The hospital-developed system analyses medical histories, laboratory results and imaging needs in real time, cutting admission delays and flagging hidden emergencies including a severe infection, suspected child abuse and a life-threatening potassium deficiency.

A new artificial intelligence system operating at Rambam Health Care Campus analyses the data of every patient in its emergency departments in real time, suggests possible diagnoses and treatments and identifies complex or unusual cases within seconds.

The system, named Shaked, has been in use at the Haifa hospital for the past six months. According to its developers, it has already shortened the time required to admit patients by one hour and reduced waiting time for specialist consultations in the adult and paediatric emergency departments by 25 minutes.

“This is the first AI system involved at highly important decision-making points in the emergency department,” said Prof. Shahar Shelly, director of Rambam’s Neurology Department and head of the artificial intelligence laboratory at the Technion’s medical faculty, who initiated and developed the system.

“We believe waiting times will be shortened significantly in the near future. The results have already amazed us.”

Artificial intelligence has assumed a growing role in health care in recent years. Hospitals and health funds are developing and adopting systems that analyze medical information, support clinical decisions and assist physicians in treating patients.

At the same time, the health system remains concerned about doctors independently using external AI products such as ChatGPT, Gemini or Claude, which may not operate within secure medical databases or under institutional oversight.

Shaked supports medical teams throughout the emergency-room process, from admission, diagnosis and treatment to the decision to hospitalize or discharge a patient.

The system consolidates each patient’s medical history, compares it with similar cases, receives and analyzes laboratory results and prioritizes imaging examinations such as CT scans in real time for all patients currently in the emergency department.

“The system sends alerts to doctors’ computers about specific patients whose condition could deteriorate within a short period, even when their clinical indicators appeared reasonable during triage,” Shelly said.

It was gum inflammation, not a stroke

Shelly described a recent case in which the system identified severe gum inflammation in a patient whose symptoms initially appeared consistent with a stroke.

“A young woman arrived with neurological symptoms characteristic of a stroke, including paralysis on one side of her face that had developed very rapidly,” he said.

“Because of the suspected stroke, we were preparing to activate the stroke protocol, including an urgent CT scan. Although neither the patient nor the doctor made the connection, the system alerted us that two weeks earlier she had undergone a complex dental procedure that could become complicated, create a localised infection in the jaw and place pressure on the facial nerve.”

(Photo: Elad Gershgoren)

An oral and maxillofacial specialist was summoned to the emergency department and determined that the woman was suffering from acute gum inflammation rather than a stroke.

Another particularly sensitive case occurred in Rambam’s pediatric emergency department after a 4-month-old girl arrived with a fractured thigh.

“The parents suggested that their 3-year-old child may have fallen on her, but the system retrieved an emergency-room visit from a month earlier involving another injury,” said Dr. Idit Pasternak, director of the pediatric emergency department at Rambam’s Ruth Rappaport Children’s Hospital.

“Analyzing the case and combining the current fracture with the previous injury immediately raised suspicion of child abuse. Without the artificial intelligence, there would certainly have been a consultation with a specialist and a discussion involving the entire team, but in this case the Shaked system raised a red flag over an injury characteristic of abuse.”

Because doctors believed the infant’s life could be at risk, they reported the case to the appropriate authorities without further delay.

A third case involved a 13-year-old boy who complained of muscle pain. According to Pasternak, the system analyzed his blood tests and alerted doctors to the possibility of severe hypokalemia combined with metabolic acidosis, a rare and life-threatening condition caused by an acute potassium deficiency.

“Without immediate treatment, acute heart and respiratory failure can develop within a short time, leading to cardiac arrest and death,” she said. “The system saved valuable time. The doctors immediately confirmed the diagnosis and began lifesaving treatment.”

‘ER lines will become a thing of the past’

Shelly said one of the principal causes of extended emergency-room stays is the wait for medical imaging, which can sometimes last several hours and create a risk of its own.

The system evaluates the patient’s symptoms, medical background, current clinical indicators and laboratory findings, compares them with existing medical literature and then recommends whether imaging is necessary and which cases should receive the highest priority.

Physicians review every recommendation and retain responsibility for the final decision.

“The system identifies specific red flags for conditions such as internal bleeding, stroke, pulmonary embolism and others,” Shelly said. “Such a case will be placed at the top of the list and marked as urgent and requiring immediate action.”

In the near future, the system is also expected to help prepare summaries of findings based on radiological interpretations.

Another tool incorporated into Shaked is a chat interface. Instead of opening dozens of previous patient visits to search for a specific piece of information, a physician can question the system about the patient’s file and medical history.

The tool operates using databases developed over decades within the hospital.

“Our AI does not have to go outside the system to mine information, but relies on the most up-to-date medical databases,” Shelly said. “This makes the teams’ work more efficient in terms of both time and precision.”

“Since we began operating the AI as a quiet adviser in the background, we have spared patients unnecessary CT scans and identified patients who need hospitalization more quickly,” he added.

“The system can refresh medical information in real time and changes with every result that arrives. That level of dynamism is the first of its kind.”

The system was developed with the support of Rambam’s management and through collaboration involving Pasternak, Emergency Department Director Dr. Alexander Strizhevsky, nursing teams and the hospital’s Digital and Information Technologies Division, headed by Hagar Trau, together with hospital developers Adi Ahituv and Keren Miron.

Dr. Michal Mekel, director of Rambam, said the hospital had placed innovation, and particularly the development of artificial intelligence, high on its list of priorities.

“We are allocating extensive resources to provide our patients with the best and most advanced treatment in the world,” she said. “We are entering a new era in medicine.”

“We believe that within a short time, emergency-room lines will become a thing of the past, and we are moving toward a future of precise and rapid medicine that reduces cognitive bias.”

In that future, she said, technology will serve as a lifesaving digital safety net, supporting doctors in their decision-making and freeing them to concentrate on what matters most: their human connection with the patient.

Haifa-based NurExone is preparing for human trials of an experimental therapy that uses engineered exosomes to restore function after spinal cord and optic nerve injuries.

“Our goal is to bring to market a therapy that could help people regain function after spinal cord injuries or restore vision following damage to the optic nerve,” says Dr. Lior Shaltiel, CEO of Haifa-based startup NurExone, which was founded on research conducted at the Technion and Tel Aviv University.

It is an ambitious goal. NurExone is attempting to tackle one of medicine’s most difficult challenges: repairing the central nervous system, where damaged nerves have long been considered largely incapable of regeneration. While doctors have learned to treat many of the complications associated with spinal cord and optic nerve injuries, they have had little success in reversing the damage itself. Patients with severe spinal cord injuries typically do not regain the ability to walk, while damage to the optic nerve often leads to permanent vision loss.

“The moment a spinal cord injury occurs – whether in a car accident or from a partial or complete severing of the cord, for example as the result of a gunshot wound – a secondary injury begins to spread,” explains Dr. Jacob Blumenthal, head of the BrainTech department at the Ehrlich Group, who was involved in the early research conducted at the Technion. “Cells in the affected area begin to die, effectively interrupting communication between the brain and the parts of the body it controls.”

Lior Shaltiel. (Photo: Michal Revivo)

Unlike the peripheral nervous system, where damaged nerves can regenerate, the central nervous system contains biological mechanisms that actively prevent damaged nerve fibers from regrowing.

“If I cut my hand and damage a peripheral nerve, it can regenerate,” explains Yoram Drucker, serial entrepreneur and chairman of NurExone. “But the central nervous system contains molecules that block the rehabilitation of damaged nerves.”

The company’s central question became straightforward: can those biological brakes be temporarily switched off?

Blocking the blocker

One of the key molecules responsible is a protein known as PTEN. Under normal circumstances, PTEN plays a critical role in protecting the body by preventing uncontrolled cell growth, helping suppress cancer. Following a nerve injury, however, the same protein also inhibits the regeneration of damaged nerve cells.

“We know there are inhibitors that prevent nerve growth,” says Drucker. “So the idea was simple: let’s inhibit the inhibitor. If we block the blocker, we allow nerves to regrow.”

The challenge is delivering such treatment directly into damaged nervous tissue without invasive surgery.

NurExone’s approach relies on exosomes, microscopic particles naturally secreted by cells to communicate with one another throughout the body.

“We use exosomes because they can do three things,” Drucker explains. “They naturally migrate toward injured tissue, they help reduce inflammation and promote healing, and they serve as delivery vehicles. We load them with the therapeutic molecule we need, they enter the damaged tissue, release their cargo, and that’s what produces the therapeutic effect.”

The therapeutic cargo consists of a genetically engineered small interfering RNA (siRNA) molecule that temporarily suppresses PTEN activity.

“By temporarily switching off PTEN, we remove the brake on nerve regeneration,” says Shaltiel. “Importantly, we do this only for a short period – roughly two to three weeks after the injury.”

From engineered tissue to exosomes

NurExone’s origins trace back to research led by Prof. Shulamit Levenberg of the Technion and Prof. Daniel Offen of Tel Aviv University, who later became co-founders of the company.

Initially, Levenberg’s team focused on implanting engineered tissue directly into damaged spinal cords.

“We suddenly saw the rats walking,” she recalls. “We expected to see some improvement, but instead they got up, stood on their own, bore weight and walked almost normally. It was an incredibly exciting moment.”

The next challenge was translating those results into a treatment suitable for people.

Performing spinal cord transplants through invasive surgery would be difficult to apply widely, prompting the researchers to ask whether they could use only the biologically active components released by the cells rather than transplanting the cells themselves.

The answer was exosomes.

“We administered only the exosomes through the nose, and again the rats recovered,” says Levenberg. “The animals began walking again. At that point we realized this could become a practical, non-invasive clinical therapy.”

Manufacturing billions of biological delivery vehicles

Producing exosomes at clinical scale presents another technological challenge.

“We grow the cells in bioreactors at 37 degrees Celsius,” explains Dr. Tali Kizhner, NurExone’s vice president of research and development. “The cells grow on tiny carriers inside the bioreactor and continuously secrete exosomes, allowing us to scale production.”

Founded in 2020, NurExone has raised approximately $20 million and is publicly traded on the Toronto Stock Exchange.

Part of the company’s strategy is to pursue regulatory pathways for orphan diseases, which can provide development incentives and potentially accelerate regulatory review.

“This is an extremely active field because there are still no effective treatments,” says Blumenthal. “The first company to successfully commercialize a therapy in this area could gain a significant competitive advantage. Strong patent protection also substantially increases the company’s value.”

The commercial opportunity could also be considerable. Drucker notes that roughly 50,000 new spinal cord injury patients are diagnosed annually across the Western world. The company is also developing treatments targeting optic nerve damage, including glaucoma, which affects an estimated 80 million people worldwide.

NurExone remains in the advanced preclinical stage and is preparing regulatory submissions to the U.S. Food and Drug Administration. The company expects to begin its first human clinical trial in Israel and the United States around 2027, subject to regulatory approval.

But Shaltiel believes the long-term vision extends beyond hospitals.

“Eventually, we’d like to move treatment from the emergency room to the ambulance,” he says. “Someone injured in a car accident could receive the therapy immediately from a paramedic to prevent the loss of nerve cells before they even arrive at the hospital.”

Levenberg shares a similar vision.

“I imagine this becoming available in every emergency department,” she says. “A patient arrives with a spinal cord injury, receives exosome treatment immediately, and we prevent further deterioration while giving the tissue the opportunity to regenerate. If that becomes possible, we could prevent many of the lifelong consequences that spinal cord injuries cause today.”

The vision remains several years away from being tested in humans. But if NurExone’s approach proves successful in clinical trials, it could challenge one of neuroscience’s longest-standing assumptions: that damage to the central nervous system is largely irreversible.

The eye can rebuild lost corneal stem cells by turning old cells back into stem cells, a discovery that could one day help restore vision.

Researchers have found that old, fully mature cells in the cornea can become stem cells again. The cells rebuilt the eye’s stem cell supply even after every last stem cell had been destroyed.

Cells that had nearly stopped dividing pulled off the reversal and kept the tissue healthy for the rest of the animal’s life.

The cornea is the clear dome at the front of the eye, and losing its stem cells can rob a person of sight.

A repair system the body runs on its own raises the prospect that some blindness could one day be reversed without a transplant from a donor.

A cornea rebuilds itself

The cornea keeps itself clear thanks to a small, hidden workforce. Its own limbal stem cells sit in a narrow ring at the outer rim, called the limbus.

Their offspring travel inward to replace surface cells as they wear away.

A team led by Professor Ruby Shalom-Feuerstein, a stem cell biologist at the Technion-Israel Institute of Technology, set out to see what happens when that workforce is gone. 

Working with mice, the researchers stripped away the entire ring of stem cells while leaving the surrounding tissue bed untouched. Through the clear cornea, they tracked the repair.

The cornea grew back. Tracing individual cells showed the new stem cells were not survivors that had escaped the scalpel but ordinary mature corneal cells that had reversed course.

Biologists call this dedifferentiation, a process in which a specialized cell reverts to a stem cell.

A remarkable transformation 

The transformation was remarkably complete.

Within about 40 days, the reverted cells were almost indistinguishable from the cornea’s native stem cells, sharing more than 99 percent of their gene activity. Ordinary corneal cells share only 85 to 90 percent.

They also behaved like the real thing, supplying fresh surface cells and keeping the cornea clear for 6 months.

That result ran against a long-held assumption. Losing a tissue’s stem cells was thought to be irreversible, ending in scarring and disease, which is why serious repair has usually meant transplanting cells from elsewhere. 

“We were surprised to discover that the cornea can regenerate itself even after the destruction of all its stem cells,” said Shalom-Feuerstein.

The oldest cells made a comeback

Cells reversing into stem cells was not new. Earlier work had caught committed cells in fast-renewing tissues doing it.

One influential study showed differentiated cells in the airway becoming stable, working stem cells after the resident ones were removed.

What stayed unclear was the reach of that ability. In those tissues the cells that reverted were young, made only days earlier, so no one knew whether genuinely old cells held the same capacity.

The cornea gave the team a way to test it. Because a corneal cell’s distance from the rim tracks its age, the researchers could pick out the eye’s oldest cells.

These were mature cells sitting at the center that had spent about four months drifting inward and had all but retired from dividing. Even these old cells made a comeback. 

When moved onto a stripped corneal rim, they switched their stem cell programs back on, rebuilt both the quiet reserve and the active working pool, and kept the cornea clear for nearly a year.

An immune cell trigger

Something had to tell those cells to change. The team traced the signal to macrophages, the immune cells that rush to any injury to clear debris and bacteria. 

They flooded the stripped rim within a day or two, and cutting their numbers by about three-quarters left the stem cells largely unable to return.

The macrophages did more than clean up. As they gathered they released signaling molecules.

Two of them, a growth factor called IGF1 and an immune messenger called CCL2, did most of the work of nudging mature cells back toward a stem cell state.

Blocking the pair stalled the repair, and adding them back restored it.

The reversal has its limits

The reversal has firm limits. It only worked when the tissue bed that houses the stem cells came through the injury intact, and it stayed within the cornea’s own family of cells.

When cells from the conjunctiva, the membrane that covers the white of the eye, moved in to repopulate the cornea’s outer rim instead, they could not transform into corneal stem cells.

The surface clouded and filled with blood vessels – the signature of blindness caused by stem cell loss.

The eye is not alone. In the skin, researchers have found that the stem cells responsible for hair color can shift in and out of a stem cell state as their surroundings change. 

The finding suggests this kind of cellular flexibility is a normal part of tissue maintenance, not just an emergency response to injury.

A path toward restoring vision

Whether human corneas can do the same is not yet settled. The experiments ran almost entirely in mice, and the team could not directly test reversal in human corneal cells.

Working with donated human tissue, they found the same macrophage molecules helped cultured human stem cells keep their identity – a sign the pathway may carry over.

The stakes are clearest for people who have lost the cornea’s stem cells to burns, infection, or disease. Their vision is already gone or fading.

The study shows that a mature cell, even a very old one, can revert to a fully functional stem cell and remain that way for life, provided the surrounding tissue and immune cells supply the right signals. 

What remains unknown is how to trigger that process on demand and whether other organs possess the same hidden regenerative ability.

“The next challenge is learning how to control it and how to use it for regenerative medicine,” said Shalom-Feuerstein.

The study is published in the journal Nature Communications.

Technion-led study combines deep learning and mathematical modelling to produce dynamic MRI images at up to one frame per second.

A group of researchers from the Technion in Israel and the United States has reported a breakthrough in MRI scanning that could significantly improve breast cancer diagnosis, according to a paper published in Nature Communications.

The researchers developed a new method, called ELITE, that accelerates and enhances MRI scans used in breast cancer imaging, a disease diagnosed in approximately 2.3 million people each year, most of them women. The approach combines artificial intelligence with advanced mathematical modelling to enable dynamic MRI imaging at what the researchers describe as unprecedented speed and accuracy.

 Dr. Eddy Solomon
Dr. Eddy Solomon. (Leo DeLuca)

The international study brings together expertise in engineering, MRI physics, artificial intelligence and clinical radiology.

Dr. Eddy Solomon of the Technion’s Faculty of Biomedical Engineering, the study’s lead author, said the research focuses on dynamic MRI, a key tool in breast cancer diagnosis. Dynamic MRI is primarily used for screening high-risk populations and is characterised by high sensitivity, with more than 90% accuracy, compared with roughly 50-60% for ultrasound and mammography combined.

However, MRI technology has long faced a fundamental limitation: producing highly detailed images requires relatively long scan times, making it difficult to track the movement of contrast material through tissue in real time. Traditional MRI systems typically generate one image every one to two minutes at best, limiting the ability to capture the rapid dynamics of contrast agents.

Dr. Solomon and his colleagues sought to bridge this gap by combining mathematical modeling that identifies structural and functional patterns in different tissues with a deep neural network (ResNet) trained to remove noise and distortions. The system also reconstructs missing information from undersampled measurements.

The result, according to the researchers, is the ability to generate one image per second.

The improved temporal resolution allows clinicians to track the movement of contrast agents almost continuously. This, the researchers say, could improve the detection of small tumours, help distinguish more accurately between benign and malignant growths, and better characterise tumour biology, including blood flow and vascular permeability.

In a study involving 54 patients, the researchers reported improved tumour visibility compared with existing methods, higher image quality, and strong diagnostic sensitivity.

They also said that shorter scan times could increase the number of patients that can be scanned using a given MRI system, potentially improving access to imaging services.

The findings are presented as a step toward faster and more precise MRI-based cancer diagnostics, though further validation and clinical deployment would be required before broader adoption.

Concrete shapes the world you live in, but its environmental toll is massive. Imagine, though, if the buildings around you could grow and breathe, helping to heal the planet. This isn’t science fiction. It’s the vision of CyanoCement. Developed by researchers from the Technion, Israel Institute of Technology, this innovative biocement uses ancient microbes to redefine how we think about construction materials.

At the heart of CyanoCement are cyanobacteria, tiny photosynthetic organisms responsible for Earth’s first oxygen-rich atmosphere. By leveraging these extraordinary capabilities, the team—Perla Armaly, Yuval Berger, Lubov Iliassafov, Keren Rosenblau, Yechezkel Kashi, and Shany Barath—crafted a process where these microbes bind minerals and precipitate calcium carbonate, creating a solid without high emissions.

Petri dish with cyanobacteria cultures on a laboratory table.

Innovative Design Meets Environmental Responsibility

This biocement doesn’t just end its environmental work once installed. It continues to capture carbon dioxide from the air, actively working against the problem of atmospheric carbon. Unlike conventional methods, CyanoCement turns construction into part of the solution.

Construction worker applying biocement to a brick wall surface.

The material is designed for facades, interior panels, and decorative structures. By focusing on non-load-bearing elements, the team keeps the project’s ambitions grounded, managing expectations with scientific precision.

Close-up of biocement texture with visible green cyanobacteria.

Visible Green: A Living, Breathing Material

The green hue of CyanoCement isn’t painted on. It’s the color of life—indicative of the cyanobacteria within. The design makes environmental benefits visible, offering a reassuring sign that sustainability is working, right before your eyes.

Lab technician measuring pH level of biocement solution.

This innovative project emerged from the Disrupt Design Lab at Technion, in collaboration with the Applied Genomics Lab, marking a significant crossover between architecture and biology. It’s a fusion rarely seen, yet wholly necessary for the future of sustainable design.

For a unique blend of nature and architecture, explore how the Sofia Pavilion blends urban landscapeswith natural elements.

Construction site using eco-friendly biocement blocks in foundation.

CyanoCement was honored with the Green Product Award, celebrated for its meaningful impact and robust research. It’s not just talk—this material has substance and intention.

Learn how ancient materials are making a comeback with Finnish designers crafting fashion from shipwreck timber.

Green biocement samples displayed on a laboratory workbench.

As we think about the future of architecture and sustainability, it’s time to reconsider the role of construction. CyanoCement poses a radical idea: buildings that are not only structures but contributors to the atmosphere. An idea that’s difficult to ignore once it takes root in your mind.

Researcher writing down observations of growing cyanobacteria samples.

Technion-developed technology allows patients to detect dangerous sleep disorders from home using a simple wearable device and artificial intelligence, eliminating need for costly lab tests.

Nearly 40% of the global population suffers from sleep-related disorders, according to the World Health Organization. Among the most serious, and often undiagnosed, conditions is sleep apnea, a disorder estimated to affect nearly one billion people worldwide.

Diagnosing sleep apnea traditionally requires an overnight stay in a specialised sleep laboratory, such as the facility at Ichilov Hospital in Tel Aviv. 

Patients are connected to multiple sensors that monitor brain activity, breathing patterns, heart rate, and oxygen levels throughout the night. The process is complex, expensive, and often inaccessible, with costs ranging from $1,170 to $11,700 depending on the clinic.

An Israeli startup, Sleep AI, aims to change that. Developed by researchers at the Technion, the technology uses a lightweight oximeter linked to a mobile app and powered by artificial intelligence. Patients can complete the test from home by simply wearing the device overnight while data is uploaded to the company’s cloud platform for analysis.

Within minutes, physicians receive a detailed medical report that not only evaluates sleep quality, but also maps sleep architecture, identifies signs of sleep apnea, and assesses cardiovascular risks linked to nighttime oxygen deprivation.

Unlike consumer smartwatches, Sleep AI is designed as a medical-grade diagnostic tool. In clinical testing conducted in sleep centres, the system demonstrated an overall accuracy rate of 89% for detecting sleep apnea, rising to 99% for moderate and severe cases.

The company is now pursuing international regulatory approvals with the goal of making sleep apnea diagnosis faster, cheaper, and more widely available — potentially even covered by health insurance in the future.

By moving diagnosis from the sleep lab to the home, Sleep AI hopes to make sleep health screening a routine part of modern medical care.

Researchers at the Technion have discovered how changes in genetic regulatory sequences can lead to alterations in the form and structure of animals – even when genetic regulatory systems are stable and resistant to change. The study, published in Science Advances, was led by Dr. Ella Preger-Ben Noon and Ph.D. candidate Areej Said-Ahmad from the Ruth and Bruce Rappaport Faculty of Medicine.

1. Dr. Ella Preger-Ben Noon (on the right) and Ph.D. candidate Areej Said-Ahmad
1. Dr. Ella Preger-Ben Noon (on the right) and Ph.D. candidate Areej Said-Ahmad צילום: רמי שלוש, דוברות הטכניון
Dr. Ella Preger-Ben Noon (on the right) and Ph.D. candidate Areej Said-Ahmad 
Dr. Ella Preger-Ben Noon (on the right) and Ph.D. candidate Areej Said-Ahmad

Photo Credit: Rami Shelush

The loss of morphological traits is a common phenomenon in evolution. Well-known examples include the loss of legs in snakes and the loss of eyes in cavefish. In many cases, such changes do not result from the loss of the genes responsible for these traits, but rather from changes in how those genes are regulated during development. However, many developmental genes are controlled by multiple regulatory sequences with overlapping activity, forming a stable and robust regulatory system.

This study addresses a fundamental question in biology: how do organisms change form over the course of evolution despite the presence of stable genetic regulatory systems? These systems rely on DNA sequences known as enhancers, which activate genes at precise times, levels, and locations during development. Enhancers often act redundantly, so that if one is impaired, others can compensate and maintain proper gene expression. This redundancy confers stability and resistance to change, but also raises a paradox: how do changes in gene expression still occur, leading to alterations in the shape and structure of organs?

To address this question, the researchers focused on Drosophila flies, particularly the species Drosophila sechellia, in which tiny hair-like structures (trichomes) have disappeared from the larval body during evolution. This trait is controlled by the shavenbaby gene, whose expression is regulated by multiple enhancers. Contrary to expectations that such a system would protect gene expression from change, the researchers found that four different enhancers of shavenbaby lost their activity over the course of evolution, each through a distinct mechanism.

Image: Closely related fruit flies can look quite different because of how a single gene is turned on or off. The larvae on the left have dense rows of tiny hairs, while those on the right have lost many of them. This difference comes from changes in how the shavenbaby gene works during early developmen
Image: Closely related fruit flies can look quite different because of how a single gene is turned on or off. The larvae on the left have dense rows of tiny hairs, while those on the right have lost many of them. This difference comes from changes in how the shavenbaby gene works during early developmen

Through detailed DNA sequence analysis and functional experiments, the researchers found that the loss of enhancer activity occurred via different molecular mechanisms, including deletion of essential sequences, loss of binding sites for activators and gain of repressor binding sites, acquisition of a silencer, and even the unmasking of pre-existing repression. In other words, the same evolutionary outcome – the loss of gene expression – was achieved through different molecular pathways within the same genomic region.

These findings demonstrate that the same evolutionary outcome can arise through multiple routes. The presence of multiple enhancers, while they contribute to stable gene expression, also creates points of vulnerability where mutations can reduce their activity. The study shows that stability does not necessarily act as a barrier to evolution, as there are diverse molecular ways to circumvent it. These insights are relevant to a wide range of biological systems and deepen our understanding of how variation in form and structure arises in nature.

Technion and Tel Aviv University researchers say BetaDescribe can turn protein sequences into plain-language insights, opening faster paths for drug discovery, biotech and materials research

Researchers from the Technion and Tel Aviv University have developed an artificial intelligence system that translates protein sequences into natural-language descriptions, a tool they say could help accelerate drug development, biotechnology and material design.

The system, called BetaDescribe, was presented in a paper published in the journal PNAS. It is designed to analyse protein sequences and generate detailed descriptions of their functions and characteristics.

Diagram illustrating the system’s operation

Protein analysis is central to medicine and biotechnology, but experimental characterization is often slow and costly. The researchers said BetaDescribe could help narrow the gap between the hundreds of thousands of proteins already characterised in laboratories and the billions, or even trillions, believed to exist in nature.

Unlike traditional methods that rely mainly on comparing unknown proteins to known sequences, BetaDescribe combines a generative model with verification and evaluation mechanisms. The researchers said this allows it to infer protein function even when a protein is not closely related to previously studied examples.

The system can describe functional properties, catalytic activity, roles in metabolic processes and possible binding sites relevant to medical and other uses. The team demonstrated the system by successfully describing six previously uncharacterized proteins.

The paper was led by doctoral student Edo Dotan under the joint supervision of Prof. Yonatan Belinkov of the Technion’s Henry and Marilyn Taub Faculty of Computer Science and Prof. Tal Pupko of Tel Aviv University’s School of Life Sciences.

Prof. Tal Pupko
Prof. Tal Pupko Photo: Chen Galili
Prof. Yonatan Belinkov
Prof. Yonatan Belinkov Photo: Hadas Parush
Doctoral student Edo Dotan
Doctoral student Edo Dotan Photo: Technion

Other co-authors included Prof. Eran Bacharach, Prof. Marcelo Ehrlich and doctoral student Iris Lyubman of Tel Aviv University.

The research was supported by the Israel Science Foundation.

Deciding whether to administer chemotherapy after surgery is one of the most challenging questions in early-stage breast cancer care. While chemotherapy can reduce the risk of recurrence, most patients do not benefit from it and may experience significant short- and long-term side effects. The central challenge is identifying, at the time of diagnosis, which patients are likely to benefit and which are not.

Researchers from the Technion—Israel Institute of Technology, together with collaborators from leading medical centres in the United States and Europe, have developed an artificial intelligence (AI) model that predicts both the risk of breast cancer recurrence and the likelihood that a patient will benefit from chemotherapy. The model analyses routine pathology slides taken at diagnosis, offering a fast, widely accessible alternative to costly genomic tests.

The study was recently published in The Lancet Oncology and presented at the European Society for Medical Oncology (ESMO) conference. It is the first AI model of its kind to be validated in a large, randomized clinical trial.

Addressing a global clinical need

Each year, approximately 2.3 million people worldwide are diagnosed with breast cancer, including about 300,000 in the United States and 5,000 in Israel. Today, genomic tests such as Oncotype DX are commonly used to guide chemotherapy decisions, but these tests are expensive, can take weeks to return results, and are unavailable to many patients globally. Their predictive accuracy is also limited, leading to both unnecessary chemotherapy and missed treatment opportunities.

The Technion-led AI model aims to address these limitations by using information already available in standard pathology samples.

How the model works

The system analyses high-resolution digital images of tumour tissue stained and examined as part of routine pathology. Using deep learning, it evaluates multiple regions of the tumour and its microenvironment, identifying visual patterns associated with cancer behaviour, including cell division, tissue structure, immune response, and features linked to treatment sensitivity or resistance.

“These are complex biological signals that the human eye cannot consistently quantify,” said Dr. Gil Shamai of the Technion’s Geometric Image Processing Laboratory, who led the study. “The model integrates many subtle cues to generate a score that reflects both recurrence risk and expected benefit from chemotherapy.”

Prof. Ron Kimmel, head of the laboratory in the Henry and Marilyn Taub Faculty of Computer Science, explained the concept: “Instead of testing genes, we look directly at the tissue. Just as eye color can be determined by looking at the eyes rather than analyzing DNA, our system extracts a visual signature from pathology images that informs optimal treatment.”

Clinical use and validation

Clinically, the process is straightforward. After diagnosis, the existing tissue sample is digitally scanned and securely analyzed by the AI system. Within minutes, the model produces a numerical score that supports shared decision-making between oncologist and patient.

While the system’s internal decision-making cannot be fully explained in simple rules, its performance has been rigorously validated. The researchers were granted rare access to tissue samples and clinical data from the TAILORx trial—one of the largest randomised breast cancer studies, involving more than 10,000 patients who were randomly assigned to receive chemotherapy or not.

“Using data from a randomised trial allowed us to test whether the model truly predicts benefit from chemotherapy, not just recurrence risk,” said Dr. Shamai.

According to Prof. Dvir Aran of the Technion’s Faculty of Biology, a co-leader of the study, “This is the first AI model shown to predict treatment benefit in breast cancer directly from pathology samples.”

The model was further validated on thousands of patients from hospitals in Israel, the United States, and Australia, including Carmel, Emek, and Sheba Medical Centres, demonstrating consistent performance across different populations, equipment, and health care systems.

Fast, affordable, and globally scalable

Unlike genomic tests, the AI-based assessment requires no additional tissue, laboratory processing, or waiting period. It can be performed in minutes in any pathology lab equipped with a digital scanner and internet access.

“In developing countries, where genomic testing is largely unavailable, this tool could dramatically expand access to personalised cancer care,” said Prof. Aran. “In high-income countries, it could reduce costs, shorten diagnosis time, and improve decision accuracy.”

Looking ahead

The research team is now advancing steps toward clinical implementation in Israel and preparing clinical trials in Brazil and India, where the potential impact is particularly large. The researchers are also working to further improve the model and extend it to additional treatments and cancer types where aggressive therapy decisions are made under uncertainty.

Based on these impressive results and the knowledge accumulated over years of groundbreaking research, the researchers now intend to establish a company that will develop tests making them significantly more accessible, accurate, and faster compared to those currently in use worldwide.

The study was led by Dr. Gil Shamai, Prof. Ron Kimmel, and Prof. Dvir Aran, in collaboration with oncologists and pathologists from institutions including Dana-Farber Cancer Institute, Mount Sinai Medical Center, the University of Chicago Medical Center, and IPATIMUP Medical Center in Portugal.

In addition to fatigue and increased hunger, living with constant sleep deprivation and stress has other effects, some long-term. Experts explain the risks – and how to limit the damage, or at least some of it

By now, this has become a daily challenge: how many hours of sleep can one get in a night riddled with air-raid alerts, racing to shelter and attempts at shuteye before being woken up again. And not just how many hours in total, but also how long one can sleep uninterrupted. All this comes before the real challenge – staying awake during the day, functioning as normally as possible and perhaps even forgetting – until the next siren – that this is an open-ended state of emergency. 

This reality has direct and indirect health implications, some immediate and clearly felt in the ability to function and in planning and concentration. In the longer run, this stressful reality, marked by constant alertness and sleep deprivation, could have a cumulative effect on other bodily systems, including the immune and cardiovascular systems, as well as mental health. 

“The professional term for what has been happening now is ‘sleep deprivation’ due to air-raid alerts,” says Prof. Yaron Dagan. “This deprivation harms two main things: one is cognitive – that is to say, everything related to thinking, perception, problem-solving, concentration and memory; the other is emotional – people are gloomier, less patient, and generally in a worse mood, which sometimes results in reckless decision-making.”

Dagan, director of the Institute for Sleep Medicine at Assuta Medical Centers, explains that healthy sleep is crucial for waking life, particularly for our cognitive system, “which reboots brain memory in order to clear it for the next 24 hours. This activity takes place in several areas in the brain, and without uninterrupted or adequate sleep – the processes served by sleep are impaired.” One stage of sleep, he emphasises, is crucial for emotional processing, learning and memory formation. “This stage occurs in 90-minute cycles, and with sleep deprivation it’s disrupted, affecting our thinking and behaviour when awake.”

Is there anything that can be done, considering that it is entirely unclear how long this routine will continue? Perhaps a nap here and there? “In principle, sleep is not a bank – you cannot not sleep for a week and then fill the deficit by sleeping for a week,” says Dagan. “What we recommend is what’s called a ‘combat nap’ – a planned 30-45-minute nap to replenish your batteries. Even if someone can’t doze off, simply lying down, closing one’s eyes and relaxing is enough. This is the best way to deal with this sleep deprivation. It cannot fully replace nighttime sleep, but it certainly helps you feel refreshed.” 

Proper or healthy sleep is not just a matter of quantity; uninterrupted sleep is just as important as getting enough hours. “Sleep that is too short or interrupted – both have the same effects and cause the same harm as sleep deprivation,” explains Prof. Giora Pillar, head of the sleep clinic in Clalit Health Services’ Haifa District and sleep researcher at the Technion’s Faculty of Medicine. “There have been studies on this. In one, students were allowed to sleep for eight hours, but their sleep was interrupted. The damage was found to be the same.” 

A vicious cycle

The immediate effects are not limited to fatigue and exhaustion. Along with sleep deprivation, unending stress is not only mental but also physiological, affecting many bodily systems. When a person remains alert for an extended period, high levels of stress hormones such as cortisol and adrenaline are secreted. Chronic exposure to these hormones can harm the immune system, increase inflammation and blood pressure and impair cardiovascular function. In addition, stress has been linked to sleep disorders (creating a vicious cycle) and to the worsening of chronic diseases such as asthma and diabetes, as well as to an increased risk of heart disease. Over time, this condition may erode physiological systems and cause an overall deterioration in health. 

Over the past two and a half years, with one operation following another and one air-raid siren after another, stress has become a familiar term. In general, it refers to a physical and emotional reaction to threatening or dangerous situations – not just wartime or physical danger, but also everyday pressures such as work overload, mental overload or difficulties in other aspects of life. In today’s reality, however, it’s almost impossible to isolate stress from sleep deprivation. “Stress is a mediating factor,” says Prof. Pillar. “It causes sleeplessness in itself, as well as many other complications.” 

In many respects, the symptoms of stress and sleep deprivation overlap or reinforce one another. In part, this connection is evident in eating patterns. Like stress, sleep deprivation is a risk factor. When sleep is reduced, levels of ghrelin (the hunger hormone) soar, while levels of leptin (the satiety hormone) fall. The result is increased hunger, especially for high-calorie, sugary and fatty foods. A 2004 study released by researchers from the University of Chicago demonstrated this clearly. The researchers hypothesised, based on their findings, that the body interprets sleep deprivation as a state of energy deficit – even if that’s not exactly the case.

Chronic overeating under such conditions can lead to weight gain, increased insulin resistance and a higher risk for type 2 diabetes, cardiovascular disease and other metabolic disorders. In addition, ongoing caloric excess, driven by fatigue, also hinders the body’s ability to regulate metabolism and balance energy. 

And the list of risks does not end there. According to Pillar, sleep deprivation also affects the immune system. “Sleepless patients or patients who sleep poorly, that is to say: people who suffer from chronic sleep disorders, are already suffering from irreversible complications,” he warns. “We will see higher rates of high blood pressure, more cases of metabolic syndromes, more diabetes, more obesity, more strokes and more cancer.” 

To a certain extent, these symptoms are reversible, as reality has proven. “Soldiers who sleep too little and then sleep through the weekend are not at risk in the long term,” Pillar illustrates. “Medical interns who sometimes work two 26-hour shifts a week make up for lost sleep and don’t develop long-term complications. That is to say, it’s reversible – up to a point.” 

However, given the current reality, which has already lasted more than a week and even a fortnight, the question becomes where the line lies beyond which the damage becomes irreversible, or only partly reversible. This is a crucial question. “We are already seeing patients whose diabetes is no longer balanced,” he says, “or who have high blood pressure.” 

A 2016 study published in the International Journal of Cardiology found a clear link between sleep duration and coronary heart disease. The findings indicate that people who sleep seven to eight hours per night are at low risk, with every one-hour reduction associated with an 11 percent increase in the risk of heart disease. These findings were reaffirmed last November in another study, published in BMC Cardiovascular Disorders, which indicated that people who sleep six hours or less are at almost twice the risk of dying from kidney or heart disease compared with those who sleep longer. 

An immune system out of balance

Over the past two decades, many studies have examined the link between sleep quality and immune system function. Among other findings, people who sleep less than six hours a night produce fewer antibodies after vaccination; on the morning after a sleepless night, a significant increase is seen in the production of inflammatory cytokines – proteins secreted by immune cells in response to infection or injury; and, in general, proper sleep strengthens anti-inflammatory and anti-viral reactions, while inflammatory signals from the immune system affect the structure and depth of sleep.

According to a 2019 study published in Nature Reviews Immunology, sleep deprivation increases activity in the sympathetic nervous system (responsible for the body’s response in situations of threat and danger), which in turn raises stress hormone levels and releases inflammatory cytokines. It was found that in chronic sleep disorders, the overall level of inflammation in the body increases, while antiviral responses grow weaker. 

“Sleep deprivation is documented as one of the main biological factors affecting the immune system (when not diseased),” says Prof. Cyrille Cohen, head of the laboratory of immunology and immunotherapy and dean of Bar-Ilan University’s Faculty of Life Sciences. “In principle, conditions such as stress and sleep deprivation do not weaken every component in the immune system but rather cause an imbalance in its function.” He says this may manifest in several ways. “For instance, you’re at a slightly higher risk of certain infections, mainly respiratory – and the recovery process may also be slower.” However, Cohen emphasizes that “the effect is usually mild, and varies greatly from person to person.”