Eye-tracking is a potentially valuable diagnostic and analytical tool for psychology and market research, but the complexity and expense of today’s technology limits its use. “The field is kind of stuck in this chicken-and-egg loop,” says Aditya Khosla, co-leader of the project and an MIT graduate student in electrical engineering and computer science, in an MIT statement. “Since few people have the external devices, there’s no big incentive to develop applications for them. Since there are no applications, there’s no incentive for people to buy the devices.”
Khosla and fellow computer science doctoral student Kyle Krafka at University of Georgia seek to break this cycle by developing an inexpensive eye tracking application on ordinary mobile devices, using their built-in cameras. Solving the problem, however, requires going beyond hardware. The system also needs to recognize and interpret small subtle eye movements, which calls for sophisticated models and software.
For their solution, Khosla, Krafka, and colleagues employ machine learning, where underlying algorithms for software are derived from patterns of individual behavior, in this case the ways people’s eyes move when they use their phones and tablets. But the developers discovered the largest available data set of gaze patterns has only 50 cases. They needed many more data points to achieve a sufficient level of accuracy.
The researchers turned to crowdsourcing to build their database. The team created an Apple iPhone app that flashes a red dot at a spot on a phone or tablet screen, then replaces it with an R or L, an instruction to swipe the screen right or left. As the participant reacts to the red dot and executes the swipe, the device’s camera captures images of the user’s face. The team recruited participants from Mechanical Turk, a crowdsourcing site offered by Amazon.com, and paid each individual a small fee.
The team’s efforts yielded data from 1,450 participants with each individual generating an average of 1,600 images. To continuously mine these data efficiently, the researchers employed a technique known as dark knowledge, which creates an interim trained network offering an approximate solution for the larger generalized model. That interim network then makes possible further training and learning on a much smaller set of new data that fits on a smartphone.
The authors say their network, called iTracker, now yields an eye-tracking prediction accuracy within 1.3 centimeters when calibrated on smartphones and 2.1 centimeters on tablets. The researchers note in an MIT statement, to achieve an accuracy within 0.5 centimeters required for commercial applications will mean capturing data from some 10,000 participants.
Khosla is co-founder and chief technologist of PathAI, a start-up enterprise applying neural networks and deep machine learning to large data sets for diagnosing cancer.
While many breath analysis devices are on the market, the team from the Texas Analog Center of Excellence at UT-Dallas — with colleagues from Ohio State and Wright State universities — are seeking a simpler and less expensive alternative to today’s bulky, complicated, and expensive systems. To meet this goal, the researchers designed their device to work on a complementary metal–oxide–semiconductor, or CMOS, chip, the familiar integrated circuit found in everyday electronic devices.
Odors encountered from exhaled breath are the result of reactions in the stomach and blood from chemicals as they reach the air in the lungs. The device detects chemical molecules in the gas exhaled from an individual, with a rotational spectrometer that transmits electromagnetic waves. As the electromagnetic waves interact with the gas molecules, they produce characteristic electronic signals that the device detects and interprets. Developers of this device they call an “electronic nose” say it can detect these gas molecules even in low concentrations.
As a proof of concept, the researchers tested the device to distinguish between acetone gas and ethanol in breath. That distinction is important, since people with type 1 diabetes produce acetone in their blood and breath, which in police breathalyzer tests is often confused with ethanol leading to DUI arrests. The researchers report their device could accurately identify acetone gas from exhaled ethanol.
The researchers believe their device has the potential to give the same results of today’s blood test without drawing blood. “Smell is one of the senses of humans and animals, and there have been many efforts to build an electronic nose,” says Sharma in a university statement. “We have demonstrated that you can build an affordable electronic nose that can sense many different kinds of smells. When you’re smelling something, you are detecting chemical molecules in the air.”
The team is designing a prototype programmable system for field testing in early 2018. The researchers expect the electronic nose to be used first in industrial settings, and later in clinics and hospitals. Eventually, the device could become a household appliance, reducing the need for diagnostic blood tests.
16 June 2016. Biomedical engineers developed techniques for growing personalized replacement facial bone from stem cells that in tests with pigs precisely fit their recipients’ faces. The team from Columbia University in New York published its findings in yesterday’s issue of the journal Science Translational Medicine (paid subscription required).
The researchers led by biomedical engineering professor Gordana Vunjak-Novakovic are seeking more productive methods for reconstructing facial bone tissue to treat accident injuries and other trauma, or deformities from cancer surgery or birth defects. Current techniques usually transplant bone and skin tissue from other parts of the body, which the authors say are difficult to precisely fit the faces of recipients, and often cause pain and complications.
Vunjak-Novakovic’s team — with colleagues from Columbia’s dentistry school, Tulane and Louisiana State Universities, and the companies epiBone and LaCell LLC — are developing a process growing new bone tissue from stem cells derived from adipose or fat tissue in the recipient. The stem cells are then grown on a scaffold made of decellularized cattle bone matrix, designed from CT scan images of the recipient’s face, using precise milling techniques guided by the images. The stem cells are cultured on the scaffold in a bioreactor, feeding nutrients, oxygen, and metabolites to grow into immature bone tissue, a process taking about 3 weeks. The new bone tissue is then transplanted into the recipient’s face, which integrates with and grows on the host bone at the injury site.
The researchers tested the process with 14 Yucatán minipigs, a type of miniature pig often used as a model in medical research for their similarities to aspects of human anatomy. The team chose bones in the ramus-condyle unit, a weight-bearing component of the jaw and skull to reconstruct in their tests. In addition, the researchers intentionally cultured the stem cells a long distance, about 1,200 miles, from the transplant surgery site to simulate realistic manufacturing and shipping conditions.
The team tested the new bone tissue 6 months after the transplants, comparing the results to pigs with similar injuries, but either without any treatment or with only the scaffolds implanted. The results show the stem-cell grown bone tissue integrated with the host bone, maintaining the same anatomical structure as the recipient. In addition, compared to the pigs without the new bone transplants, the recipients grew more bone tissue with normal blood vessel development.
“The need is huge, especially for congenital defects, trauma, and bone repair after cancer surgery,” says Vunjak-Novakovic in a university statement. “The quality of the regenerated tissue, including vascularization with blood perfusion, exceeds what has been achieved using other approaches.”
The researchers are now investigating addition of cartilage to regenerate tissue for more complex injuries, and advancing the technology through more preclinical stages. Vunjak-Novakovic and first author Sarindr Bhumiratana are co-founders of the spin-off enterprise epiBone Inc. in New York that licenses the bone-regeneration technology and is planning for clinical trials.
In the following video, Bhumiratana and epiBone CEO Nina Tandon tell more about the technology.
Additive manufacturing system at Lawrence Livermore National Lab (Jamie Douglas, Livermore Lab)
15 June 2016. Ansys Inc., a developer of engineering simulation software and services, is starting an additive manufacturing lab at University of Pittsburgh’s engineering school. The project is initially funded by a grant from America Makes, but further financial details about the collaboration between the Pittsburgh-based company and the university were not disclosed.
Additive manufacturing is 3-D printing in industrial form, with systems often consisting of modeling software, computer controller, printing platform, and raw materials. Production sites read data from the modeling software, with the computer-controlled printer putting down successive layers of powder, metal, liquid, or plastic to create an object. While the concept is simple, additive manufacturing can be designed to meet complex production needs, such as individualized products and eventually human organs.
The Ansys Additive Manufacturing Research Laboratory will be a 1,200 square-foot facility at Pittsburgh’s engineering school for solving issues facing advanced applications of the technology. Ansys and the university say the lab plans to investigate the use of lasers with metal in 3-D printing, which can result in unexpected melting, as well as deformation in the metal from rapid heating and cooling. The collaboration is expected to simulate 3-D processes to better understand how deformations happen, and prevent their occurrence.
Ansys offers a 3-D modeling software package called SpaceClaim that, according to the company, interacts with computer-assisted design software to analyze designs and simulate industrial 3-D printing for testing alternative designs and highlighting problems before production, as well as speed edits and repair. Ansys says the computational tools in its software make it possible to test a wide range of designs, materials, flows, and shapes before producing even the first prototype.
The partnership will support research and training with Ansys staff involving 3-D laser printing with metals, as well as alloys and polymers. Pittsburgh faculty and students will have access to Ansys resources to simulate and analyze stress and fatigue on materials in producing aircraft, vehicles, and medical devices. The lab will also be open to other businesses in the biomedical, aerospace, and defense industries.
The university and Ansys say their collaboration is a result of funding from National Additive Manufacturing Innovation Institute, also known as America Makes, a public-private partnership promoting additive manufacturing. America Makes, in Youngstown, Ohio, has members from industry, academia, government, and non-government organizations.
Lal and colleagues from UC-San Diego’s Center for Excellence in Nanomedicine and Engineering are seeking a way of shrinking genetic analysis technologies into portable diagnostics equipment that can detect biomarkers for diseases resulting from mutations, variations in DNA sequences or RNA transcribed from the genetic code. Current equipment for detecting genetic variations called single nucleotide polymorphisms or SNPs are found in stationary lab equipment, which while accurate and effective, are expensive and time-consuming. The authors note that the growing use of precision medicine will likely mean increased demand for detection of genetic defects, thus the need for rapidly increasing the supply of these systems.
SNPs note differences in nucleotides, the DNA base building blocks identified by the letters A, C, G, and T. Most SNPs occur normally in the genetic code, but some of these variations are associated with inherited diseases, autoimmune conditions, and neurodegenerative disorders, as well as more common heart disease, cancer, and diabetes. Most of the current systems for detecting SNPs in specimen samples use enzymes, requiring expensive equipment in a lab setting. Efforts to develop alternatives to enzyme analysis, say the authors, so far encounter problems in specificity, returning too many false negative results.
The UC-San Diego team developed its device to search DNA or RNA strands and identify specific SNP sequences. The device analyzes individual strands of DNA or RNA looking for the SNP. When a targeted SNP sequence is encountered and binds to the chip, the device analyzes complementary strands replacing those with weak binding until a perfect match in the sequence is found. The matching strand then binds to the chip, which generates a signal indicting the device detected the targeted SNP.
The researchers built the device on graphene, a material closely related to graphite like that used in pencils. The material is very light, strong, chemically stable, and can conduct both heat and electricity, with applications in electronics, energy, and health care. With graphene, say the authors, they can replace measurement of chemical fluorescence used in standard lab systems, and algorithms in software for analysis, with electronic matching of complementary DNA or RNA strands. Thus graphene makes it possible to shrink the the device to a single chip, which in proof-of-concept tests returned results that correlated with standard lab systems using fluorescence measurement.
The authors say their electronic approach analyzing double DNA or RNA strands is also more reliable than other portable devices that sequence single strands. They report the graphene chip can sequence strands up to 47 nucleotides in length, which the authors say are the longest strands analyzed so far.
“We expected that with a longer probe, we can develop a reliable sequence-specific SNP detection chip,” says Lal in a university statement. “Indeed, we’ve achieved a high level of sensitivity and specificity with the technology we’ve developed.” For their next steps, the team plans to scale-up the technology and add wireless transmission. Later, the researchers hope to develop the chip into a clinical diagnostics tool for mobile and point-of-care devices.
Type 2 diabetes is a disorder where the pancreas produces some, but not enough insulin, or the body cannot process insulin, and accounts for some 90 percent of all diabetes cases. Diabetes can also affect the heart and blood vessels, leading to coronary heart disease, heart attack, and stroke. According to International Diabetes Federation, cardiovascular disease is the most common cause of death in people with diabetes.
Liraglutide is in a class of drugs known as glucagon-like peptide-1 receptor agonists, hormones that generate a greater insulin response, and thus help regulate blood glucose levels. These drugs also reduce glucagon that raises glucose levels and induce satiety, or feeling, full after eating, thus help control weight. Liraglutide is a long-acting drug, prescribed for people who find it difficult to control their blood glucose levels through diet and exercise and need extra help.
The clinical trial enrolled 9,340 individuals with type 2 diabetes at 410 sites in 32 countries. Participants were randomly assigned to receive injections of liraglutide or a placebo under the skin each day, in doses starting from 0.6 milligrams and increasing to 1.8 milligrams over 3 weeks. In addition to either liraglutide or placebo, individuals in the trial took other drugs to control their diabetes, high blood pressure, or cholesterol. Participants were on average 64 years old and nearly three-quarters (73%) had a previous history of heart disease.
After the initial treatments, the research team led by John Buse at University of North Carolina medical school in Chapel Hill, followed up with participants after 30 and 90 days, then every 6 months from 3.5 to 5 years. Individuals in the trial took electrocardiograms and gave blood and urine samples at the beginning of the study, then at each of the follow-up points, where the researchers checked for signs of cardiovascular or kidney disease.
The results show a 22 percent reduction in death from cardiovascular causes among participants receiving liraglutide compared to placebo recipients. Liraglutide recipients also had a 13 percent lower risk of non-fatal heart attack or non-fatal stroke, and 15 percent lower risk of mortality for any reason. In addition, participants receiving liraglutide had a 22 percent lower risk of advanced diabetic kidney disease than their counterparts receiving the placebo.
Percentages of participants experiencing adverse effects were about the same in both groups. Instances of abnormal tissue growth were slightly higher among liraglutide recipients, while individuals receiving the placebo had somewhat more cases of pancreatitis. In neither case were the differences large enough to be statistically reliable.
FDA first approved liraglutide for control of blood glucose levels in 2010, but the trial results highlight the drug’s additional value to people with diabetes. “Type 2 diabetes treatments that can also reduce cardiovascular risk are important,” says Buse in Novo Nordisk statement, “since cardiovascular disease is the leading cause of death worldwide in this patient population.”
High-resolution CT scan of lungs with idiopathic pulmonary fibrosis (IPF Editor, Wikimedia Commons)
14 June 2016. A clinical trial testing a drug treating chronic cough — coughs continuing for 8 weeks or more — will begin recruiting participants in the U.K. to join other sites in the U.S. The U.K. sites are under the direction of National Institute for Health Research, testing a small molecule, or low molecular weight, compound code-named AF-219, developed by biotechnology company Afferent Pharmaceuticals in San Mateo, California.
Chronic cough is condition where an upper respiratory disease causes a cough lasting a few weeks, but the nerve fibers remain irritated and fail to return to their normal state, causing a coughing reflex. The same mechanism, says Afferent, may be at work in disorders such as idiopathic pulmonary fibrosis or whooping cough, where non-productive coughing serves no useful purpose.
AF-219 is Afferent’s lead product, designed to address proteins known as P2X3 receptors that overproduce in nerve fibers when hyper-sensitized, involving in this case nerve fibers in lung tissue. The company says chronic cough may affect as many as 10 percent of the population, particularly among people with lung disorders that do not respond to treatments directly addressing those conditions. Afferent is also developing drugs targeting P2X3 receptors involved in chronic pain, cardiovascular diseases, and other disorders.
The company conducted a proof-of-concept trial with 24 individuals in the U.K. having persistent chronic cough, testing AF-219 against a placebo. The results, reported in The Lancet, show participants receiving AF-219 in tablets twice a day reduced their cough frequency by 75 percent.
The new clinical trial is recruiting 200 individuals in the U.S. and U.K., testing AF-219 at 3 dosage levels against a placebo for 12 weeks. The U.K. site is in Manchester, under the direction of University of Manchester medical school professor Jaclyn Smith, who led the proof-of-concept study. Smith is also an inventor of the VitaloJAK, a wearable medical device that records cough frequency without audio or video recordings, also used in this trial.
“Previously, studies relied on patient reported outcomes, which are not always reliable,” says Smith in a National Institute for Health Research statement. “This may lead to effective drugs being dismissed due to inaccurate reporting and, I believe, is a contributing factor to the lack of interest from big pharma companies in investigating new cough treatments.”
Smith adds, “We are just beginning to understand how the nerves in the airways are involved in pathologic cough such as chronic cough. With recent developments in the technology to effectively measure coughs and this important new drug, we have started to see real progress in this area.”
Last week, the pharmaceutical company Merck agreed to buy Afferent Pharmaceuticals for $500 million in cash, and another $750 million in milestone payments, connected in part to the development and commercialization of AF-219.
The consortium seeks to create faster as well as more reliable and consistent techniques for generating human cells used in treatments and tissue engineering, as well as medical devices, drug discovery platforms, and organ-on-chip models that test drugs for toxicity. The consortium says several federal agencies spent as much as $3 billion on research in regenerative medicine that yielded promising cell-based technologies, but the ability to produce human cells in sufficient quantities and in consistent high quality, needed for day-to-day use of those technologies, is lagging behind.
The road map, originally published in February 2016, offers a strategy through 2025 for participants in the cell manufacturing community — pharmaceutical and biotech companies, medical device developers, academic institutions, private foundations, and federal agencies — to produce this capability at a large scale and high, consistent quality. The strategy covers manufacturing of autologous cells provided by patients for retransplant, donated or allogenic cells, and pluripotent stem cells that transform or differentiate from an immature state to working cells in the body.
Achieving this goal, says the consortium, requires the cell manufacturing community to work along two tracks. The first track upgrades current technologies for cell processing, preservation and distribution, and quality assurance. Cell processing technologies cover screening and selection methods, culture media, cell expansion methods and equipment, separation techniques, and cell modification and differentiation methods.
Cell preservation and distribution includes cryopreservation techniques, storage technologies, and systems for tracking the movement of cell products. Processes for monitoring and quality assurance cover cell attribute testing and measurement, as well as data analytics and management.
A second simultaneous track aims to strengthen the industry’s business infrastructure. One part of this track aims to upgrade the industry’s workforce through training, higher education, and more cross-industry collaboration. Another part of strengthening the foundation aims to enhance standardization and regulation, through development of quality standards, more consistency in the supply chain, and improved regulatory strategies.
The road map calls for additional investments in advanced cellular manufacturing to maintain U.S. leadership in the field, noting similar national initiatives in the U.K., Canada, Australia, and Germany. While the document did not spell out specific numbers, the consortium estimates the need for “several hundred million dollars a year” for the next 10 years. The report says its members are prepared to provide monetary and in-kind support, but “additional or matched external or federal funding would multiply the impact of the consortium’s efforts.”
National Cell Manufacturing Consortium was established in 2014, and now has some 60 representatives from industry, government, and not-for-profit organizations. The group is funded by an advanced manufacturing technology grant from National Institute of Standards and Technology.
Clinic in Haiti treating cholera patients in 2010 (Kendra Helmer, USAID)
13 June 2016. Food and Drug Administration approved for marketing the U.S. a vaccine to prevent cholera for travelers to regions in the world where contaminated food or water is causing the disease. Vaxchora by PaxVax Inc. in Redwood City, California is the first vaccine approved by the agency to prevent cholera.
According to World Health Organization, from 1.4 to 4.3 million cases of cholera occur each year, killing 28,000 to 142,000 people. The disease is caused by Vibrio cholerae bacteria, found in contaminated food and water, particularly in low-resource regions with weak public sanitation facilities, and in humanitarian crisis areas, where normal water supplies and sanitation systems are disrupted. Cholera is characterized by diarrhea and vomiting, which if severe, can lead to dehydration.
Vaxchora is an oral vaccine, distributed as a powder, mixed with bottled water, and taken 10 days before travel to regions experiencing cholera. The vaccine has live, but weakened cholera bacteria from a subgroup labeled 01, which according to WHO accounts for a majority of cases. The original research for Vaxchora was conducted at University of Maryland medical school by microbiologists Myron Levine and James Kaper, and licensed to PaxVax for development.
The vaccine was tested in a series of clinical trials, including a late-stage study with nearly 200 adult healthy volunteers in the U.S. Participants in the trial were given either Vaxchora, code-named at the time PXVX0200, or a placebo, then exposed to Vibrio cholerae bacteria. The results show 10 days after taking the vaccine, 90 percent participants were protected against cholera when exposed to the bacteria. Likewise, 90 days after the vaccine, 80 percent were protected. Participants receiving the placebo were given immediate care with antibiotics and fluid replacement.
In other clinical studies in the U.S. and Australia, says FDA, healthy participants given the vaccine developed antibodies indicating protection against cholera. Several trials tested Vaxhora’s safety, with tiredness, headache, abdominal pain, nausea, vomiting, lack of appetite and diarrhea as the most common side effects. The vaccine has not been tested, however, in people already exposed to cholera bacteria, or with other bacterial sub-types.
FDA gave Vaxchora priority review and fast-track status to expedite its review through the agency. In addition, the vaccine received special priority review consideration to encourage development of treatments and vaccines for tropical diseases often neglected by drug companies. The company expects to have Vaxchora on the market during the third quarter of 2016.
Ingber and colleagues developed the new technique to make it easier to discover Staphylococcus aureus, or staph bacteria, and other pathogens causing infections contracted in hospitals and clinics. Staph bacteria are associated with osteoarthritis infections, sometimes resulting from joint replacements, where infections can lead to painful inflammation. Samples of tissue or joint fluid from patients, however, are complex and pathogens in theses samples are often difficult to isolate and identify.
The Wyss team, with former institute colleagues in France, adapted a process designed earlier to clean blood of sepsis bacteria, causing dangerous infections. That method uses an engineered molecule called FcMBL acting like a natural protein known as mannose-binding lectin that binds to carbohydrates found in a broad range of bacteria and viruses including those associated with sepsis. Wyss researcher Michael Super, a co-author of the new study, genetically engineered mannose-binding lectin to combine with molecules from the Fc region in antibodies that extends their persistence in blood serum.
As reported in Science & Enterprise, Ingber and Super founded Opsonix Inc., a spin-off company to commercialize the sepsis-cleaning technology. That technology includes coating nanoscale beads with the engineered mannose-binding lectin, then magnetizing the coated beads to attract microbes and toxins.
When the researchers tried those techniques to isolate staph bacteria in tissue samples from a biobank of osteoarthritis patients, they discovered the magnetized and coated beads cannot bind directly to bacteria in the joint fluids. One reason is the viscosity of the joint fluid, but also the presence of immune cells and proteins in the samples that mask carbohydrate molecules on bacteria that serve as binding targets for FcMBL.
To improve the performance and efficiency of the beads, the researchers wash the clinical samples in an enzyme cocktail before exposure to the FcMBL. The pre-treatment cocktail is made of proteases, enzymes that break down the peptides holding together the amino acids in proteins masking the binding carbohydrates. The pre-treatment also uses hyaluronidase, another enzyme that helps break down the viscous nature of the joint fluids.
The combination of pre-treatment and FcMBL beads makes it possible to isolate the staph bacteria from the samples within 2 hours. Once isolated, the bacteria or other pathogens can then be tested with standard lab procedures.
Ingber says in a Wyss Institute statement that the technique can be extended to other pathogens and specimen samples, such as blood, urine, sputum, and cerebral spinal fluid. “In addition to saving more lives,” adds Ingber, “this new method also should reduce the use of broad-spectrum antibiotic therapies, or sub-optimal regimens, and thereby decrease development of antibiotic-resistant organisms that become a more general threat in the long run.”
Science and Enterprise offers researchers, administrators, business people, government officials, and interested observers news and insights into the value chain connecting scientific discovery with the marketplace.