Artificial intelligence takes to the sky above Canada’s forests

Canada’s forests are scattered in research labs across the nation.

In one laboratory, forest fires weave their way through computer code, ending up on screens in pixels and probabilities.

In another, algorithms are learning to “see” conifers in ones and zeroes.

Both laboratories are leveraging artificial intelligence (AI) to understand and protect Canada’s forests.

In a study published in the Journal of Unmanned Vehicle Systems, York University PhD student Sowmya Natesan and colleagues outline their development of an AI algorithm that accurately identifies conifer species from images of trees.

Construction and paper industries around the world use conifer wood. Canadian pines, firs, spruce, and cedars are big exports, critically important to the country’s economy.

A key way to ensure that Canada’s forests remain productive is to “understand the presence and status of trees” according to Udayalakshmi Vepakomma, lead scientist at FPInnovations and co-author of the article.

Improved knowledge of tree species means that the forest industry can better manage and use these natural resources.

Foresters fly unmanned aerial vehicles (UAVs) with cameras attached, taking pictures of tree canopies. They use image processing software to try and identify the trees, but often fail.

Current software requires high-quality pictures, but the quality of pictures captured by UAVs often varies. It depends on factors like camera angle, season, and lighting conditions.

The team studied forest canopies of Ontario’s Petawawa Research Forest. They collected hundreds of images over three years in different seasons, times of day, sunlight amount, and camera angles. These initial images were used to “train” the software so it could learn to correctly identify tree species.

More training data means more information for the software to learn from and develop faster pattern recognition and more accurate identification.

“Data processing is the heart of the system—this is what will lead to species identification.”

When put to the test, the AI software encountered previously unseen images of trees. It “sieved” them through various computational layers, each identifying features such as leaf outline and branch shape, curves, and colour. The software referred to its training data and then, with a degree of certainty, determined the number of trees in each image that are conifers.

The program is a good learner—when asked to decipher new tree images of varying quality, it identified conifer species with 84% accuracy.

“The importance of our work does not lay with data collection and images,” says York University geomatics engineering professor Costas Armenakis and principal investigator of the study. “Data processing is the heart of the system—this is what will lead to species identification.”

Smart AI flying in UAVs above Canadian forests are also lending a helping hand in combating forest fires, the frequency of which are expected to increase in North America over the next century because of climatic change.

A forest fire, in its early stages, may not even have a flame or it could be hidden beneath the forest canopy. But smoke appears earlier than flame and can be detected from far away. At night, however, smoke becomes invisible.

PhD student F.M. Anim Hossain and colleagues at Concordia University’s Department of Mechanical, Industrial, and Aerospace Engineering have developed an AI-based fire detection method that identifies both flame and smoke from aerial images taken by UAVs.

“You are trying to decrease detection time and increase efficiency with which fire fighters can combat fires as early as possible,” says professor Youmin M. Zhang and principal investigator of the study published in the Journal of Unmanned Vehicle Systems.

Recent fires in Australia, North America, Southern Europe, and Southeast Asia—millions of hectares burnt—give us a glimpse of the future: countless individuals evacuated, left homeless, or their lives taken.

“We want to equip UAVs with autonomous decision-making capabilities, and automatically warn fire authorities even before fires have started.”

The researchers gathered aerial forest fire images and videos from all over the world, including recent wildfires of California and Australia and those in the Canadian provinces of British Columbia and Alberta.

Much like Natesan and colleagues’ work, hundreds of images were used to train the software, and hundreds of images were used to test it.

The AI principle is also similar—evaluate each image for colour, shape, texture, or motion, for example, and then assign a probability to the final evaluation of smoke or fire. The software is highly precise in detecting both.

Importantly, the AI does not require much computational power and processes results in real-time. Onboard a UAV, this is important. It frees the UAV’s computer to do other tasks such as maintain flight control, navigate around fires, and communicate quickly to fire teams about imminent dangers.

Forest fires can quickly change from small sparks into out-of-control blazes. The ability for AI to quickly assess and evaluate its environment and relay its decision to fire authorities will be critical.

“We want to equip UAVs with autonomous decision-making capabilities, and automatically warn fire authorities even before fires have started,” Youmin says.

Original article:

Artificial intelligence takes to the sky above Canada’s forests

New study uncovers how berberine combats diabetes and obesity

A few centimeters below ground, roots of Chinese goldthread tirelessly manufacture Olympic ring-like structures made up of carbon, hydrogen, nitrogen, and chlorine into a chemical compound called berberine hydrochloride.

Berberine has been known to reduce blood sugar, improve heart health, and ameliorate obesity, but the exact mechanism has remained elusive. In a new paper published in the Canadian Journal of Physiology and Pharmacology, Anil Poudel and colleagues from Central Michigan University put the tiny chemical compound under a battery of experiments. Their paper unlocks the chemical’s potential in battling metabolic syndrome—a cluster of diseases including obesity, insulin resistance, hypertension, and cardiovascular disease.

Chinese goldthread is one of the 50 fundamental herbs of traditional Chinese medicine, and the first recorded medicinal use of berberine dates back almost 2000 years. Other plants that produce berberine include barberry, prickly poppy, and tree turmeric. The new research could potentially help scientists develop novel therapeutic interventions for people suffering from metabolic syndrome.

According to the 2017 World Obesity Federation report, a third of Earth’s population is overweight and obese. By 2025, 34% of adults in Canada will live with obesity. Treating obesity-related health problems costs Canada roughly $34 billion each year. Troublingly, only 40 out of almost 80,000 physicians in the country are formally trained and certified to address obesity and weight management.

When the body is under attack, the immune system secretes chemicals known as cytokines, chemical messengers that coordinate communication between the immune system and damaged tissues. Cytokines help flush out pathogens or restore body functions to normal levels. This, in a nutshell, is known as an inflammatory response. Poudel and his team set about to study the anti-inflammatory effects of berberine, which were well documented but poorly understood.

The authors examined cells that eventually form the body’s muscles. Importantly, these cells metabolize almost 80% of the body’s glucose—some with insulin’s help and others independent of it.

People suffering from obesity have abnormal inflammatory responses and produce high amounts of cytokines. The high concentration of cytokines can cause insulin resistance—a major contributor to type 2 diabetes and metabolic syndrome.

To find out just how berberine saves the day, Poudel and his colleagues bathed muscle cells in a broth of pro-inflammatory cytokines together with berberine.

They found that berberine boosted anti-inflammatory pathways—specifically, they identified the exact sequence of chemicals that berberine communicates with to effectively shut down the abnormal immune response or, in some cases, even reverse it.

Cytokines also reduce almost 70% of mitochondrial function leading to muscle weakness and lethargy. The authors discovered that berberine restores mitochondrial function and pinpointed the exact biochemical pathways.

Ultimately, Poudel and his team found that berberine’s fight against diabetes and metabolic syndrome is a delicate balance among multiple chemical signals, holding some back while encouraging others to step up.

The intricate process, clearly outlined for the first time, may be the first therapeutic targets of berberine in treating insulin resistance.

New study uses social media to track spread of dengue fever

By the fall of 2014, almost 15,000 cases of dengue fever broke out across Guangzhou, one of China’s largest cities. The metropolis went on to become one of the hardest-hit places by 2014’s global dengue fever outbreak.

During that time, social media was packed with discussions, comments, and posts about the epidemic, spread by a mosquito-borne virus. A digital footprint captured people’s hopes, fears, and attempts to understand the outbreak.

Now, researchers from China and the United States have followed these footprints to conduct a pioneering study. In a new paper published in Geomatica, Junfang Gong and colleagues analyzed social media activity during Guangzhou’s dengue fever outbreak to understand when, where, and how social media posts originated and spread.

By understanding how social media data originate during disease outbreaks and combining it with public health statistics, health officials can design and implement intervention programs to minimize the physical, social, and economic toll of such events.

Previous public health research has used the occurrence of social media posts to predict and monitor the spread of diseases, but few studies have focused on how social media discussions change and spread over time as diseases progress or abate. So far, researchers have had snapshots of how discussion of the disease disseminates online, but Gong and his colleagues’ research provides moving pictures.

For the study, authors collected social media posts on dengue fever outbreak from China’s most popular micro-blogging website, Sina – think a combination of Twitter and Facebook.

Gong and colleagues used a machine-learning algorithm called text mining to discover and extract various topics in social media posts. They statistically analyzed how posts on particular topics – “Southeast Asia” or news events such as “reports of dengue fever in 2014, China” or “prevention of dengue fever” were being noticed, followed, and discussed via comments, reports, and endorsements over time.

According to their research, it took more than 20 days for discussions related to dengue fever to become a hot topic on micro-blogs. The initial period lasted approximately 30 days. After that time, many more micro-blogs on the topic were posted. Users continued the discussion on dengue fever until gradually losing interest in the discussion in a phase that lasted 20 days.

The authors summarize that “given that the spread of dengue fever typically started in late spring and leveled off in late summer or early autumn, the 80-day time period explained the diffusion process very well.”

The authors also used location information of social media posts to figure out if posts with similar topics were originating from similar geographical locations.

By tracking the place and time of origin of the social media posts, the authors were also able to explain the spread of dengue fever in various locations both inside and outside the city. For the latter, they hypothesize that modern transportation networks would move carriers of the virus from one place to another, even over lengthy distances.

Understanding the process of where, when and how social media posts spread will be critical for leveraging social media to spread public-health related information more accurately and efficiently.

Study lays out first-ever, holistic guide to teaching future-forward Canadian engineers

There are a number of things going on under an engineer’s helmet—figuring out the right mixture of steel and concrete, juggling multiple construction schedules, and keeping on top of paperwork, just to name a few. Engineering goes well beyond math and physics.

Today’s engineers are managers and leaders, foreman and contractors, and they are required to maintain good relations between teams while keeping up with the latest advances in technology. Yet many departments in Canadian universities—particularly in departments such as construction and building—are failing to produce well-rounded engineers.

Conrad Boton and colleagues from Ecole de Technologie Supérieure (ETS), Montreal, have a solution.

Their recent paper A framework for Building Information Modeling implementation in engineering education published in the Canadian Journal of Civil Engineering is the first-ever comprehensive how-to guide for universities teaching an engineering framework known as Building Information Modeling (BIM).

BIM, according to the authors, anticipates ever-changing demands of the construction industry while teaching students core engineering skills. It helps engineers manage construction projects more efficiently and in more cost-effective ways. Engineers, contractors, and foremen use BIM to work collectively to design and construct facilities.

Unlike automotive and aerospace industries, Canada’s Architecture, Engineering, and Construction sector, valued at $171-billion, suffers from low rates of productivity and has been slow to adopt innovative methods like BIM for one critical reason—the lack of reliable personnel that can implement BIM.

The root of the problem, the authors suggest, lays in an outdated engineering curricula.

University departments are often unaware of what BIM is. Many are not qualified to teach it. In some cases, universities lack adequate resources to implement it. Boton and colleagues worked with ETS-Montreal, a Canadian engineering school, to implement BIM on a trial basis.

The paper not only points out the challenges the trial program faced but also highlights the advantages and positive feedback the school received from students.

The authors suggest the following:

Focus on three critical things: 1. Technology—including software and equipment, 2. Processes—communication and information exchange between different teams, and 3. Policy—regulations, guidelines, and contracts.

Start slow but start early: Introduce specific BIM content in modules before a full implementation and, critically, introduce those courses beginning at the undergraduate level. “The bachelor degree level is where engineering core competencies are expected to be acquired, including both foundational and industry-specific competencies,” the authors state.

Mix interactive classroom activities with the classic “chalk and talk” approach, which is often unaware of current industrial practices and does not teach teamwork and communication skills necessary for graduates. The industry is changing rapidly—knowledge, according to the authors, is no longer generated by academic research but by advances in the industry.

In addition to faculty members, have industry professionals who work with local construction industries as teachers. The latter often have valuable knowledge of good business practices, can help students connect theory and practice, and can assist in modifying the curricula based on current trends.

The authors suggest applying the ETS-Montreal case study to other academic institutions in Canada. Currently, they are working with the school and monitoring the response from students, faculty members, and local industries as well as improving the curricula based on continual feedback from both students and industry professionals that will continue to inform how the curricula can be improved.

A High-Tech Window into the Microbial World

To better understand some of the tiniest creatures known to us we need to make cutting-edge technology even sharper and more accessible.

Joost T.P. Verhoeven and colleagues from Memorial University of Newfoundland have done just that in a recent paper published in the Canadian Journal of Microbiology.

The paper describes two methods, ViDiT (Virus Discovery Ion Torrent) and CACTUS (not an acronym), that help biologists extract and read, respectively, viral and bacterial DNA in a quicker, cheaper, and easier way compared with current methods.

Biologists often deal with less than ideal instances of microbial samples ranging from contaminated soil to improperly stored clinical swabs. Often, DNA is hardly detectable in such samples.

ViDiT is sensitive enough to detect low-quality and low quantities of DNA but also versatile enough to extract genetic material from a wide variety of samples.

CACTUS does not require complicated software or expensive hardware—something out of reach for many laboratories. In fact, it can even work on virtual cloud-computing platforms, which means a laboratory anywhere in the world can analyze their data at a low cost.

To prove the effectiveness and versatility of ViDiT and CACTUS, Verhoeven and colleagues collected tissue samples from wild birds of six different species from all over Newfoundland as part of studies on the avian influenza A virus.

ViDiT and CACTUS detected previously known viral genetic material in addition to discovering never-before-described viruses. In each case, they could also read the entire viral gene sequence.

Deep-sea samples are also challenging to work with because it is difficult to extract high-quality genetic material from these microbes that live in extreme environments. The team was successful in reading and deciphering a diverse microbial community in a deep-sea carnivorous sponge—their results closely matched previously published data.

Their work comes at a critical time.

Rising cross-border trade, international tourism, and climate change have contributed to a significant rise in infectious disease outbreaks over the past several decades.

According to the World Bank, the spread of infectious disease is one of the top global economic risks—the annual global cost of moderate to severe pandemics is roughly U.S. $570 billion.

Every month, the World Health Organization receives 5,000 early warning disease signals from around the world, of which at least 30 are closely investigated for their potential to cause epidemics.

Whether it is tracking emerging diseases or finding new microbes, ViDiT and CACTUS are robust, flexible, and cost-effective solutions that put the microbial community at biologists’ fingertips quickly and accurately.

A Taste of the Wild—Deciphering the Lobster Mushroom

Nutty, sweet smell with a distinctive taste—this is a typical culinary description of lobster mushrooms, prized by amateur cooks and chefs alike. Yet behind the tasty morsels is a hostile and unforgiving microbial world. The parasitic fungus Hypomyces lactifluorum infects and colonizes the mushroom Russula brevipes—white, bitter and non-edible—and transforms it into the orange, culinary wonder known as the lobster mushroom. In a pioneering study in Botany, combining molecular biology and biochemistry, Genevieve Laperriere and colleagues from the University of Quebec at Trois-Rivières track the infection process and colonization of Russula brevipes by the parasitic fungus Hypomyces lactifluorum, and decipher the complex taste profile of lobster mushrooms resulting from that infection. Their work will help strengthen the large-scale industrial use and commercialization of lobster mushrooms in Canada, one of the world’s largest exporters of fresh mushrooms with an industry value of $1 billion.

Laperriere et al. extracted DNA from infected (edible) and non-infected (non-edible) mushrooms collected from various sites around Quebec. They found that the infected and edible lobster mushroom mostly contains the DNA of the parasitic fungus with only trace amounts of Russula brevipes DNA.

Laperriere and his team also measured intermediate products of chemical reactions, or metabolites, in infected and non-infected mushrooms. Metabolites help determine how fungi look and taste, and whether they are fit to eat. They found that through the course of its infection, the parasitic fungus completely alters the diversity and amount of metabolites in Russula brevipes.

For example, terpenes, small chemical products synthesized by Russula brevipes, are completely absent from lobster mushrooms infected by the parasitic fungus. Russula brevipes produces these chemicals partly for its protection and defense against infection. As Laperriere and his team found, the complete absence of terpenes from lobster mushrooms indicates a successful infection by the parasitic fungus. Interestingly, some varieties of terpenes are responsible for the pungent and bitter taste of Russula brevipes. Laperriere and his team speculate that during the course of infection, these chemicals get converted into other more flavourful compounds, making lobster mushrooms more palatable.