Mayo Clinic uses AI to reduce miss rate of precancerous polyps in colorectal cancer screening

Artificial intelligence reduced by twofold the rate at which precancerous polyps were missed in colorectal cancer screening, reported a team of international researchers led by Mayo Clinic. The study is published in Gastroenterology.

Most colon polyps are harmless, but some overtime develop into colon or rectal cancer, which can be fatal if found in its later stages. Colorectal cancer is the second most deadly cancer in the world, with an estimated 1.9 million cases and 916,000 deaths worldwide in 2020, according to the World Health Organization. A colonoscopy is an exam used to detect changes or abnormalities in the large intestine (colon) and rectum.

Between February 2020 and May 2021, 230 study participants each underwent two back-to-back colonoscopies on the same day at eight hospitals and community clinics in the U.S., U.K., and Italy. One colonoscopy used AI; the other, a standard colonoscopy, did not.

The rate at which precancerous colorectal polyps are missed has been estimated to be 25%. In this study, the miss rate was 15.5% in the group that had the AI colonoscopy first. The miss rate was 32.4 % in the group that had standard colonoscopy first. The AI colonoscopy detected more polyps that were smaller, flatter, and in the proximal and distal colon.

 "Colorectal cancer is almost entirely preventable with proper screening," says senior author Michael B. Wallace, M.D., division chair of gastroenterology and hepatology at Sheikh Shakhbout Medical City in Abu Dhabi, United Arab Emirates, and the Fred C. Andersen Professor of Medicine at Mayo Clinic in Jacksonville, Fla. "Using artificial intelligence to detect colon polyps and potentially save lives is welcome and promising news for patients and their families."

In addition, false-negative rates were 6.8% in the group that had the AI colonoscopy first. It was 29.6% in the group that had standard colonoscopy first. A false-negative result indicates that you do not have a particular condition, when in fact you do.

Thompson shows how changes in vegetation shaped global temperatures over last 10,000 years

Follow the pollen. Records from past plant life tell the real story of global temperatures, according to research from a climate scientist at Washington University in St. Louis, Missouri.

Warmer temperatures brought plants — and then came even warmer temperatures, according to new model simulations published April 15 in Science Advances.

Alexander Thompson, a postdoctoral research associate in earth and planetary sciences in Arts & Sciences, updated simulations from an important climate model to reflect the role of changing vegetation as a key driver of global temperatures over the last 10,000 years. Thompson

Thompson had long been troubled by a problem with models of Earth’s atmospheric temperatures since the last ice age. Too many of these simulations showed temperatures warming consistently over time.

But climate proxy records tell a different story. Many of those sources indicate a marked peak in global temperatures that occurred between 6,000 and 9,000 years ago.

Thompson had a hunch that the models could be overlooking the role of changes in vegetation in favor of impacts from atmospheric carbon dioxide concentrations or ice cover.

“Pollen records suggest a large expansion of vegetation during that time,” Thompson said.

“But previous models only show a limited amount of vegetation growth,” he said. “So, even though some of these other simulations have included dynamic vegetation, it wasn’t nearly enough of a vegetation shift to account for what the pollen records suggest.”

In reality, the changes to vegetative cover were significant.

Early in the Holocene, the current geological epoch, the Sahara Desert in Africa grew greener than today — it was more of a grassland. Other Northern Hemisphere vegetation including the coniferous and deciduous forests in the mid-latitudes and the Arctic also thrived.

Thompson took evidence from pollen records and designed a set of experiments with a climate model known as the Community Earth System Model (CESM), one of the best-regarded models in a wide-ranging class of such models. He ran simulations to account for a range of changes in vegetation that had not been previously considered.

“Expanded vegetation during the Holocene warmed the globe by as much as 1.5 degrees Fahrenheit,” Thompson said. “Our new simulations align closely with paleoclimate proxies. So this is exciting that we can point to Northern Hemisphere vegetation as one potential factor that allows us to resolve the controversial Holocene temperature conundrum.”

Understanding the scale and timing of temperature change throughout the Holocene is important because it is a period of recent history, geologically speaking. The rise of human agriculture and civilization occurred during this time, so many scientists and historians from different disciplines are interested in understanding how early and mid-Holocene climates differed from the present day.

Thompson conducted this research work as a graduate student at the University of Michigan. He is continuing his research in the laboratory of climate scientist Bronwen Konecky at Washington University.

“Overall, our study emphasizes that accounting for vegetation change is critical,” Thompson said. “Projections for future climate change are more likely to produce more trustworthy predictions if they include changes in vegetation.”

Birmingham University’s BlueBEAR predicts pollution from cooking emissions

Organic aerosols – such as those released in cooking – may stay in the atmosphere for several days, because of nanostructures formed by fatty acids as they are released into the air.

By identifying the processes which control how these aerosols are transformed in the atmosphere, scientists will be able to better understand and predict their impact on the environment and the climate.

Experts at the Universities of Birmingham and Bath have used instruments at the Diamond Light Source and the Central Laser Facility, both based at the Harwell Campus in Oxford, to probe the behavior of thin films of oleic acid – and unsaturated fatty acid commonly released when cooking.

In the study, published in Atmospheric Chemistry and Physics, they were able to analyze the particular molecular properties that control how rapidly aerosol emissions can be broken down in the atmosphere.

Then, using a theoretical model combined with experimental data the team was able to predict the number of times aerosols generated from cooking may hang around in the environment.

These types of aerosols have long been associated with poor air quality in urban areas, but their impact on human-made climate change is hard to gauge. That’s because of the diverse range of molecules found within aerosols, and their varying interactions with the environment.

By identifying the nanostructure of molecules emitted during cooking that slows down the break-up of organic aerosols, it becomes possible to model how they are transported and dispersed into the atmosphere.

Lead author Dr. Christian Pfrang, of the University of Birmingham’s School of Geography, Earth and Environmental Sciences, said: “Cooking aerosols account for up to 10 percent of particulate matter (PM) emissions in the UK. Finding accurate ways to predict their behavior will give us much more precise ways to also assess their contribution to climate change.”

Co-author Dr. Adam Squires, of the University of Bath, said: “We’re increasingly finding out how molecules like these fatty acids from cooking can organize themselves into bilayers and other regular shapes and stacks within aerosol droplets that float in the air, and how this completely changes how fast they degrade, how long they persist in the atmosphere, and how they affect pollution and weather.”