in internet search.

Biden signed an executive order on July 9, 2021, to enforce antitrust laws arising from the anticompetitive effects of dominant internet platforms. The order also targeted the acquisition of nascent competitors, the aggregation of data, unfair competition in attention markets and the surveillance of users. The Biden-Harris administration has filed antitrust cases against Apple and Google.

The Biden-Harris administration’s merger guidelines in 2023 outlined rules to determine when mergers can be considered anticompetitive. While both administrations filed antitrust cases, the Biden administration’s antitrust push appears stronger in terms of its impact in potentially reorganizing or even orchestrating a breakup of dominant companies such as Google.

Cryptocurrency

The candidates have different approaches to regulation. Late in his administration, Trump tweeted in support of cryptocurrency regulation. Also late in Trump’s administration, the federal Financial Crimes Enforcement Network proposed regulations that would have required financial firms to collect the identity of any cryptocurrency wallet to which a user sent funds. The regulations were not enacted.

Trump has since shifted his position on cryptocurrencies. He has criticized existing U.S. laws and called for the United States to be a Bitcoin superpower. The Trump campaign is the first presidential campaign to accept payments in cryptocurrencies.

The Biden-Harris administration, by contrast, has laid out regulatory restrictions on cryptocurrencies with the Securities and Exchange Commission, which brought about a series of enforcement actions. The White House vetoed the Financial Innovation and Technology for the 21st Century Act that aimed to clarify accounting for cryptocurrencies, a bill favored by the cryptocurrency industry.

Data privacy

Biden’s AI executive order calls on Congress to adopt privacy legislation, but it does not a legislative framework to do so. The Trump White House’s American AI Initiative executive order mentions privacy only in broad terms, calling for AI technologies to uphold “civil liberties, privacy, and American values.” The order did not mention how existing privacy protections would be enforced.

Across the U.S., several states have tried to pass legislation addressing aspects of data privacy. At present, there is a patchwork of statewide initiatives and a lack of comprehensive data privacy legislation at the federal level.

The paucity of federal data privacy protections is a stark reminder that while the candidates are addressing some of the challenges posed by developments in AI and technology more broadly, a lot still remains to be done to regulate technology in the public interest.

Overall, the Biden administration’s efforts at antitrust and technology regulation seem broadly aligned with the goal of reining in technology companies and protecting consumers. It’s also reimagining monopoly protections for the 21st century. This seems to be the chief difference between the two administrations.The Conversation

Anjana Susarla, Professor of Information Systems, Michigan State University

This article is republished from The Conversation under a Creative Commons license. Read the original article.

Read More

The post Comparing the Trump and Harris records on technology regulation appeared first on .com

Continue Reading

Trending