본문 바로가기
대메뉴 바로가기
KAIST
Newsletter Vol.24
Receive KAIST news by email!
View
Subscribe
Close
Type your e-mail address here.
Subscribe
Close
KAIST
NEWS
유틸열기
홈페이지 통합검색
-
검색
KOREAN
메뉴 열기
ANG
by recently order
by view order
Image Analysis to Automatically Quantify Gender Bias in Movies
Many commercial films worldwide continue to express womanhood in a stereotypical manner, a recent study using image analysis showed. A KAIST research team developed a novel image analysis method for automatically quantifying the degree of gender bias in films. The ‘Bechdel Test’ has been the most representative and general method of evaluating gender bias in films. This test indicates the degree of gender bias in a film by measuring how active the presence of women is in a film. A film passes the Bechdel Test if the film (1) has at least two female characters, (2) who talk to each other, and (3) their conversation is not related to the male characters. However, the Bechdel Test has fundamental limitations regarding the accuracy and practicality of the evaluation. Firstly, the Bechdel Test requires considerable human resources, as it is performed subjectively by a person. More importantly, the Bechdel Test analyzes only a single aspect of the film, the dialogues between characters in the script, and provides only a dichotomous result of passing the test, neglecting the fact that a film is a visual art form reflecting multi-layered and complicated gender bias phenomena. It is also difficult to fully represent today’s various discourse on gender bias, which is much more diverse than in 1985 when the Bechdel Test was first presented. Inspired by these limitations, a KAIST research team led by Professor Byungjoo Lee from the Graduate School of Culture Technology proposed an advanced system that uses computer vision technology to automatically analyzes the visual information of each frame of the film. This allows the system to more accurately and practically evaluate the degree to which female and male characters are discriminatingly depicted in a film in quantitative terms, and further enables the revealing of gender bias that conventional analysis methods could not yet detect. Professor Lee and his researchers Ji Yoon Jang and Sangyoon Lee analyzed 40 films from Hollywood and South Korea released between 2017 and 2018. They downsampled the films from 24 to 3 frames per second, and used Microsoft’s Face API facial recognition technology and object detection technology YOLO9000 to verify the details of the characters and their surrounding objects in the scenes. Using the new system, the team computed eight quantitative indices that describe the representation of a particular gender in the films. They are: emotional diversity, spatial staticity, spatial occupancy, temporal occupancy, mean age, intellectual image, emphasis on appearance, and type and frequency of surrounding objects. Figure 1. System Diagram Figure 2. 40 Hollywood and Korean Films Analyzed in the Study According to the emotional diversity index, the depicted women were found to be more prone to expressing passive emotions, such as sadness, fear, and surprise. In contrast, male characters in the same films were more likely to demonstrate active emotions, such as anger and hatred. Figure 3. Difference in Emotional Diversity between Female and Male Characters The type and frequency of surrounding objects index revealed that female characters and automobiles were tracked together only 55.7 % as much as that of male characters, while they were more likely to appear with furniture and in a household, with 123.9% probability. In cases of temporal occupancy and mean age, female characters appeared less frequently in films than males at the rate of 56%, and were on average younger in 79.1% of the cases. These two indices were especially conspicuous in Korean films. Professor Lee said, “Our research confirmed that many commercial films depict women from a stereotypical perspective. I hope this result promotes public awareness of the importance of taking prudence when filmmakers create characters in films.” This study was supported by KAIST College of Liberal Arts and Convergence Science as part of the Venture Research Program for Master’s and PhD Students, and will be presented at the 22nd ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW) on November 11 to be held in Austin, Texas. Publication: Ji Yoon Jang, Sangyoon Lee, and Byungjoo Lee. 2019. Quantification of Gender Representation Bias in Commercial Films based on Image Analysis. In Proceedings of the 22nd ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW). ACM, New York, NY, USA, Article 198, 29 pages. https://doi.org/10.1145/3359300 Link to download the full-text paper: https://files.cargocollective.com/611692/cscw198-jangA--1-.pdf Profile: Prof. Byungjoo Lee, MD, PhD byungjoo.lee@kaist.ac.kr http://kiml.org/ Assistant Professor Graduate School of Culture Technology (CT) Korea Advanced Institute of Science and Technology (KAIST) https://www.kaist.ac.kr Daejeon 34141, Korea Profile: Ji Yoon Jang, M.S. yoone3422@kaist.ac.kr Interactive Media Lab Graduate School of Culture Technology (CT) Korea Advanced Institute of Science and Technology (KAIST) https://www.kaist.ac.kr Daejeon 34141, Korea Profile: Sangyoon Lee, M.S. Candidate sl2820@kaist.ac.kr Interactive Media Lab Graduate School of Culture Technology (CT) Korea Advanced Institute of Science and Technology (KAIST) https://www.kaist.ac.kr Daejeon 34141, Korea (END)
2019.10.17
View 20537
Sungjoon Park Named Google PhD Fellow
PhD candidate Sungjoon Park from the School of Computing was named a 2019 Google PhD Fellow in the field of natural language processing. The Google PhD fellowship program has recognized and supported outstanding graduate students in computer science and related fields since 2009. Park is one of three Korean students chosen as the recipients of Google Fellowships this year. A total of 54 students across the world in 12 fields were awarded this fellowship. Park’s research on computational psychotherapy using natural language processing (NLP) powered by machine learning earned him this year’s fellowship. He presented of learning distributed representations in Korean and their interpretations during the 2017 Annual Conference of the Association for Computational Linguistics and the 2018 Conference on Empirical Methods in Natural Language Processing. He also applied machine learning-based natural language processing into computational psychotherapy so that a trained machine learning model could categorize client's verbal responses in a counseling dialogue. This was presented at the Annual Conference of the North American Chapter of the Association for Computational Linguistics. More recently, he has been developing on neural response generation model and the prediction and extraction of complex emotion in text, and computational psychotherapy applications.
2019.09.17
View 5773
Two More Cross-generation Collaborative Labs Open
< President Sung-Chul Shin (sixth from the left) and Professor Sun Chang Kim (seventh from the left) at the signboard ceremony of KAIST BioDesigneering Laboratory > KAIST opened two more cross-generation collaborative labs last month. KAIST BioDesigneering Laboratory headed by Professor Sun Chang Kim from the Department of Biological Sciences and Nanophotonics Laboratory led by Professor Yong-Hee Lee from the Department of Physics have been selected to receive 500 million KRW funding for five years. A four-member selection committee including the former President of ETH Zürich Professor Emeritus Ralph Eichler and Professor Kwang-Soo Kim of Harvard Medical School conducted a three-month review and evaluation for this selection to be made. With these two new labs onboard, a total of six cross-generation collaborative labs will be operated on campus. The operation of cross-generation collaborative labs has been in trial since March last year, as one of the KAIST’s Vision 2031 research innovation initiatives. This novel approach is to pair up senior and junior faculty members for sustaining research and academic achievements even after the senior researcher retires, so that the spectrum of knowledge and research competitiveness can be extended to future generations. The selected labs will be funded for five years, and the funding will be extended if necessary. KAIST will continue to select new labs every year. One of this year’s selectees Professor Sun Chang Kim will be teamed up with Professor Byung-Kwan Cho from the same department and Professor Jung Kyoon Choi from the Department of Bio and Brain Engineering to collaborate in the fields of synthetic biology, systems biology, and genetic engineering. This group mainly aims at designing and synthesizing optimal genomes that can efficiently manufacture protein drug and biomedical active materials. They will also strive to secure large amounts of high-functioning natural active substances, new adhesive antibacterial peptides, and eco-friendly ecological restoration materials. It is expected that collaboration between these three multigenerational professors will help innovate their bio-convergence technology and further strengthen their international competitiveness in the global bio-market. Another world-renowned scholar Professor Yong-Hee Lee of photonic crystal laser study will be joined by Professor Minkyo Seo from the same department and Professor Hansuek Lee from the Graduate School of Nanoscience and Technology. They will explore the extreme limits of light-material interaction based on optical micro/nano resonators, with the goal of developing future nonlinear optoelectronic and quantum optical devices. The knowledge and technology newly gained from the research are expected to provide an important platform for a diverse range of fields from quantum communications to biophysics. (END)
2019.09.06
View 7210
Enhanced Natural Gas Storage to Help Reduce Global Warming
< Professor Atilhan (left) and Professor Yavuz (right) > Researchers have designed plastic-based materials that can store natural gas more effectively. These new materials can not only make large-scale, cost-effective, and safe natural gas storage possible, but further hold a strong promise for combating global warming. Natural gas (predominantly methane) is a clean energy alternative. It is stored by compression, liquefaction, or adsorption. Among these, adsorbed natural gas (ANG) storage is a more efficient, cheaper, and safer alternative to conventional compressed natural gas (CNG) and liquefied natural gas (LNG) storage approaches that have drawbacks such as low storage efficiency, high costs, and safety concerns. However, developing adsorptive materials that can more fully exploit the advantages of ANG storage has remained a challenging task. A KAIST research team led by Professor Cafer T. Yavuz from the Graduate School of Energy, Environment, Water, and Sustainability (EEWS), in collaboration with Professor Mert Atilhan’s group from Texas A&M University, synthesized 29 unique porous polymeric structures with inherent flexibility, and tested their methane gas uptake capacity at high pressures. These porous polymers had varying synthetic complexities, porosities, and morphologies, and the researchers subjected each porous polymer to pure methane gas under various conditions to study the ANG performances. Of these 29 distinct chemical structures, COP-150 was particularly noteworthy as it achieved a high deliverable gravimetric methane working capacity when cycled between 5 and 100 bar at 273 K, which is 98% of the total uptake capacity. This result surpassed the target set by the United States Department of Energy (US DOE). COP-150 is the first ever structure to fulfil both the gravimetric and volumetric requirements of the US DOE for successful vehicular use, and the total cost to produce the COP-150 adsorbent was only 1 USD per kilogram. COP-150 can be produced using freely available and easily accessible plastic materials, and moreover, its synthesis takes place at room temperature, open to the air, and no previous purification of the chemicals is required. The pressure-triggered flexible structure of COP-150 is also advantageous in terms of the total working capacity of deliverable methane for real applications. The research team believed that the increased pressure flexes the network structure of COP-150 showing “swelling” behavior, and suggested that the flexibility provides rapid desorption and thermal management, while the hydrophobicity and the nature of the covalently bonded framework allow these promising materials to tolerate harsh conditions. This swelling mechanism of expansion-contraction solves two other major issues, the team noted. Firstly, when using adsorbents based on such a mechanism, unsafe pressure spikes that may occur due to temperature swings can be eliminated. In addition, contamination can also be minimized, since the adsorbent remains contracted when no gas is stored. Professor Yavuz said, “We envision a whole host of new designs and mechanisms to be developed based on our concept. Since natural gas is a much cleaner fuel than coal and petroleum, new developments in this realm will help switching to the use of less polluting fuels.” Professor Atilhan agreed the most important impact of their research is on the environment. “Using natural gas more than coal and petroleum will significantly reduce greenhouse gas emissions. We believe, one day, we might see vehicles equipped with our materials that are run by a cleaner natural gas fuel,” he added. This study, reported in Nature Energy on July 8, was supported by National Research Foundation of Korea (NRF) grants ( NRF-2016R1A2B4011027, NRF-2017M3A7B4042140, and NRF-2017M3A7B4042235). < Suggested chemical structure of COP-150 > < Initial ingredients (left) and final product (right) of COP-150 synthesis > < Comparison of highest reported volumetric working capacities > (END)
2019.08.09
View 24475
Deep Learning-Powered 'DeepEC' Helps Accurately Understand Enzyme Functions
(Figure: Overall scheme of DeepEC) A deep learning-powered computational framework, ‘DeepEC,’ will allow the high-quality and high-throughput prediction of enzyme commission numbers, which is essential for the accurate understanding of enzyme functions. A team of Dr. Jae Yong Ryu, Professor Hyun Uk Kim, and Distinguished Professor Sang Yup Lee at KAIST reported the computational framework powered by deep learning that predicts enzyme commission (EC) numbers with high precision in a high-throughput manner. DeepEC takes a protein sequence as an input and accurately predicts EC numbers as an output. Enzymes are proteins that catalyze biochemical reactions and EC numbers consisting of four level numbers (i.e., a.b.c.d) indicate biochemical reactions. Thus, the identification of EC numbers is critical for accurately understanding enzyme functions and metabolism. EC numbers are usually given to a protein sequence encoding an enzyme during a genome annotation procedure. Because of the importance of EC numbers, several EC number prediction tools have been developed, but they have room for further improvement with respect to computation time, precision, coverage, and the total size of the files needed for the EC number prediction. DeepEC uses three convolutional neural networks (CNNs) as a major engine for the prediction of EC numbers, and also implements homology analysis for EC numbers if the three CNNs do not produce reliable EC numbers for a given protein sequence. DeepEC was developed by using a gold standard dataset covering 1,388,606 protein sequences and 4,669 EC numbers. In particular, benchmarking studies of DeepEC and five other representative EC number prediction tools showed that DeepEC made the most precise and fastest predictions for EC numbers. DeepEC also required the smallest disk space for implementation, which makes it an ideal third-party software component. Furthermore, DeepEC was the most sensitive in detecting enzymatic function loss as a result of mutations in domains/binding site residue of protein sequences; in this comparative analysis, all the domains or binding site residue were substituted with L-alanine residue in order to remove the protein function, which is known as the L-alanine scanning method. This study was published online in the Proceedings of the National Academy of Sciences of the United States of America (PNAS) on June 20, 2019, entitled “Deep learning enables high-quality and high-throughput prediction of enzyme commission numbers.” “DeepEC can be used as an independent tool and also as a third-party software component in combination with other computational platforms that examine metabolic reactions. DeepEC is freely available online,” said Professor Kim. Distinguished Professor Lee said, “With DeepEC, it has become possible to process ever-increasing volumes of protein sequence data more efficiently and more accurately.” This work was supported by the Technology Development Program to Solve Climate Changes on Systems Metabolic Engineering for Biorefineries from the Ministry of Science and ICT through the National Research Foundation of Korea. This work was also funded by the Bio & Medical Technology Development Program of the National Research Foundation of Korea funded by the Korean government, the Ministry of Science and ICT. Profile: -Professor Hyun Uk Kim (ehukim@kaist.ac.kr) https://sites.google.com/view/ehukim Department of Chemical and Biomolecular Engineering -Distinguished Professor Sang Yup Lee (leesy@kaist.ac.kr) Department of Chemical and Biomolecular Engineering http://mbel.kaist.ac.kr
2019.07.09
View 32413
Wearable Robot 'WalkON Suit' Off to Cybathlon 2020
Standing upright and walking alone are very simple but noble motions that separate humans from many other creatures. Wearable and prosthetic technologies have emerged to augment human function in locomotion and manipulation. However, advances in wearable robot technology have been especially momentous to Byoung-Wook Kim, a triplegic for 22 years following a devastating car accident. Kim rejoiced after standing upright and walking again by putting on the ‘WalkON Suit,’ the wearable robot developed by Professor Kyoungchul Kong’s team. Even more, Kim won third prize in the powered exoskeleton race at Cybathlon 2016, an international cyborg Olympics hosted by ETH Zurich. Now Kim and Professor Kong’s team are all geared up for the Cybathlon Championship 2020. Professor Kong and his startup, Angel Robotics, held a kickoff ceremony for Cybathlon 2020 at KAIST on June 24. The 2020 championship will take place in Switzerland. Only pilots with complete paralysis of the legs resulting from spinal cord injuries are eligible to participate in the Cybathlon, which takes place every four years. Pilots compete against each other while completing everyday tasks using technical assistance systems in six different disciplines: a brain-computer interface race, a functional electrical stimulation bike race, a powered arm prosthesis race, a powered leg prosthesis race, a powered exoskeleton race, and a powered wheelchair race. The 2016 championship drew 66 pilots from 56 teams representing 25 countries. In the powered exoskeleton race, pilots complete everyday activities such as getting up from a sofa and overcoming obstacles such as stairs, ramps, or slopes and up to four pilots compete simultaneously on tracks to solve six tasks; and the pilot that solves the most tasks in the least amount of time wins the race. (Kim, a triplegic for 22 years demonstrates walking and climbing the stairs (below photo) wearing the WalkOn Suit during the media day on June 21 at KAIST.) Kim, who demonstrated walking and climbing the stairs wearing the WalkON Suit during the media day for the Cybathlon 2020 kickoff ceremony on June 21 at KAIST, said, “I have been confined to a wheelchair for more than 20 years. I am used to it so I feel like the wheelchair is one of my body parts. Actually, I don’t feel any big difficulties in doing everyday tasks in wheelchair. But whenever I face the fact that I will never be able to stand up with my own two legs again, I am so devastated.” He continued, “I still remember the day when I stood up with my own two legs by myself after 22 years. It was beyond description.” The market for wearable robots, especially for exoskeleton robots, is continuing to grow as the aging population has been a major challenge in almost every advanced country. The global market for these robots expects to see annual growth of 41.2% to 8.3 billion US dollars by 2025. Healthcare wearable robots for the elderly and rehabilitation take up the half of the market share followed by wearable robots for industrial and defense purposes. Professor Kong from the Department of Mechanical Engineering and his colleagues have developed two wearable robot systems in 2014: The "WalkON Suit" for complete paraplegics and “Angel Suit” for those with partial impairment in walking ability such as the elderly and rehabilitation patients. Professor Kong said after 15 years of basic research, the team is now able to develop its own distinct technologies. He said their robots are powered by non-resistant precision drives with algorithms recognizing the user’s moving intention. Incorporated with prosthetic devices technology from the Severance Rehabilitation Hospital, their control technology has led to the production of a customizable robot suit optimized for each user’s physical condition. The WalkON Suit, which boasts a maximum force of 250 Nm and maximum rotation speed of 45 RPM, gives the user high-energy efficiency modeled after the physiology of the human leg. It allows users to walk on flat ground and down stairs, climb up and down inclines, and sit and lie down. Currently the battery lasts five to six hours for locomotion and the approximate 25 kg of robot weight still remains a technical challenge to upgrade. Professor Kong’s team has grafted AR glass technology into the WalkOn Suit that one of his pilots put on for the torch relay of the PyongChang Paralympics in 2018. His team is now upgrading the WalkON Suit 4.0 for next year’s competition. Severance Rehabilitation Hospital will help the seven pilots with their training. Professor Kong said his goal is to make robots that can make people with disabilities much more independent. He stressed, “Wearable robots should be designed for each single user. We provide a very good graphical user interface so that we can design, check, and also verify our optimized design for users’ best performance.” (Seven pilots and Professor Kong (fifth from left in second row) pose with guests who joined the Cybathlon 2020 kickoff ceremony. President Shin (fifth from right) made a congratulatory remarks during the ceremony.)
2019.06.25
View 36611
Efficiently Producing Fatty Acids and Biofuels from Glucose
Researchers have presented a new strategy for efficiently producing fatty acids and biofuels that can transform glucose and oleaginous microorganisms into microbial diesel fuel, with one-step direct fermentative production. The newly developed strain, created by Distinguished Professor Sang Yup Lee and his team, showed the highest efficiency in producing fatty acids and biodiesels ever reported. It will be expected to serve as a new platform to sustainably produce a wide array of fatty acid-based products from glucose and other carbon substrates. Fossil fuels, which have long been energy resources for our daily lives, are now facing serious challenges: depletion of their reserves and their role in global warming. The production of sustainable bio-based renewable energy has emerged as an essential alternative and many studies to replace fossil fuels are underway. One of the representative examples is biodiesel. Currently, it is mainly being produced through the transesterification of vegetable oils or animal fats. The research team engineered oleaginous microorganisms, Rhodococcus opacus, to produce fatty acids and their derivatives that can be used as biodiesel from glucose, one of the most abundant and cheap sugars derived from non-edible biomass. Professor Lee’s team has already engineered Escherichia coli to produce short-chain hydrocarbons, which can be used as gasoline (published in Nature as the cover paper in 2013). However, the production efficiency of the short-chain hydrocarbons using E. coli (0.58 g/L) fell short of the levels required for commercialization. To overcome these issues, the team employed oil-accumulating Rhodococcus opacus as a host strain in this study. First, the team optimized the cultivation conditions of Rhodococcus opacus to maximize the accumulation of oil (triacylglycerol), which serves as a precursor for the biosynthesis of fatty acids and their derivatives. Then, they systematically analyzed the metabolism of the strain and redesigned it to enable higher levels of fatty acids and two kinds of fatty acid-derived biodiesels (fatty acid ethyl esters and long-chain hydrocarbons) to be produced. They found that the resulting strains produced 50.2, 21.3, and 5.2 g/L of fatty acids, fatty acid ethyl esters, and long-chain hydrocarbons, respectively. These are all the highest concentrations ever reported by microbial fermentations. It is expected that these strains can contribute to the future industrialization of microbial-based biodiesel production. “This technology creates fatty acids and biodiesel with high efficiency by utilizing lignocellulose, one of the most abundant resources on the Earth, without depending on fossil fuels and vegetable or animal oils. This will provide new opportunities for oil and petroleum industries, which have long relied on fossil fuels, to turn to sustainable and eco-friendly biotechnologies,” said Professor Lee. This paper titled “Engineering of an oleaginous bacterium for the production of fatty acids and fuels” was published in Nature Chemical Biology on June 17. This work was supported by the Technology Development Program to Solve Climate Changes on Systems Metabolic Engineering for Biorefineries from the Ministry of Science and ICT through the National Research Foundation (NRF) of Korea (NRF-2012M1A2A2026556 and NRF-2012M1A2A2026557). (Figure: Metabolic engineering for the production of free fatty acids (FFAs), fatty acid ethyl esters (FAEEs), and long-chain hydrocarbons (LCHCs) in Rhodococcus opacus PD630. Researchers have presented a new strategy for efficiently producing fatty acids and biofuels that can transform glucose and oleaginous microorganisms into microbial diesel fuel, with one-step direct fermentative production.) # # # Source: Hye Mi Kim, Tong Un Chae, So Young Choi, Won Jun Kim and Sang Yup Lee. Engineering of an oleaginous bacterium for the production of fatty acids and fuels. Nature Chemical Biology ( https://www.nature.com/nchembio/ ) DOI: 10.1038/s41589-019-0295-5 Profile Dr. Sang Yup Lee leesy@kaist.ac.kr Distinguished Professor at the Department of Chemical and Biomolecular Engineering KAIST
2019.06.19
View 44363
Engineered Microbial Production of Grape Flavoring
(Image 1: Engineered bacteria that produce grape flavoring.) Researchers report a microbial method for producing an artificial grape flavor. Methyl anthranilate (MANT) is a common grape flavoring and odorant compound currently produced through a petroleum-based process that uses large volumes of toxic acid catalysts. Professor Sang-Yup Lee’s team at the Department of Chemical and Biomolecular Engineering demonstrated production of MANT, a naturally occurring compound, via engineered bacteria. The authors engineered strains of Escherichia coli and Corynebacetrium glutamicum to produce MANT through a plant-based engineered metabolic pathway. The authors tuned the bacterial metabolic pathway by optimizing the levels of AAMT1, the key enzyme in the process. To maximize production of MANT, the authors tested six strategies, including increasing the supply of a precursor compound and enhancing the availability of a co-substrate. The most productive strategy proved to be a two-phase extractive culture, in which MANT was extracted into a solvent. This strategy produced MANT on the scale of 4.47 to 5.74 grams per liter, a significant amount, considering that engineered microbes produce most natural products at a scale of milligrams or micrograms per liter. According to the authors, the results suggest that MANT and other related molecules produced through industrial processes can be produced at scale by engineered microbes in a manner that would allow them to be marketed as natural one, instead of artificial one. This study, featured at the Proceeding of the National Academy of Sciences of the USA on May 13, was supported by the Technology Development Program to Solve Climate Changes on Systems Metabolic Engineering for Biorefineries from the Ministry of Science and ICT. (Image 2. Overview of the strategies applied for the microbial production of grape flavoring.)
2019.05.15
View 51172
Professor Park at UPC-Barcelona Tech Receives Jeong Hun Cho Award
Professor Hyuk Park was honored to be the recipient of the Jeong Hun Cho Award which was presented at the Universitat Politècnica de Catalunya Barcelona Tech. The award recognizes young scientists in the field of aerospace engineering. Professor Park, a graduate of KAIST’s Department of Mechanical Engineering in 2001, earned his MS and PhD at the Gwangju Institute of Science and Technology, and works at the Castelldefels School of Telecommunications and Aerospace Engineering at UPC-Barcelona Tech. He won this year’s award, which honors former PhD candidate Jeong Hun Cho at the Aerospace Engineering Department who died in a lab accident in 2003. Professor Park also received 25 million KRW prize money. Cho’s family endowed the award and scholarship in his memory. Since 2005, the scholarship has selected three young scholars every year who specialize in aerospace engineering from Cho’s alma maters of KAIST, Korea University, and Kongju National University High School. Professor Park was selected as this year’s awardee in recognition of his studies of synthetic-aperture radar (SAR) satellite radiometer system, remote sensing radio frequency interference reduction system development, and 3CAT series research. The Award Committee also chose three students for scholarships: PhD candidate Sang-Woo Chung from the Department of Aerospace Engineering at KAIST with 4 million KRW, PhD candidate Eun-Hee Kang from the School of Mechanical Engineering at Korea University with 4 million KRW, and Chan-Ho Song from Kongju National University High School with 3 million KRW.
2019.05.14
View 5158
KAIST Identifies the Cause of Sepsis-induced Lung Injury
(Professor Pilhan Kim from the Graduate School of Medical Science and Engineering) A KAIST research team succeeded in visualizing pulmonary microcirculation and circulating cells in vivo with a custom-built 3D intravital lung microscopic imaging system. They found a type of leukocyte called neutrophils aggregate inside the capillaries during sepsis-induced acute lung injury (ALI), leading to disturbances and dead space in blood microcirculation. According to the researchers, this phenomenon is responsible for tissue hypoxia causing lung damage in the sepsis model, and mitigating neutrophils improves microcirculation as well as hypoxia. The lungs are responsible for exchanging oxygen with carbon dioxide gases during the breathing process, providing an essential function for sustaining life. This gas exchange occurs in the alveoli, each surrounded by many capillaries containing the circulating red blood cells. Researchers have been making efforts to observe microcirculation in alveoli, but it has been technically challenging to capture high-resolution images of capillaries and red blood cells inside the lungs that are in constant breathing motion. Professor Pilhan Kim from the Graduate School of Medical Science and Engineering and his team developed an ultra-fast laser scanning confocal microscope and an imaging chamber that could minimize the movement of a lung while preserving its respiratory state. They used this technology to successfully capture red blood cell circulation inside the capillaries of animal models with sepsis. During the process, they found that hypoxia was induced by the increase of dead space inside the lungs of a sepsis model, a space where red blood cells do not circulate. This phenomenon is due to the neutrophils aggregating and trapping inside the capillaries and the arterioles. It was also shown that trapped neutrophils damage the lung tissue in the sepsis model by inhibiting microcirculation as well as releasing reactive oxygen species. Further studies showed that the aggregated neutrophils inside pulmonary vessels exhibit a higher expression of the Mac-1 receptor (CD11b/CD18), which is a receptor involved in intercellular adhesion, compared to the neutrophils that normally circulate. Additionally, they confirmed that Mac-1 inhibitors can improve inhibited microcirculation, ameliorate hypoxia, while reducing pulmonary edema in the sepsis model. Their high-resolution 3D intravital microscope technology allows the real-time imaging of living cells inside the lungs. This work is expected to be used in research on various lung diseases, including sepsis. The research team’s pulmonary circulation imaging and precise analytical techniques will be used as the base technology for developing new diagnostic technologies, evaluating new therapeutic agents for various diseases related to microcirculation. Professor Kim said, “In the ALI model, the inhibition of pulmonary microcirculation occurs due to neutrophils. By controlling this effect and improving microcirculation, it is possible to eliminate hypoxia and pulmonary edema – a new, effective strategy for treating patients with sepsis.” Their 3D intravital microscope technology was commercialized through IVIM Technology, Inc., which is a faculty startup at KAIST. They released an all-in-one intravital microscope model called ‘IVM-CM’ and ‘IVM-C’. This next-generation imaging equipment for basic biomedical research on the complex pathophysiology of various human diseases will play a crucial role in the future global bio-health market. This research, led by Dr. Inwon Park from the Department of Emergency Medicine at Seoul National University Bundang Hospital and formally the Graduate School of Medical Science and Engineering at KAIST, was published in the European Respiratory Journal (2019, 53:1800736) on March 28, 2019. Figure 1. Custom-built high-speed real-time intravital microscope platform Figure 2. Illustrative schematic and photo of a 3D intravital lung microscopic imaging system Figure 3. Aggregation of neutrophils and consequent flow disturbance in pulmonary arteriole in sepsis-induced lung injury
2019.05.07
View 42582
KAIST Unveils the Hidden Control Architecture of Brain Networks
(Professor Kwang-Hyun Cho and his team) A KAIST research team identified the intrinsic control architecture of brain networks. The control properties will contribute to providing a fundamental basis for the exogenous control of brain networks and, therefore, has broad implications in cognitive and clinical neuroscience. Although efficiency and robustness are often regarded as having a trade-off relationship, the human brain usually exhibits both attributes when it performs complex cognitive functions. Such optimality must be rooted in a specific coordinated control of interconnected brain regions, but the understanding of the intrinsic control architecture of brain networks is lacking. Professor Kwang-Hyun Cho from the Department of Bio and Brain Engineering and his team investigated the intrinsic control architecture of brain networks. They employed an interdisciplinary approach that spans connectomics, neuroscience, control engineering, network science, and systems biology to examine the structural brain networks of various species and compared them with the control architecture of other biological networks, as well as man-made ones, such as social, infrastructural and technological networks. In particular, the team reconstructed the structural brain networks of 100 healthy human adults by performing brain parcellation and tractography with structural and diffusion imaging data obtained from the Human Connectome Project database of the US National Institutes of Health. The team developed a framework for analyzing the control architecture of brain networks based on the minimum dominating set (MDSet), which refers to a minimal subset of nodes (MD-nodes) that control the remaining nodes with a one-step direct interaction. MD-nodes play a crucial role in various complex networks including biomolecular networks, but they have not been investigated in brain networks. By exploring and comparing the structural principles underlying the composition of MDSets of various complex networks, the team delineated their distinct control architectures. Interestingly, the team found that the proportion of MDSets in brain networks is remarkably small compared to those of other complex networks. This finding implies that brain networks may have been optimized for minimizing the cost required for controlling networks. Furthermore, the team found that the MDSets of brain networks are not solely determined by the degree of nodes, but rather strategically placed to form a particular control architecture. Consequently, the team revealed the hidden control architecture of brain networks, namely, the distributed and overlapping control architecture that is distinct from other complex networks. The team found that such a particular control architecture brings about robustness against targeted attacks (i.e., preferential attacks on high-degree nodes) which might be a fundamental basis of robust brain functions against preferential damage of high-degree nodes (i.e., brain regions). Moreover, the team found that the particular control architecture of brain networks also enables high efficiency in switching from one network state, defined by a set of node activities, to another – a capability that is crucial for traversing diverse cognitive states. Professor Cho said, “This study is the first attempt to make a quantitative comparison between brain networks and other real-world complex networks. Understanding of intrinsic control architecture underlying brain networks may enable the development of optimal interventions for therapeutic purposes or cognitive enhancement.” This research, led by Byeongwook Lee, Uiryong Kang and Hongjun Chang, was published in iScience (10.1016/j.isci.2019.02.017) on March 29, 2019. Figure 1. Schematic of identification of control architecture of brain networks. Figure 2. Identified control architectures of brain networks and other real-world complex networks.
2019.04.23
View 36527
KAIST Develops Analog Memristive Synapses for Neuromorphic Chips
(Professor Sung-Yool Choi from the School of Electrical Engineering) A KAIST research team developed a technology that makes a transition of the operation mode of flexible memristors to synaptic analog switching by reducing the size of the formed filament. Through this technology, memristors can extend their role to memristive synapses for neuromorphic chips, which will lead to developing soft neuromorphic intelligent systems. Brain-inspired neuromorphic chips have been gaining a great deal of attention for reducing the power consumption and integrating data processing, compared to conventional semiconductor chips. Similarly, memristors are known to be the most suitable candidate for making a crossbar array which is the most efficient architecture for realizing hardware-based artificial neural network (ANN) inside a neuromorphic chip. A hardware-based ANN consists of a neuron circuit and synapse elements, the connecting pieces. In the neuromorphic system, the synaptic weight, which represents the connection strength between neurons, should be stored and updated as the type of analog data at each synapse. However, most memristors have digital characteristics suitable for nonvolatile memory. These characteristics put a limitation on the analog operation of the memristors, which makes it difficult to apply them to synaptic devices. Professor Sung-Yool Choi from the School of Electrical Engineering and his team fabricated a flexible polymer memristor on a plastic substrate, and found that changing the size of the conductive metal filaments formed inside the device on the scale of metal atoms can make a transition of the memristor behavior from digital to analog. Using this phenomenon, the team developed flexible memristor-based electronic synapses, which can continuously and linearly update synaptic weight, and operate under mechanical deformations such as bending. The team confirmed that the ANN based on these memristor synapses can effectively classify person’s facial images even when they were damaged. This research demonstrated the possibility of a neuromorphic chip that can efficiently recognize faces, numbers, and objects. Professor Choi said, “We found the principles underlying the transition from digital to analog operation of the memristors. I believe that this research paves the way for applying various memristors to either digital memory or electronic synapses, and will accelerate the development of a high-performing neuromorphic chip.” In a joint research project with Professor Sung Gap Im (KAIST) and Professor V. P. Dravid (Northwestern University), this study was led by Dr. Byung Chul Jang (Samsung Electronics), Dr. Sungkyu Kim (Northwestern University) and Dr. Sang Yoon Yang (KAIST), and was published online in Nano Letters (10.1021/acs.nanolett.8b04023) on January 4, 2019. Figure 1. a) Schematic illustration of a flexible pV3D3 memristor-based electronic synapse array. b) Cross-sectional TEM image of the flexible pV3D3 memristor
2019.02.28
View 6990
<<
첫번째페이지
<
이전 페이지
1
2
3
4
5
6
7
8
9
10
>
다음 페이지
>>
마지막 페이지 24