Autism may have two distinct subtypes based on brain connectivity patterns

Amazing stuff! I procrastinated on blogging about this one!

"... Some neuroscientists have been exploring the possibility that this well-documented diversity partly reflects differences in the brain's organization and underlying neurobiology. However, so far only a few studies have been able to link differences in autistic behavior to specific neurobiological processes. ...

Their findings ... led to the identification of two distinct autism subtypes characterized by distinct connectivity patterns. ...

"Autism is extremely heterogeneous clinically, and for many years imaging studies have also reported heterogeneous and sometimes apparently conflicting findings: some studies found reduced functional connectivity, others found increased connectivity, and others found more complex patterns. ..."

From the abstract:

"It is often assumed that phenotypic heterogeneity in autism reflects underlying pathobiological variation. However, direct evidence supporting this link is lacking. Leveraging cross-species functional neuroimaging, we show that brain dysconnectivity patterns in autism can be parsed into biologically dissociable subtypes.

Specifically, we found that functional magnetic resonance imaging (fMRI) connectivity alterations in 20 distinct genetic mouse models of autism cluster into hypoconnectivity-dominant and hyperconnectivity-dominant subtypes.

These subtypes are linked to distinct biological pathways, with hypoconnectivity being associated with synaptic dysfunction and hyperconnectivity reflecting transcriptional and immune-related alterations.

Here we identified analogous hypoconnectivity and hyperconnectivity subtypes in a multicenter human fMRI dataset of n = 940 individuals with idiopathic autism and n = 1,036 neurotypical individuals. The human autism subtypes are highly replicable, are associated with distinct functional network architectures and behavioral profiles and recapitulate the synaptic and immune-related pathways identified in the rodent dataset.

Our work provides a new empirical framework for targeted subtyping of the autism spectrum."

Autism may have two distinct subtypes based on brain connectivity patterns

Nature Neuroscience: Autism subtypes identified using cross-species functional connectivity analyses "It is often assumed that phenotypic heterogeneity in autism reflects underlying pathobiological variation. However, direct evidence supporting this link is lacking. Leveraging cross-species functional neuroimaging, we show that brain dysconnectivity patterns in autism can be parsed into biologically dissociable subtypes."

Autism subtypes identified using cross-species functional connectivity analyses (open access)

Fig. 1: fMRI connectivity in 20 autism mouse models clusters into dominant hypoconnectivity and hyperconnectivity subtypes.

Can a faint glow from cell metabolism flag cancer or neurodegenerative disease long before symptoms appear

Recommendable! This is about a survey and a research article. This could be a breakthrough!

"Every living thing emits a faint glow from cell metabolism, and researchers now hope the ultraweak light can flag cancer or neurodegenerative disease long before symptoms appear."

"In 2025, quantum physicist Daniel Oblak and his colleagues published images of mice that made headlines around the world. The researchers put four anaesthetized mice inside a dark chamber and pointed a sensitive camera at them. It picked up a stream of photons emitted from the animals’ skin: a light signal too faint to be seen by the human eye, but visible in the camera images as a ghostly, mouse-shaped glow. ...

The idea that cells might communicate through extremely low-intensity light was first suggested in the 1920s by Russian biologist Alexander Gurwitsch ...

Biophotons are produced during aerobic metabolism, in which cells use oxygen to break down nutrients and extract energy. This mostly occurs in mitochondria. During metabolic reactions, short-lived chemicals produced from oxygen, called reactive oxygen species (ROS), generate intermediates that decompose to produce molecules in an ‘excited’ electronic state. As these relax into more stable arrangements, they can release energy in the form of photons.

A high-energy ‘singlet’ state of oxygen can emit a photon of red light as it decays, for example, whereas excited carbonyl (C=O) groups glow in the blue-green range. ..."

From the abstract:

"Biophotons are non-thermal and non-bioluminescent ultraweak photon emissions, first hypothesised by Gurwitsch as a regulatory mechanism in cell division, and then experimentally observed in living organisms.

Today, two main hypotheses explain their origin:
stochastic decay of excited molecules and coherent electromagnetic fields produced in biochemical processes.

Recent interest focuses on the role of biophotons in cellular communication and disease monitoring.

This study presents the first campaign of biophoton emission measurements from cultured astrocytes and glioblastoma cells ... 

The statistical analyses of the collected data revealed a clear separation between cellular signals and dark noise, confirming the high sensitivity of the apparatus. The Diffusion Entropy Analysis (DEA) was applied to the data to uncover dynamic patterns, revealing anomalous diffusion and long-range memory effects that may be related to intercellular signaling and cellular communication.

These findings support the hypothesis that biophoton emissions encode rich information beyond intensity, reflecting metabolic and pathological states.

The differences revealed by applying the Diffusion Entropy Analysis to the biophotonic signals of Astrocytes and Glioblastoma are highlighted and discussed in the paper.

This work lays the groundwork for future studies on neuronal cultures and proposes biophoton dynamics as a promising tool for non-invasive diagnostics and the study of cellular communication."

Wednesday, July 29, 2026 - Join The Flyover

All living things emit a faint glow. Could this light be useful? "Ultra-weak ‘biophotons’ might be used to diagnose disease, or could even represent a new signalling mechanism in cells."

First Experimental Measurements of Biophotons from Astrocytes and Glioblastoma Cell Cultures (open access)

A sensitive camera picks up biophoton emissions from a mouse (alive, left; dead, right).

Figure 2. Comparison plots of biophotonic signals emitted by astrocyte (green) and glioblastoma (red) cell cultures concerning the background (dark counts in blue) performed with the LNF machine

Unlocking the Medical Potential of Menstrual Blood. Really!

A disturbing article! When ideology results in absurdity!

Notice the Johns Hopkins Bloomberg School of Public Health does not mention women at all in this article as if transgender men have menstrual blood!

P.S. I believe, the medical analysis of menstrual blood has been done for a long time. Thus, this article is dated!

"An emerging field of study and innovation is stemming from a long-discarded biological resource: menstrual blood.  

A growing number of researchers and biotech companies are exploring the potential of such blood samples as a tool for understanding fertility, diagnosing disease, and studying the body’s healing systems.

Recent developments include:

A system that collects and analyzes blood from used tampons, creating a “liquid biopsy” without invasive procedures.

A self-collection kit that uses a tampon to screen for HPV.

An FDA-approved pad used for hemoglobin A1c monitoring.  

A menstrual cup used for endometriosis diagnosis.  

Broader insights: Researchers are also studying whether menstruation may provide a “blueprint” for the body’s response to inflammation and tissue repair.

Reasons to Be Cheerful"

Global Health NOW: Hope and Headwinds in the HIV Fight; and Unlocking the Medical Potential of Menstrual Blood

Tampons Are Becoming a Valuable Medical Tool "Menstrual blood is usually treated as waste. But it could be key to helping scientists understand how our bodies heal."

Fitness tracker early on indicate pregnancy and miscarriage in women before conventional pregnancy tests

Good news!

"... For many women, the first sign that they might be pregnant is a period that doesn't arrive when expected. This is around the same time it's possible to take a pregnancy test. With the exception of some tests that claim to be able to detect a pregnancy early, most aren't able to spot the change in hormones until the first day of a missed period.

But Ravika is one of hundreds of women who are reporting online that they are seeing the signs earlier than this, through the data on their fitness trackers.

Fitness trackers - including smart rings, smart watches and chest-mounted trackers - have surged in popularity in the UK. According to YouGov research, 36% of British adults in January 2026 owned and used a fitness tracker - nearly twice the proportion in July 2019.

Finnish-US company Oura has the highest sales in the smart ring market, which also includes India-based Ultrahuman, and Hong Kong-based RingConn.

Predicting miscarriages as well as pregnancies

One of the key signs of pregnancy is a change to body temperature - which means trackers that can monitor body temperature are popular with those trying to conceive.

In a typical cycle, a woman's body temperature will be elevated until just before her period comes, when it drops sharply. In early pregnancy, body temperature usually stays high. If a fitness tracker measures body temperature, it can pick up on this early symptom.

As well as her stats, Ravika started reading Reddit posts from women who said their Oura ring had predicted not only their pregnancy but their miscarriage as well, through the ring's "major signs" notification. ..."

My fitness tracker knew I was pregnant before I did

Source

Brain glutamate changes could link cannabis use to a higher risk of psychosis

More bad news for e.g. recreational cannabis use! Got to be careful using such drugs!

"... Researchers ... recently carried out a study exploring the possibility that brain glutamate levels could partly explain the relationship between cannabis use and psychotic symptoms. Their paper, published in Molecular Psychiatry, suggests that differences in glutamate levels could be one of the biological pathways linking cannabis use to psychosis. ...

As part of their study, ... examined 79 people with varying levels of psychosis vulnerability. Some of these participants had no known psychiatric conditions and were not experiencing any mental health-related symptoms. Others were considered at a high risk of experiencing psychosis. Finally, a portion of participants were diagnosed with psychosis.

All the participants completed clinical assessments to determine whether they were experiencing positive psychosis symptoms (i.e., hallucinations and delusions), negative psychosis symptoms (i.e., low motivation, a reduced ability to feel pleasure, social withdrawal and a lack of emotional expression), or mood-related symptoms. The researchers also collected information about the patients' cannabis use and analyzed their urine to detect any recent use of the substance.

Finally, the researchers examined the participants' brains using a technique called ultra-high-field 7 Tesla magnetic resonance spectroscopy (1HMRS). This technique allowed them to measure the concentration of specific chemicals in the brain, particularly in a brain region called the anterior cingulate cortex (ACC). ..."

From the abstract:

"Cannabis use is linked to elevated psychosis risk, yet the neurobiological mechanisms that couple use to symptom expression remain unclear. Because glutamatergic dysregulation has been implicated in both cannabis effects and psychosis vulnerability, we examined whether brain glutamate relates to dimensional psychosis symptoms as a function of cannabis use across the psychosis spectrum.

Seventy-nine participants—typically developing controls, clinical high-risk individuals, and patients with psychosis—completed dimensional clinical assessments, detailed cannabis use surveys, urine toxicology, and ultra-high-field 7T magnetic resonance spectroscopy (1HMRS) of the anterior cingulate cortex (ACC).

Linear models assessed the main and interactive effects of ACC glutamate and cannabis use on psychopathology symptoms.

Self-reported cannabis use showed good concordance with urine toxicology, with strongest agreement among frequent users.

Both lower ACC glutamate and higher cannabis use were independently associated with positive and negative psychosis symptoms.

Notably, lower glutamate levels were associated with higher positive symptoms in cannabis users but not cannabis non-users.

Exploratory analyses suggested interactions for depressive and manic symptoms, indicating that glutamatergic abnormalities may amplify the overall severity of cannabis-related symptoms.

Sensitivity analyses revealed lower ACC glutamate in psychosis patients—especially cannabis users—highlighting diagnostic group differences and reinforcing the link between cannabis exposure and glutamatergic dysfunction. These findings implicate ACC glutamatergic dysfunction as a transdiagnostic correlate of symptom burden, particularly in those with psychosis who are cannabis users. Glutamate-targeted interventions and longitudinal designs will be needed to examine causal pathways linking cannabis exposure to psychosis-relevant outcomes."

Brain glutamate changes could link cannabis use to a higher risk of psychosis

Cannabis use and glutamate across the psychosis spectrum: in vivo evidence from 7T proton magnetic resonance spectroscopy (open access)

Fig. 1: Individual metabolite quantification and voxel overlap in single-voxel spectroscopy (SVS) ¹HMRS.

Fig. 4: Glutamate × cannabis interactions predicting dimensional psychosis symptoms.

A portable ultrasound system could make reliable breast imaging more accessible and available for home use

Good news! Cancer is history (soon)!

Again the ideologues and demagogues at MIT demean women  as "people"!!! Horrible!

"For people [women] at high risk of developing breast cancer, yearly mammograms may not be enough to detect tumors early. To make earlier diagnosis easier, an MIT team has developed portable detectors based on ultrasound, which could be used much more frequently. ..."

From the abstract:

"Breast cancer screening and longitudinal monitoring require imaging technologies that are portable, operator-independent, and suitable for frequent use, a capability not fully met by mammography or conventional ultrasound.

We present a three-dimensional (3D) portable ultrasound system for real-time examination (3D PURE) that overcomes key limitations in volumetric breast imaging through advances in transducer design, acoustic materials, and adaptive beamforming.

A box-array design incorporating a corner-gap offset geometry suppresses peak crosstalk (by 3.73 dB at the corner-most element), prevents preamplifier saturation, and supports higher transmit voltages (up to 24 V).

A custom flowable backing layer (impedance 6.12 MRayl; attenuation 7.56 dB mm−1 MHz−1) integrates around fragile wirebonds, reduces inter-element crosstalk by ~4.5 dB throughout the array, and improves axial and lateral/elevational resolutions by ~200 µm and ~70 µm, respectively.

Layered Aberration-Correction Reconstruction (LACR), an adaptive 3D beamformer, compensates for heterogeneous speed-of-sound (SoS) in the breast, reducing depth localization error by 2 mm and aberration defocusing by 70 µm on average at a 5 cm depth.

Nine of the ten participants in an in vitro study showed improved microtarget detection efficiency with 3D PURE relative to a conventional 2D system (p = 0.0215), analogous to detection of microcalcifications.

These results, combined with in vivo imaging validation of various phenotypes, highlight the potential of 3D PURE for reliable breast imaging.

Furthermore, a vision-guided computer interface, MyFUS, ensures self-guided, user-friendly, and operator-independent probe positioning for longitudinal monitoring."

A portable ultrasound system could make reliable breast imaging more accessible | MIT News | Massachusetts Institute of Technology "The new technology, which generates high-resolution, 3D images of breast tissue, requires no expertise to operate and could be used at home."

Portable, real-time 3D ultrasound for operator-independent breast imaging (open access)

Portable, real-time 3D ultrasound for operator-independent breast imaging (contains link to PDF file)

Fig. 1: Overview of the 3D portable ultrasound for real-time examination (PURE) for portable and operator-independent monitoring.

Andes hantavirus PCR test created in two days

Good news! Impressive! This is only the beginning!

"Researchers have managed to develop a diagnostic PCR test for the Andes hantavirus in just two days. The disease sickened more than 10 people and killed three aboard the cruise ship MV Hondius last month. A PCR test is important to track the spread of the virus, which can take up to 42 days to become symptomatic. At the moment, the US Centers for Disease Control and Prevention is using a blood test to detect antibodies in infected people, but it doesn’t work in those not yet showing symptoms."

"... Some other countries have used PCR tests to detect hantavirus, but in the U.S., the Centers for Disease Control and Prevention does not yet have a validated one for diagnosis. The CDC is developing such a test, but in the meantime, the agency has been using a blood test that can detect antibodies in infected people who are symptomatic, but it can’t detect low levels of the virus in asymptomatic people. ..."

P.S. I was unable to find an original news release for this story.

Nature Briefing: Translational Research

Inside the race to develop a hantavirus PCR test "Researchers at the Nebraska Public Health Laboratory worked round the clock to develop a test for the Andes virus at the center of the deadly cruise ship outbreak"

Molecular quantum nanosensors can precisely measure the temperature inside of individual living cells

Amazing stuff! This could be breakthrough!

"A ‘thermometer’ made of quantum nanosensors can detect the temperature in individual cells. With it, researchers have found that temperature varies by as much as 1 ºC in different parts of a cancer cell.

The sensors could also detect signs of oxidative stress and the production of free radicals, reactive molecules that can damage DNA.

The crystal fragments that made up the tool didn’t seem to damage cancer cells when injected or absorbed, but the researchers want to make them even smaller so they can “spy on cellular activities” ... “It’s a fantastic milestone,” s... “It will be so useful for disease detection.”"

"... The MoQN is a technique utilizing a special kind of quantum crystal which is injected into a tissue to be monitored. The crystal reads the surrounding temperature, and this signal is read by an optical magnetic process called optically detected magnetic resonance (ODMR). Conceptually, it has some things in common with the way MRI machines work ... 

The MoQN is nontoxic and can read temperatures within living cells. Also, by applying a quantum property known as molecular spin, MoQNs are fabricated with incredible precision, yielding great homogeneity leading to very high accuracy. ..."

From the abstract:

"Quantitatively mapping temperature within living cells is essential for understanding subcellular biophysical processes; however, existing intracellular quantum sensors such as nanodiamonds with nitrogen-vacancy centers, quantum dots, and fluorescent proteins face limitations in material heterogeneity, cytotoxicity, and thermometric specificity.

Here, we present molecular quantum nanosensors (MoQNs) as a next-generation platform for intracellular quantum sensing. MoQNs embed pentacene molecular spin qubits within para-terphenyl nanocrystals coated with Pluronic F-127, yielding a coherent spin system with molecular-level uniformity and long spin coherence times under physiological conditions.

By chemically suppressing hyperfine interactions, we enhance spectral resolution and demonstrate spatially resolved absolute temperature sensing inside the cytoplasm and nuclei of cancer cells.

MoQNs thus offer a chemically tunable, biologically compatible platform for quantum-level detection of thermal and biochemical states of intracellular environments."

Nature Briefing: Cancer

Quantum ‘thermometer’ takes temperatures inside living cancer cells "Nanosensors showed that different parts of the cell varied by as much as 1 ºC." (no public access)

Tiny molecular quantum thermometer can measure the insides of cells (original news release) "A new way to record temperatures at microscopic scales with incredibly high precision"

Molecular quantum nanosensors functioning in living cells (open access)

Intracellular quantum sensing. A diagram showing how MoQN can report the temperature from within cells. In this example, a cancer cell containing MoQN can be measured by an external ODMR when a small microwave signal is applied to activate the MoQN.

Fig. 1. Schematics of in-cell ODMR using MoQNs and CW-ODMR and pulsed-ODMR characterization.

Fig. 4. Absolute temperature measurement inside the nuclei of cancer cells.

AI is starting to beat human doctors at making correct diagnoses in the emergency room

Good news! Finally, human doctors/phycisians get some serious competition to the benefit of our health!

"... This vision of AI-assisted emergency health care may soon be reality. In a new study, researchers show that a type of AI known as a large language model (LLM) often outperformed physicians at diagnosing complex and potentially life-threatening conditions, including decreased blood flow to the heart, even in the fast-moving stages of real ER care when information is limited, they report today in Science.

In early ER cases, the model identified the correct or a very close diagnosis in about 67% of cases, compared with roughly 50% to 55% for physicians. And the technology is only getting better. ..."

From the editor's summary and abstract:

"Editor’s summary

Computational tools for medical decision support have been advancing over time, mainly by serving as resources for limited applications. Machine learning tools for autonomous interpretation of clinical cases have also been gradually improving over time.

Brodeur et al. pitted a large language model, the OpenAI o1 series, directly against hundreds of physicians at different levels of training and experience on a variety of clinical cases ranging from published patient vignettes to evaluations of brand-new emergency room patients, as well as on clinical tasks including both diagnosis and planning of clinical management  ... Across a variety of scenarios and applications, the large language model outperformed both human physicians and older models, suggesting its potential utility for clinical care.  ...

Abstract

More than 65 years ago, complex clinical diagnostic reasoning cases were introduced as the gold standard for the evaluation of expert medical computing systems, a standard that has held ever since.

In this study, we report the results of a physician evaluation of a large language model (LLM) on challenging clinical cases across five experiments with a baseline of hundreds of physicians.

We then report a real-world study comparing human expert and artificial intelligence (AI) second opinions in randomly selected patients in the emergency room of a major tertiary academic medical center.

In all experiments, the LLM outperformed physician baselines and displayed continued improvement from prior generations of AI clinical decision support. Our study suggests that LLMs have eclipsed most benchmarks of clinical reasoning, motivating the urgent need for prospective trials."

AI is starting to beat doctors at making correct diagnoses | Science | AAAS "Large language model excels at clinical decisions, even in fast pace of a simulated ER"

AI can reason like a physician—what comes next? (Perspective, open access) "Text-based AI can think like a physician; the challenge is achieving safe clinical implementation" [This Perspective has no abstract!]

Performance of a large language model on the reasoning tasks of a physician (no public access)

Superhuman performance of a large language model on the reasoning tasks of a physician (similar preprint, published 12/14/2024, open access) 

Extracting More health Information from Exhaled Breath

Amazing stuff!

"... The first thing a smart mask does to collect data is cool exhaled breath using a hydrogel—a water-based substance that can retain water while maintaining its structure—and capture the resulting liquid, called exhaled breath condensate (EBC). Once EBC is collected, sensors analyze the samples for biomarkers of health, such as metabolites, pathogens, and inflammatory indicators, and transmit the data wirelessly to a personal phone, tablet, or computer. But in previous versions of the mask, the hydrogel dried up after a few hours, making continuous monitoring over time difficult. ..."

From the abstract:

"Exhaled breath condensate (EBC) offers a non-invasive window into respiratory and systemic metabolism, yet wearable EBC systems are constrained by unstable moisture harvesting, short sensor lifetimes in humidity and battery dependence.

Here we present EBClite, a battery-free smart mask for multiday EBC biomarker monitoring, using lactate as a model analyte. The platform integrates a regenerable antidrying porous hydrogel that achieves sustained and reactivatable breath condensation over several days, extending operational lifetime while reducing material cost.

EBClite further incorporates a long-term stable electrochemical lactate sensor and an ultrathin quasi-two-dimensional perovskite solar cell with a power-to-weight ratio of 10 W g−1 and a power conversion efficiency exceeding 30% under ambient indoor illumination, enabling autonomous operation across diverse lighting conditions. Human studies demonstrate strong correlations between EBC and blood lactate levels during exercise and carbohydrate intake, enabling real-time tracking of metabolic fluctuations. EBClite provides a sustainable, user-friendly platform for continuous respiratory and metabolic monitoring."

Extracting More Information from Exhaled Breath - www.caltech.edu "Exhaled breath can provide a treasure trove of health information, offering a noninvasive window to both respiratory microenvironments and systemic physiological states. But collecting such data is a challenge."

A battery-free smart mask for long-term exhaled breath biochemical sensing (no public access)

Illustration of the origin and transport pathway of metabolites (for example, lactate) from blood to the airway surface liquid, where they diffuse across the epithelium and are incorporated into exhaled vapor before condensing as non-volatile metabolites in exhaled breath condensate.

The EBClite smart mask can analyze the chemicals in one's breath in real time.

No evidence ADHD is being over-diagnosed in the UK, say experts. Really!

What is ADHD anyway? Is this not a broad spectrum disorder that allows almost any kid to be diagnosed as such?

How much of normal children's behavior is diagnosed as ADHD? I bet, too much!

When is an underdiagnosis an overdiagnosis?

The more patients, the better for the business of doctors!

"In a paper, published today in the British Journal of Psychiatry, a group of experts say there is no robust evidence that ADHD is over-diagnosed in the UK. They refute the view that ‘nowadays everyone has ADHD’, which is gaining traction in public discourse and has been amplified by some leading politicians, as demand rises for NHS assessments and services. ...

“While many more people with ADHD are being recognised and treated, we are failing to support many more. Overdiagnosis is not a problem, but misdiagnosis may be as people are driven into the private sector by long waits; and sadly, missed diagnoses remain common.” ..."

From the abstract:

"When thoroughly assessed, the prevalence of attention-deficit hyperactivity disorder (ADHD) in children/adolescents is estimated at 5%.

There is no evidence that ADHD is over-diagnosed in the UK. Indeed, available data point to under-diagnosis, even though rigorous updated post-COVID-19 pandemic data are not available.

Some cases may be misdiagnosed due to low-quality assessment, poor adherence to national guidance or inappropriate differential diagnosis. Beyond the controversy around over- or under-diagnosis and over-medicalisation of ordinary behaviours or emotions, the main issue is that UK clinical services cannot adequately support individuals with ADHD who need help. There is a risk that the narrative claiming ‘ADHD is over-diagnosed’ could be used to deny people with properly-diagnosed ADHD the care they deserve."

No evidence ADHD is being over-diagnosed, say experts | University of Cambridge "Experts are warning that far from being over-diagnosed, people with ADHD are waiting too long for assessment, support and treatment."

ADHD (over) diagnosis: fiction, fashion and failure (open access)

One night of sleep vital signs may Predict Illness several years ahead and over 100 health conditions

Amazing stuff!

"Difficulty sleeping often precedes heart disease, psychiatric disorders, and many other illnesses. Researchers used data gathered during sleep studies to detect such conditions.

What’s new: SleepFM is a system that classifies Alzheimer’s, Parkinson’s, prostate cancer, stroke, congestive heart failure, and many other conditions based on a person’s vital signs while asleep — as much as 6 years before they show symptoms. ...

Input/output: Recordings of one night of sleep in, disease classifications out

Architecture: Convolutional neural network encoder, transformer, LSTM

Performance: Can accurately classify over 130 conditions ...

How it works: SleepFM comprises a convolutional neural network (CNN), transformer, and LSTM. The authors trained the system in two stages:

(i) to encode patterns in sleep data and

(ii) to classify diseases.

The training data comprised roughly 585,000 hours of sleep-study recordings that included, in addition to each patient’s age and sex, signals of activity in the brain, heart, respiratory system (airflow, snoring, and blood oxygen level), and leg muscles. The data was mostly proprietary but included public datasets.

The authors trained the CNN and transformer together.  ...

The authors added the LSTM and separately trained it, given 9 hours of sleep data as well as the subject’s age and sex, to classify more than 1,000 diseases.

Results: The authors compared SleepFM’s performance on a proprietary test set to the same system without pretraining and a vanilla neural network that was trained on only demographic information.

Across 14 general categories of disease ..."

From the abstract:

"Sleep is a fundamental biological process with broad implications for physical and mental health, yet its complex relationship with disease remains poorly understood. Polysomnography (PSG)—the gold standard for sleep analysis—captures rich physiological signals but is underutilized due to challenges in standardization, generalizability and multimodal integration.

To address these challenges, we developed SleepFM, a multimodal sleep foundation model trained with a new contrastive learning approach that accommodates multiple PSG configurations.

Trained on a curated dataset of over 585,000 hours of PSG recordings from approximately 65,000 participants across several cohorts, SleepFM produces latent sleep representations that capture the physiological and temporal structure of sleep and enable accurate prediction of future disease risk.

From one night of sleep, SleepFM accurately predicts 130 conditions with a C-Index of at least 0.75 (Bonferroni-corrected P < 0.01), including all-cause mortality (C-Index, 0.84), dementia (0.85), myocardial infarction (0.81), heart failure (0.80), chronic kidney disease (0.79), stroke (0.78) and atrial fibrillation (0.78).

Moreover, the model demonstrates strong transfer learning performance on a dataset from the Sleep Heart Health Study—a dataset that was excluded from pretraining—and performs competitively with specialized sleep-staging models such as U-Sleep and YASA on common sleep analysis tasks, achieving mean F1 scores of 0.70–0.78 for sleep staging and accuracies of 0.69 and 0.87 for classifying sleep apnea severity and presence.

This work shows that foundation models can learn the language of sleep from multimodal sleep recordings, enabling scalable, label-efficient analysis and disease prediction."

The New Open-Weights Leader, Big AI’s Political Influence, Predicting Illness, Faster Reasoning

New AI model predicts disease risk while you sleep "Stanford Medicine scientists and their colleagues created the first artificial intelligence model that can predict more than 100 health conditions from one night’s sleep."

A multimodal sleep foundation model for disease prediction (open access)

Fig. 1: Overview of SleepFM framework.

A portable ultrasound sensor may enable earlier detection of breast cancer

Amazing stuff!

Notice MIT too uses the horrible ideological term "people" for women!!!

"For people [women] who are at high risk of developing breast cancer, frequent screenings with ultrasound can help detect tumors early. MIT researchers have now developed a miniaturized ultrasound system that could make it easier for breast ultrasounds to be performed more often, either at home or at a doctor’s office.

The new system consists of a small ultrasound probe attached to an acquisition and processing module that is a little larger than a smartphone. This system can be used on the go when connected to a laptop computer to reconstruct and view wide-angle 3D images in real-time. ..."

From the abstract:

"Effective resource-constrained volumetric ultrasound imaging requires compact, low-power systems capable of wide-angle real-time 3D imaging to accommodate small changes in placement by the operator. However, obtaining such images requires an excessive O(N2) channel count, bulky electronics, and high power consumption.

We introduce an end-to-end system architecture to enable high-resolution, real-time 3D ultrasound imaging in a portable form factor. We present: a convolutional optimally distributed array (CODA) geometry that drastically reduces the number of elements (from 1024 to 128), a novel chirped data acquisition (cDAQ) architecture that enhances imaging depth while operating with a 25.3 dB lower transmit amplitude than a pulsed system, and an associated new signal processing methodology.

We experimentally demonstrate our system's ability to perform deep (> 11 cm), high axial resolution (< 600 µm), and wide-angle (57°) imaging, while simultaneously reducing power consumption (29.6x reduction) and drive voltage (18 V).

We validated our system in vitro and further performed in vivo human trials, demonstrating the ability to detect both tumors and cysts in breast tissue. This new architectural approach will unlock a new class of medical devices with enhanced diagnostic and long-term monitoring capabilities and open up future wearable designs of real-time 3D ultrasound systems."

A portable ultrasound sensor may enable earlier detection of breast cancer | MIT News | Massachusetts Institute of Technology "The new system could be used at home or in doctors’ offices to scan people who are at high risk for breast cancer."

Real-Time 3D Ultrasound Imaging with an Ultra-Sparse, Low Power Architecture (open access)

Fig. 1 Real-time 3D ultrasound imaging system architecture.

AI-generated sensors open new paths for early cancer detection

Good news! Detecting cancer the earlier, the better! Cancer is history (soon)!

Apparently, the featured research article does not really describe new early cancer detection. Or in other words, there is a mismatch!

"... The researchers developed an AI model to design peptides (short proteins) that are targeted by enzymes called proteases, which are overactive in cancer cells. Nanoparticles coated with these peptides can act as sensors that give off a signal if cancer-linked proteases are present anywhere in the body.

Depending on which proteases are detected, doctors would be able to diagnose the particular type of cancer that is present. These signals could be detected using a simple urine test that could even be done at home. ...

More than a decade ago, Bhatia’s lab came up with the idea of using protease activity as a marker of early cancer. The human genome encodes about 600 proteases, which are enzymes that can cut through other proteins, including structural proteins such as collagen. They are often overactive in cancer cells, as they help the cells escape their original locations by cutting through proteins of the extracellular matrix, which normally holds cells in place.

The researchers’ idea was to coat nanoparticles with peptides that can be cleaved by a specific protease. These particles could then be ingested or inhaled. As they traveled through the body, if they encountered any cancer-linked proteases, the peptides on the particles would be cleaved.

Those peptides would be secreted in the urine, where they could be detected using a paper strip similar to a pregnancy test strip. Measuring those signals would reveal the overactivity of proteases deep within the body. ...

The researchers have used this approach to demonstrate diagnostic sensors for lung, ovarian, and colon cancers. ..."

From the abstract:

"Proteases, enzymes that play critical roles in health and disease, exert their function through the cleavage of peptide bonds. Identifying substrates that are efficiently and selectively cleaved by target proteases is essential for studying protease activity and for harnessing it in protease-activated diagnostics and therapeutics.

However, the vast design space of possible substrates (c.a. 2010 amino acid combinations for a 10-mer peptide) and the limited accessibility of high-throughput activity profiling tools hinder the speed and success of substrate design.

We present CleaveNet, an end-to-end AI pipeline for the design of protease substrates. Applied to matrix metalloproteinases, CleaveNet enhances the scale, tunability, and efficiency of substrate design. CleaveNet generates peptide substrates that exhibit sound biophysical properties and capture not only well-established but also previously-uncharacterized cleavage motifs.

To control substrate design, CleaveNet incorporates a conditioning tag that steers peptide generation towards desired cleavage profiles, enabling targeted design of efficient and selective substrates. CleaveNet-generated substrates were validated experimentally through a large-scale in vitro screen, even in the challenging case of designing highly selective substrates for MMP13. We envision that CleaveNet will accelerate our ability to study and capitalize on protease activity, paving the way for in silico design tools across enzyme classes."

AI-generated sensors open new paths for early cancer detection | MIT News | Massachusetts Institute of Technology "Nanoparticles coated with molecular sensors could be used to develop at-home tests for many types of cancer."

Deep learning guided design of protease substrates (open access)

Fig. 1: A deep learning approach for protease substrate design.

The Lancet: Microbubble-Enhanced Transcranial Focused Ultrasound

Recommendable!

Ultrasound probe maps real-time blood flow across entire organs in 4D

Amazing stuff! Good news!

"Microcirculation – the flow of blood through the smallest vessels – is responsible for distributing oxygen and nutrients to tissues and organs throughout the body. Mapping this flow at the whole-organ scale could enhance our understanding of the circulatory system and improve diagnosis of vascular disorders. With this aim, researchers at the Institute Physics for Medicine Paris ... have combined 3D ultrasound localization microscopy (ULM) with a multi-lens array method to image blood flow dynamics in entire organs with micrometric resolution, reporting their findings in Nature Communications. ..."

"Researchers develop an ultrasound probe capable of visualising an entire organ in 4D

For the first time, whole-organ blood flow dynamics can be mapped at micrometric resolution using an ultrasound multi-lens probe — demonstrated in vivo in the heart, kidney, and liver.

Developed at Physics for Medicine Paris, this technology enables real-time 4D imaging of vascular dynamics across entire organs. ..."

From the abstract:

"Mapping microcirculation at the whole-organ scale in 3D is crucial for understanding vascular pathologies and improving diagnostics. Although 3D ultrasound localization microscopy (ULM) enables microscopic resolution by localizing intravenously injected microbubbles in small animal models, visualizing entire organs in large animals or humans remains challenging due to limited field of view, low sensitivity, and probe technological complexity.

Here, we demonstrate how a multi-lens array method overcomes these limitations. Combined with 3D ULM, it maps and quantifies large vascular volumes (up to 120 × 100 × 82 mm³) at high spatial resolution (125–200 µm) with a volumetric acquisition rate of 312 Hz, using low-cost technology.

This approach enables deeper insights into hemodynamics from large vessels to pre-capillary arterioles, by providing vast and rich datasets of whole-organ vascularization.

It could also facilitate diagnosis of microcirculation disorders and monitoring of small-vessel disease treatments by addressing key limitations of current imaging modalities."

Ultrasound probe maps real-time blood flow across entire organs – Physics World

Major breakthrough in medical imaging just published in Nature Communications (original news release)

Multi-lens ultrasound arrays enable large scale three-dimensional micro-vascularization characterization over whole organs (open access)

Fig. 1: Definition, simulation study of the multi-lens array probe and comparison to conventional approaches.

Fig. 3: 3D Coronary mapping of an isolated porcine heart.

Schizophrenia is linked to iron and myelin deficits in the brain, implicate oligodendrocyte dysfunction

Good news!

"... Using these experimental techniques, the researchers' results suggested iron and myelin anomalies that affected specific regions in the brains of individuals diagnosed with schizophrenia, including the caudate, putamen, and globus pallidus. Their findings are aligned with those of some earlier studies and could help to paint a clearer picture of disease pathophysiology. ...

"This was most significant in regions rich in oligodendrocytes. As oligodendrocytes utilize iron to synthesize myelin, this links oligodendrocyte dysfunction to schizophrenia, highlighting the mechanism underlying this as an important research area." ..."

From the abstract:

"Iron—the most abundant magnetic brain substance—is essential for many biological processes, including dopamine and myelin synthesis.

Quantitative susceptibility mapping (QSM) MRI has recently linked altered subcortical magnetic susceptibility (χ) to schizophrenia. Since χ is increased by iron and decreased by myelin, abnormal levels of either could underlie these QSM differences.

In white matter tracts, magnetic susceptibility anisotropy (δχ) serves as a myelin-specific marker that is insensitive to iron content. To clarify the origin of case-control χ differences, we employed QSM in 85 individuals with schizophrenia, from first-episode mental health teams, and 86 healthy controls.

A subset also underwent diffusion tensor imaging (DTI) to calculate subcortical tissue mean diffusivity, which inversely correlates with myelin concentration and fractional anisotropy.

White matter δχ was calculated by combining QSM and DTI. Schizophrenia was associated with lower subcortical χ (d = −0.36, p = 0.023). This was significant in the caudate nucleus (d = −0.37, p = 0.037), putamen (d = −0.36, p = 0.037), globus pallidus (d = −0.57, p = 0.001), and SN-VTA (as previously reported).

Additionally, schizophrenia was linked to higher subcortical mean diffusivity (d = 0.44, p = 0.018), and lower white matter δχ (d = −0.37, p = 0.047). These findings suggest that both subcortical iron and brain myelin levels are lower in schizophrenia.

By comparing our voxelwise χ maps with postmortem gene expression data, we reveal that regions with lower subcortical χ in schizophrenia are enriched for oligodendrocyte-related genes (p < 0.001). As oligodendrocytes are both the most iron-rich brain cells and essential for myelin synthesis, our results implicate oligodendrocyte dysfunction in schizophrenia pathophysiology."

Schizophrenia is linked to iron and myelin deficits in the brain, neuroimaging study finds

The role of low subcortical iron, white matter myelin, and oligodendrocytes in schizophrenia: a quantitative susceptibility mapping and diffusion tensor imaging study

Fig. 1: Overview of methods testing the two competing hypotheses.

Fig. 2: Subcortical magnetic susceptibility (χ) t-score map on the T1-weighted Montreal Neurological Imaging (MNI) template at the annotated slice level and orientation.

Exact Sciences launches Cancerguard, a liquid biopsy cancer test for 50 different cancers sold directly to consumers

Good news, but the price tag is $689!

"Consumers can now purchase Exact Sciences’ liquid biopsy early detection test, Cancerguard, and get an indication of whether they might have one of 50 different cancers included in the test’s analysis. ..."

Exact Sciences launches Cancerguard, a liquid biopsy cancer test | STAT "Blood test starts new era, company says, though there are questions about data"

Exact Sciences Launches Cancerguard™, First-of-Its-Kind Multi-Cancer Early Detection Blood Test (original news release)

New scanning technique reveals ADHD brain differences

Good news!

"A new study significantly strengthens the case that attention-deficit/hyperactivity disorder (ADHD) brains are structurally unique, thanks to a new scanning technique known as the traveling-subject method. ...

A team of ... scientists ... has corrected the inconsistencies in brain scans of ADHD individuals, where mixed results from magnetic resonance imaging (MRI) studies left researchers unable to say for certain whether neurodivergency could be identified in the lab. Some studies reported smaller gray matter volumes in children with ADHD compared to those without, while others showed no difference or even larger volumes. With some irony, it's been a gray area for diagnostics and research. ..."

From the abstract:

"Brain imaging studies for attention-deficit/hyperactivity disorder (ADHD) have not always yielded consistent findings, potentially owing to measurement bias in magnetic resonance imaging (MRI) scanners.

This study aimed to elucidate the structural brain characteristics in children with ADHD by addressing measurement bias in multi-site MRI data using the harmonization method, traveling-subject (TS) approach.

The MRI data of 14 traveling subjects, 178 typically developing (TD) children, and 116 children with ADHD were collected from multiple sites. The TS method and ComBat were used to correct for measurement bias.

Gray matter volumes were estimated using FreeSurfer, and the ADHD and TD groups were compared using mixed-effect models. Compared to raw data, the TS method significantly reduced measurement bias while maintaining sampling bias. In contrast, ComBat effectively reduced measurement bias but also significantly decreased sampling bias. 

TS-corrected data showed decreased brain volumes in the frontotemporal regions in the ADHD group compared to the TD group. Specifically, significant volumetric reductions were found in the right middle temporal gyrus in children with ADHD (TS-corrected data: β = −0.255, FDR [family discovery rate] p = 0.001).

These results demonstrate that the TS method effectively reduces measurement bias across MRI scanners, ensuring reliable findings in multi-site studies. The observed frontotemporal volume reductions in ADHD, especially in the right middle temporal gyrus, highlight the reliability of findings obtained with TS correction."

New scanning method reveals ADHD brain differences

Novel Accurate Approach Improves Understanding of Brain Structure in Children with ADHD (original news release) "Magnetic resonance imaging often yields inconsistent results when assessing the brain’s structural characteristics in children with attention deficit/hyperactivity disorder (ADHD). To address this, scientists from Japan have used a harmonization method called traveling-subject (TS) to reduce measurement bias in brain imaging datasets from multiple sites. The TS harmonized datasets showed significant reductions in measurement bias and revealed apparent volumetric changes in specific brain regions, indicating promise for developing a more robust diagnosis for ADHD."

Brain structure characteristics in children with attention-deficit/hyperactivity disorder elucidated using traveling-subject harmonization (open access)

Fig. 3: The brain regions with significant differences between the ADHD and TD groups using multiple methods of correction in the mixed-effects model.

Can you smell Parkinson's disease? Yes, after 38-53 weeks of training

Maybe you need to train for a long time if you don't have a dog's nose! 😊

We need machine learning & AI to train artificial noses!

"[two dogs] reliably identified Parkinson’s disease in patients based on their odor, per a new study in The Journal of Parkinson's Disease. ..."

From the abstract:

"A definitive diagnostic test for Parkinson's disease (PD) remains elusive, so identification of potential biomarkers can facilitate diagnosis and early intervention. Two dogs were trained to distinguish between dry skin swabs obtained from people with Parkinson's (PwP) and control participants.

After 38–53 weeks of training on 205 samples, the dogs were tested in a double-blind trial using 60 control and 40 target (drug-naïve PwP) samples.

They each showed high sensitivity (70% and 80%) and specificity (90% and 98%). This supports previous findings that dogs can be trained to reliably detect the odor of PD."

Global Health NOW: Accelerating Alzheimer’s Research; Replacing Aid With ‘Sin Taxes’; and Molar Express

Dogs can detect Parkinson’s years before symptoms—with 98% accuracy

Trained dogs can detect the odor of Parkinson's disease (open access)