Roundup: AI Tool in NIH-Funded Study May Advance Routine Medical Scans; and More News

NIH: AI Tool Could Help Doctors Detect Disease Earlier from CT Scans

A new artificial intelligence (AI) tool under research funded by the National Institutes of Health (NIH) may soon help doctors get more information from routine medical scans — possibly even predicting certain diseases years before symptoms appear.

Researchers recently developed a machine-learning model called Merlin that analyzes CT scans, a common type of medical imaging. A CT (computed tomography) scan uses multiple X-ray images taken from different angles to create detailed pictures of the inside of the body, according to a news release from the NIH. Doctors often rely on these scans to look for injuries, tumors, infections, and other health problems.

Traditionally, a radiologist—a doctor trained to interpret medical images—reviews the scan and writes a report explaining what they see. However, that process can take time and may require additional tests before a diagnosis is confirmed. With a growing shortage of physicians in the United States, researchers are exploring ways to help doctors review scans more efficiently.

The new AI model could help streamline that process.

A Powerful New Type of Medical AI

Merlin belongs to a new category of AI systems called “foundation models.” These models are trained using extremely large datasets so they can perform many different tasks instead of just one specialized job.

To train Merlin, researchers used more than 15,000 abdominal CT scans along with their accompanying radiology reports and nearly one million medical diagnosis codes from Stanford University. A diagnosis code is a standardized label doctors use to classify diseases and medical conditions in health records.

By studying both the images and written reports, Merlin learned to connect patterns seen in scans with specific health conditions.

Researchers then tested the system using over 50,000 CT scans from four hospitals that the AI had never seen before. The goal was to see how well Merlin could interpret the scans compared with human-generated medical conclusions.

Predicting Disease Risk Years in Advance

The results were promising, the NIH said. Across hundreds of medical conditions, Merlin correctly predicted which scan was more likely to be linked to a particular diagnosis more than 81 percent of the time, outperforming several other AI systems designed specifically for that task.

The model also showed potential in predicting future disease risk. When researchers looked at CT scans from people who appeared healthy at the time, Merlin could identify those who were more likely to develop conditions such as diabetes, osteoporosis (bone thinning), or heart disease within the next five years about 75 percent of the time.

Scientists believe the AI may be spotting subtle warning signs—sometimes called biomarkers, or measurable indicators of disease—that are difficult for the human eye to detect.

Gut Bacteria May Influence Cholesterol Levels, New Study Suggests

Scientists are learning that the bacteria living in our digestive system may play a larger role in heart health than previously thought. A new study published in Microbiology Spectrum found that people with dyslipidemia—an unhealthy balance of fats in the blood—have different gut bacteria and microbial functions compared with people who have normal cholesterol levels.

Dyslipidemia refers to abnormal levels of lipids (fats) in the blood, including high cholesterol or triglycerides. It is a major risk factor for cardiovascular disease, which includes conditions such as heart attacks and strokes.

Researchers analyzed stool samples from 1,384 adults, including 895 people with dyslipidemia and 489 without the condition. Using a technique called shotgun metagenomic sequencing—a method that reads genetic material from all microorganisms in a sample—the scientists examined the types of bacteria in the gut and the biological functions they perform.

Differences in Gut Bacteria

The study found clear differences in the gut microbiome, the community of microorganisms living in the digestive tract.

One bacterial species, Bacteroides caccae, was more common in people with dyslipidemia. Previous research has linked this bacterium to inflammation, which may contribute to metabolic problems and changes in how the body processes fats.

In contrast, people without dyslipidemia had higher levels of bacteria — such as Coprococcus eutactus, Coprococcus catus, and Blautia obeum. These microbes help produce short-chain fatty acids (SCFAs)—beneficial molecules created when gut bacteria break down dietary fiber.

SCFAs help regulate metabolism, support gut health, and may influence cholesterol levels. Lower amounts of these beneficial bacteria could therefore disrupt lipid regulation.

Changes in ‘Metabolic Pathways’

Beyond differences in bacterial species, the study also found changes in the metabolic pathways of gut microbes in people with dyslipidemia. Metabolic pathways are the chemical processes cells use to produce energy and build molecules.

Participants with dyslipidemia had lower activity in pathways involved in glycogen production, which is important for energy storage, and peptidoglycan biosynthesis, a process involved in bacterial cell wall formation that may influence immune responses.

Researchers also observed lower levels of pseudouridine, a molecule involved in RNA metabolism, which is essential for normal cellular function.

Another finding was an increase in genes related to a pathway linked to bacterial pathogenicity — meaning the potential ability of microbes to cause disease.

Possible Link to Antibiotic Resistance

The study also examined the gut resistome, the collection of antibiotic-resistance genes found in gut bacteria.

People with dyslipidemia showed a slight increase in the tetQ gene, which provides resistance to tetracycline antibiotics. While the increase was small, it suggests a potential relationship between metabolic health and antibiotic resistance within the microbiome.

What It Means for Heart Health

The findings do not prove that gut bacteria cause dyslipidemia. However, the study strengthens evidence that the microbiome may influence lipid metabolism and cardiovascular risk.

The researchers say these results could eventually lead to microbiome-based strategies for managing metabolic diseases. Such approaches might include diet changes that support beneficial bacteria, probiotics, or other treatments designed to modify the gut microbiome.

For now, experts emphasize that maintaining heart health still relies on well-known lifestyle habits, including a balanced diet, regular physical activity, and medical treatment when necessary.

Only Half of U.S. Adults know About At-Home Colorectal Cancer Screenings, Survey Finds

Colorectal cancer is now the leading cause of cancer death for adults under 50 in the U.S.. However, a new national survey reveals significant gaps in public awareness and screening that are contributing to this health crisis.

The Colorectal Cancer Alliance, a nonprofit organization, commissioned the survey and found that while most cancers are seeing declining death rates, the rate for colorectal cancer continues to climb. In response, they are calling for action during March, which is National Colorectal Cancer Awareness Month.

Gaps in Knowledge and Care

The survey uncovered some concerning issues. One-third of all Americans, and nearly half of adults under 45, reported that a doctor had dismissed their stomach or bowel symptoms. This suggests that patient concerns, particularly from younger individuals, are not always being taken seriously.

Additionally, even when screening is recommended, more than a quarter of adults don't follow through. A major knowledge gap may be a factor: only half of Americans know that convenient at-home screening tests are an option. This lack of awareness could lead to dangerous delays in diagnosis.

Many people are also unaware of basic facts. About 52 percent don't know that colorectal cancer is highly preventable, and 37 percent are not familiar with the different screening options. These include colonoscopies, stool tests (which can be done at home), and CT scans of the colon.

Overcoming Barriers to Screening

What stops people from getting screened? The most common reasons cited were cost, fear, and a lack of knowledge about the process. However, the survey also found what motivates people to act. Over 60 percent of respondents said they would be more likely to get screened if they had a better understanding of their personal risk.

The good news is that people are becoming more willing to talk about it. A large majority of Americans said they would share their screening experiences to encourage others, viewing it as an act of love.

Know the Risks and Take Action

While the most common symptom of colorectal cancer is no symptom at all, it's important to be aware of potential warning signs. These can include changes in bowel habits, rectal bleeding, abdominal pain, fatigue, or unexplained weight loss.

Screening is recommended to begin at age 45. However, you should talk to your doctor about starting sooner if you have risk factors. These include a family history of the disease, certain genetic conditions, or chronic inflammatory bowel diseases like Crohn’s disease.

Experts emphasize that this cancer is often curable when detected early. Starting conversations with your doctor, taking symptoms seriously, and getting screened on time are the most effective ways to save lives.

Learn more about colorectal screenings.