Virus Response Insights Offer MS Treatment Hope

A study has revealed how white blood cells in the body react to the Epstein-Barr virus (EBV) in a way that drives autoimmunity in multiple sclerosis (MS), which could help identify and design effective therapies for the condition.

The findings, in Science Translational Medicine, shine a light on immune mechanisms underlying the well-established link between infection with the virus and MS.

Nearly all MS patients are seropositive for EBV, a widespread virus that commonly causes glandular fever otherwise known as infectious mononucleosis­.

The research suggests T helper cells in the immune system—also called CD4+ cells—primarily react to particular EBV viral components.

These responses were reduced by B cell depletion therapy, which has known MS therapeutic benefit, suggesting it may be particularly useful for the disease.

“By identifying a readily measurable, peripherally accessible immune response linked to disease, this work provides a foundation for the rational design and monitoring of EBV-targeted vaccines and antivirals for MS,” added the researchers.

MS is a chronic inflammatory disease in which the central nervous system is attacked by the body’s own immune system, damaging the fatty myelin sheath protecting nerves and causing other harm that can affect vision, movement and cognition.

Kjetil Bjornevik, PhD, from Harvard T. H. Chan School of Public health in Boston, and colleagues examined how the virus could affect immune responses by studying healthy individuals, people with treatment-naive MS, and those with MS receiving disease-modifying therapies.

The team found that the CD4+ T cells in the group with MS predominantly targeted several viral components, in particular the late lytic capsid and glycoprotein antigens that are components of EBV viral particles.

The EBV-specific CD4+ T cell response in untreated MS patients was twice that of healthy control individuals, further implicating this response. Responses to other herpes viruses remained similar, suggesting a specific role for EBV in MS immunopathology.

B cell depletion therapy based on anti-CD20 antibodies reduced CD4+ T cell responses to the virus 2.5-fold in two MS groups comprising a total of 69 patients. It also eliminated detectable viral shedding in saliva, consistent with B cells serving as the primary viral reservoir.

The authors concluded: “Together, our findings establish EBV viral particle antigens as the dominant targets of CD4+ T cell responses in MS and demonstrate that these responses are modifiable by current therapies.”

The post Virus Response Insights Offer MS Treatment Hope appeared first on Inside Precision Medicine.

The Download: OpenAI unveils GPT-Red and heat pumps rise in the US

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

Meet GPT-Red: an LLM super-hacker OpenAI built to make its models safer

OpenAI has built an LLM super-hacker called GPT-Red that it uses as a sparring partner to help its other models boost their defenses against cyberattacks. 

It automates a type of safety evaluation for software systems known as red-teaming, which is typically done by a team of human testers. The aim is to find as many different ways to break or hijack a system as possible.

OpenAI gave MIT Technology Review an exclusive peek into the system. Find out how it could keep the company ahead of human attackers.

—Will Douglas Heaven

Why heat pumps are still so hot in the US

—Casey Crownhart

It feels as if it should be illegal to even think about heating appliances during the height of summer, but we need to talk about heat pumps. 

The appliances use electricity for heating, they’re incredibly efficient, and they’re on the rise. In the US, their sales have doubled over the past 15 years, according to a new report. They’re also winning the heating race against fossil fuels, outpacing natural-gas furnaces by 32% during the first quarter of 2026. 

These stats are especially striking at this moment, because a key tax credit for heat pumps just ended. So why are heat pumps still so hot? Read the full story for the answer.

This article is from The Spark, our weekly climate tech newsletter. Sign up to receive it in your inbox every Wednesday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Elon Musk discreetly bought a $1 billion gas turbine firm to power Grok
He acquired fossil fuel company APR Energy in May. (Electrek)
+ The most likely application will be powering AI data centers. (Engadget)
+ The deal was revealed through an FTC filing. (Gizmodo)
+ What will power AI’s growth? (MIT Technology Review)
 
2 A hack shows the Suno AI music generator scraped YouTube, Deezer
It scraped decades’ worth of music to train its models. (404 Media)
+ The hacked is a unique look into the black boxes powering GenAI. (CNET)
+ AI is coming for music, too. (MIT Technology Review)
 
3 Thinking Machines has launched an open-weight AI model
Inkling offers a US alternative to China’s open-source models. (Reuters $)
+ It’s the first AI model built by Thinking Machines. (WSJ $)
+ The startup was founded by former OpenAI CTO Mira Murati. (Axios)
 
4 Europe is narrowing its ambitions for tech independence
Manufacturing and research show promise, but funding is a problem. (NYT $)
+ Earnings are strong, but an AI gap persists. (Reuters $)
+ India is also scrambling for AI independence. (MIT Technology Review)
 
5 Earth is absorbing energy at a rate that’s alarming climate scientists
The planet is taking in more heat than models predicted. (Economist $)
+ The legal case for climate justice is growing. (MIT Technology Review)
 
6 The AI backlash has tech executives fearing for their lives
Violent threats against AI firms are spilling into the real world. (WSJ $)
+ An anti-AI movement is growing globally. (MIT Technology Review)
 
7 A Moroccan intelligence insider exposed widespread Pegasus use
Including to target journalists, activists, and foreign politicians. (Guardian)

8 AI is powering citizen-led disaster relief from afar for Venezuela
It’s helping to locate missing people and coordinate relief. (Rest of World)
 
9 Thermodynamic computers could turn noise into useful calculations
They may offer a cooler, more efficient way to process information. (Quanta)

10 An engineer has explained every ’90s computer in Jurassic Park
Fans have debated the technology in the film for decades. (Ars Technica)

Quote of the day

“We hit pause because the communities powering AI should share in its success. Maybe that’s a novel concept in Washington.” 

—New York Gov. Kathy Hochul responds on X to President Donald Trump’s criticism of her state’s new data center moratorium.

One More Thing


Will we ever trust robots?

Robotics firm Prosper is developing a humanoid called Alfie to perform tasks in homes, hospitals, and hotels. The company’s founder, Shariq Hashme, has identified trustworthiness as the top design priority—and first hurdle to clear before humanoids can live up to their hype.

Hashme believes one essential tactic to get people to put their trust in Alfie is to build a detailed character from the ground up—something humanlike but not too human. But the robot’s reliance on remote human operators raises broader questions about privacy, labour, and whether society will truly accept humanoids in our private spaces.

Read the full story on the humanoid trust dilemma.

—James O’Donnell

We can still have nice things

A place for comfort, fun, and distraction to brighten up your day. (Got any ideas? Drop me a line.)

+ Meet the man behind the world’s most beautiful books.
+ 3D printing has revived Roman Britain’s most popular board game.
+ “Lucha Libro” is an imaginative idea to boost literacy: staging live wrestling matches in US libraries.
+ Over 100 years after the death of legendary explorer Ernest Shackleton, the wreck of his final ship has been photographed for the first time.

Mechanical characterization of vertically aligned carbon nanotube forest microelectrodes for neural interfacing

IntroductionThis study aims to explore the mechanical properties of porous microelectrodes formed from vertically aligned carbon nanotube (CNT) forests. Specifically, we investigate the range of effective CNT-based microelectrode (ME) moduli that can be fabricated and identify moduli within that range that significantly reduce strain on brain tissue during micromotion.Materials and methodsTo address these questions, we developed a micromechanical measurement method, known as the dual deflection (DD) test, which is compatible with microelectrode array (MEA) form factors and can measure a wide range of moduli with a 30% uncertainty. Using the DD test with small deflections, we measured the effective Young’s modulus of freestanding CNT microelectrodes (MEs) fabricated with different carbon infiltration times (0, 15, and 30 s) at 900°C. We also developed a static 10 μm deflection finite element analysis (FEA) model to compare the brain tissue strain induced by probes with the maximum (1.7 GPa), median (72 MPa), and minimum (3.9 MPa) measured CNT moduli, along with the modulus of silicon (165 GPa) for comparison.ResultsThe DD test results showed mean effective moduli of 19.6 ± 14.5 MPa, 67.7 ± 22.7 MPa, and 168 ± 62.3 MPa for arrays fabricated with 0, 15, and 30 s infiltrations, respectively. The FEA model revealed that probes with the maximum CNT modulus induced similar strain to the silicon probes at the tip, while probes with the minimum and median CNT moduli showed minimal strain at the tip.DiscussionThese findings suggest that CNT microelectrodes with moduli in the tens of MPa range, achievable through 15 s of carbon infiltration, can significantly reduce brain tissue strain. Additionally, we consistently observed that microelectrodes with 15 s of infiltration were apparently undamaged after deflection, making them mechanically promising candidates for neural probe arrays.

Advances in biomarkers for diagnosing and prognosticating disorders of consciousness

Disorders of Consciousness (DoC) resulting from brain injury comprise a spectrum of clinical syndromes, where the level of impairment varies considerably depending on lesion location, etiology, severity, and individual patient factors. These differences substantially influence both rehabilitation strategies and long-term prognosis. Current diagnostic assessment relies primarily on behavioral scales, supplemented by electrophysiological and neuroimaging studies; however, these approaches remain limited in objectivity, sensitivity, and accessibility. There is therefore a clinical need for biologically informative biomarkers to improve diagnostic precision and prognostic stratification in DoC. This narrative review synthesizes recent advances in biomarker research, encompassing proteomic, metabolomic, and microRNAs (miRNAs) signatures, across multiple biological specimens. We evaluate the findings spanning exploratory to validation stages, and discuss their translational potential, providing a valuable reference for future large-scale, multicenter investigations.

Neural speech encoding in fetal alcohol spectrum disorder: an exploratory study using frequency-following responses

BackgroundFetal alcohol spectrum disorder (FASD) is associated with neurodevelopmental impairments, including listening difficulties not always explained by peripheral hearing loss, suggesting alterations at the level of neural auditory processing. The frequency-following response (FFR) provides an objective measure of neural speech encoding and may offer insight into auditory function in this population.MethodsTwenty-five normal-hearing participants were included: 11 individuals with FASD and 14 controls. Speech-evoked FFRs were recorded using a 160 ms /da/ stimulus at 80 dB SPL with a stimulation rate of 4.35/s. Pitch tracking, stimulus–response correlation, response latency, and signal quality were analyzed using non-parametric statistics. Given the exploratory nature of the study and uncontrolled demographic variables — including a significant age difference between groups — findings should be interpreted as preliminary and hypothesis-generating.ResultsCompared to controls, individuals with FASD showed reduced pitch-tracking consistency and lower stimulus–response correlation, with the most robust finding being a large-effect reduction in F0-range correlation (R[70–120 Hz]: p < 0.001, rank-biserial r = 0.823). Prolonged latencies were observed across multiple response components, and signal-to-noise ratio tended to be lower in the FASD group.ConclusionThis exploratory proof-of-concept study provides preliminary evidence of altered neural speech encoding in individuals with FASD despite normal peripheral hearing. Given uncontrolled confounds, observed differences cannot be specifically attributed to FASD. These findings establish the feasibility of FFR in this population and provide the empirical foundation for future controlled research on auditory biomarkers and intervention monitoring in FASD.

The vagus nerve as a neurovisceral interface: a comprehensive review

The vagus nerve is the longest cranial nerve and a key component of the autonomic nervous system, functioning as a neurovisceral interface between the brain and peripheral organs. Despite well-defined anatomy, the mechanisms underlying its integrative roles in cardiovascular, metabolic, and neuropsychiatric regulation remain incompletely understood. This narrative review synthesizes current evidence on the anatomical organization, physiological functions, and clinical relevance of the vagus nerve, focusing on cardiac autonomic control, gastrointestinal and metabolic regulation, the gut–brain axis, and vagus nerve stimulation. In the cardiovascular system, it interacts with sympathetic pathways within the cardiac plexus and intrinsic cardiac nervous system to regulate heart rate and conduction. In the gastrointestinal system, it coordinates motility, secretion, and metabolic homeostasis through nutrient- and hormone-sensitive pathways. Within the gut–brain axis, emerging evidence highlights rapid neuroepithelial signaling and microbiota-dependent modulation mediated by vagal circuits. The vagus nerve stimulation represents a promising therapeutic strategy for restoring autonomic balance, although challenges remain in fiber selectivity and clinical variability. Advances in multi-omics approaches are beginning to reveal the molecular heterogeneity of vagal neurons, but significant gaps persist due to limited human anatomical data. In conclusion, the vagus nerve functions as a multidimensional integrative system, and a deeper understanding of its structure and molecular organization is essential for developing precise neuromodulatory therapies.

Beyond visual inspection: the deep learning revolution in quantitative cerebrovascular imaging

The rising global burden of cerebrovascular disease, propelled by an aging population, highlights the inherent limitations of conventional, labor-intensive diagnostic paradigms. In the context of time-sensitive stroke management, variability in image interpretation and the high rate of misclassification, particularly during the assessment of transient ischemic attack (TIA), underscore the urgent need for more consistent and efficient diagnostic solutions. Artificial intelligence (AI), particularly deep learning (DL), offers a transformative pathway by automating the analysis of complex neurovascular imaging. Here, we conduct a comprehensive examination of how DL is revolutionizing stroke-related image analysis, moving beyond general assertions of potential to discuss specific technical implementations. We systematically detail the evolution from traditional segmentation algorithms to advanced deep learning architectures—such as U-Net, DeepMedic, and their variants—in performing critical tasks. These tasks encompass the automated segmentation of intracranial and extracranial (carotid) arteries, the quantification of stenosis and plaque burden, and the hemodynamic assessment of vascular lesions across modalities including MRA, CTA, and DSA. By synthesizing landmark studies, our analysis delineates three core aspects: the technological trajectory of DL models in achieving expert-level accuracy in vascular feature extraction in controlled studies; the clinical translation of these tools into diagnostic, prognostic, or therapeutic procedural planning workflows; and the persistent challenges and future directions, including data standardization, model generalizability, and multimodal integration. This review posits that DL represents not merely an assistive technology but a foundational cornerstone for the next generation of precision cerebrovascular medicine. It holds the potential to bridge critical gaps in diagnostic speed, objectivity, and accessibility, provided its development and validation are guided by rigorous, interdisciplinary collaboration.

A study on the relationship between research stress, research anxiety, research performance, and job satisfaction among Chinese healthcare professionals and its influencing mechanisms: a national multi-center survey

BackgroundChinese healthcare professionals face dual pressures from clinical duties and research activities, which may increase research-related anxiety and reduce job satisfaction. While occupational stress has been well studied, the mechanisms linking research stress to job satisfaction are unclear. Research performance may also play a key role, but its impact remains uncertain. This study explores the relationships between research stress, research anxiety, research performance, and job satisfaction, and identifies the pathways connecting these variables.MethodsA cross-sectional survey was conducted among healthcare professionals (nurses, physicians, and pharmacists) from eastern, central, and western regions of China to assess research-related conditions and job satisfaction (JS). Job satisfaction was measured using a modified validated scale. Research stress (RS), research anxiety (RA), and research performance (RP) were assessed using self-developed instruments. Structural equation modeling (SEM) was employed to perform path analysis and mediation analysis.ResultsA total of 924 healthcare professionals were included in the study. Over 85% reported working more than 40 hours per week, and 66.5% rated their health status as fair or poor. Regarding the types of research conducted by healthcare professionals, clinical research accounted for the highest proportion (63.2%), followed by basic science research (9.8%), while health services research and community-based research were relatively less common. The mean scores of JS, RS, RA, and RP were 3.40 ± 0.88, 2.56 ± 0.70, 2.34 ± 0.87, and 3.55 ± 0.82, respectively. Path analysis revealed that research stress was positively associated with research anxiety and negatively associated with both research performance and job satisfaction. Research anxiety was also negatively associated with research performance and job satisfaction, whereas research performance was positively associated with job satisfaction. Mediation analysis indicated that research stress was associated with job satisfaction both directly and indirectly through research anxiety and research performance.ConclusionsResearch stress is negatively associated with job satisfaction among healthcare professionals through increased research anxiety, whereas higher research performance is positively associated with job satisfaction. Hospitals and healthcare institutions should optimize the research environment, strengthen psychological support, and enhance research resource allocation to reduce research stress, improve research performance, and ultimately increase job satisfaction.

ADHD as a disorder of operational capacity: a buffered-state framework for sustained engagement instability

Attention-deficit/hyperactivity disorder (ADHD) is commonly conceptualized as a disorder of executive dysfunction or impaired motivation. In clinical practice, however, many individuals with ADHD describe a different core difficulty: they can initiate tasks appropriately, understand goals, and exert effort, yet struggle to remain engaged as time-on-task and internal strain accumulate. These failures are often abrupt, ego-dystonic, and strongly context-dependent, posing challenges for models based solely on static deficits or gradual depletion. Building on this phenomenology, we propose a constraint-level framework that conceptualizes ADHD as a disorder of operational capacity—the finite tolerance for sustaining stable mental engagement under cumulative cognitive and emotional demand. Within this framework, sustained engagement is modeled as a buffered state that remains stable while cumulative load is within tolerance but becomes increasingly fragile near saturation and may undergo nonlinear collapse into disengagement. We formalize this governing constraint as Operational Economy (OE)—a regulatory, not energetic, constraint principle—and describe a pre-volitional Global Switching Framework (GSF) that mediates rapid state transitions once engagement stability is exceeded. This formulation integrates and extends cognitive-energetic and state-regulation accounts while explicitly differentiating the OE/GSF framework from these predecessors by its emphasis on threshold-dependent dynamics, bounded tolerance, and asymmetric re-engagement. This framework offers a coherent account of clinical features such as intact initial performance with late-stage disengagement, pronounced time-on-task effects, and context-dependent hyperfocus. The framework generates falsifiable predictions regarding sensitivity to cumulative demand, stress-related acceleration of disengagement, pharmacological effects that preferentially extend engagement endurance rather than normalize peak executive performance, and individual differences across ADHD subtypes and comorbid presentations. These predictions are consistent with existing empirical findings on time-dependent variability in ADHD and provide a basis for future experimental validation.