\n<\/iframe>\n<\/div><\/figure>\n\n\n\nBOSTON (Nov. 14, 2020) \u2013 Psychiatrists are faced with the difficult task of making diagnoses without having any physical evidence to draw upon every day. No blood tests or biological assessment of any kind can produce conclusive evidence that one has mental illness, let alone that it\u2019s in development.\u00a0\u00a0\u00a0<\/p>\nResearchers from around the world are trying to solve this problem by tapping into the power of data and machine learning. Among them is Prof. Kiyoto Kasai of The University of Tokyo\u2019s Neuropsychiatry Department, who presented his lecture at the fifth Japan-U.S. Science Forum in Boston held online on Nov. 14, 2020. Prof. Kasai has been using artificial intelligence (AI) for meta-analysis of neuroimaging data in his research on psychosis in adolescence, and has discovered that the volume of the globus palladium increases in schizophrenic patients. In the meantime, Dr. Mitsuo Kawato, the other Science Forum lecturer and the director of Advanced Telecommunications Research Institute in Kyoto, used AI to identify distinct brain network patterns in patients suffering from mental disorders.\u00a0<\/p>\n\u201cBig-data science has definitely arrived in the neuroscience field,\u201d said Harvard University professor Takao Hensch, as he made an introductory presentation at the forum.\u00a0<\/p>\nHensch, who is a professor of Molecular Cellular Biology at Harvard\u2019s Center for Brain Science, pointed out how enormously complex the interworking of microcircuits is, given that plasticity and dynamics of circuits and various other factors all come into play.\u00a0<\/p>\n\u201cWe need computational approaches to tackle this information overload, as well as to come up with new principles that are involved in brain development and brain disorders,\u201d Prof. Hensch said.\u00a0<\/p>\nFor the fifth year, the Science Forum brought the Japanese and U.S. science communities together. Though held virtually due to concerns related to new coronavirus, the forum, which took on the theme of \u201cAI for Clinical Translational Research,\u201d provided the opportunity for researchers from both countries and elsewhere to meet and share inspiration via the event platform.\u00a0<\/p>\nJointly organized by the Japan Society for the Promotion of Science (JSPS)\u2019s Washington Office, the Consulate General of Japan in Boston, and the United Japanese Researchers Around the World (UJA), the forum began in 2016 as a way to showcase Japan\u2019s scientific undertakings and promote global partnerships. UJA President Takeya Adachi said his personal experience of promoting translational research on rare genetic disorders has shown the importance of global sharing of data and expertise. Both Mr. Setsuo Ohmori, Consul General of Japan in Boston, and Mark Eliott, vice provost of Harvard University, said the significance of transnational collaborations is becoming more apparent than ever amid the global pandemic.\u00a0<\/p>\n\u201cThe researchers who are here today are all working on powerful, promising tools for tackling the challenges of today\u2019s world,\u201d Mr. Ohmori said.\u00a0<\/p>\nCalling such collective effort a \u201cglobal research enterprise,\u201d Dr. Eliott said the enterprise is \u201cheld together by the intellectuals and personal connections that meetings such as this one today make possible.\u201d\u00a0<\/p>\nDr. Kohji Hirata, the director of the JSPS Washington Office, thanked all involved in the organizing of the forum for making the virtual event happen.<\/p>\nHarnessing the power of AI to take neuroscience to a\u00a0 whole new level<\/strong><\/span><\/p>\nThe past year has seen the Covid-19 pandemic turn into a global mental health crisis. Neuropsychiatric disorders had already accounted for more than 28 percent of work productivity loss in one\u2019s lifetime due to diseases in North America prior to the new coronavirus outbreaks, Prof. Hensch said. With the increasing societal burden of mental health comes the growing desire among researchers to understand when and how illness and disorders begin to develop in a person.\u00a0<\/p>\n\u201cWe think that mental illnesses have origins in early life no matter how late in life they manifest themselves,\u201d Prof. Hensch said, adding that people\u2019s behaviors as adults are \u201csculpted during the critical periods of early brain development, leading from even before birth through adolescence.\u201d\u00a0<\/p>\nResearchers are taking advantage of recent technological advancements to understand how neural circuits are wired in healthy individuals.\u00a0<\/p>\n\u201cThis is also an opportunity to learn the principles of next-generation artificial intelligence, which might actually leverage some of the ideas and create the possibility for super intelligent behaviors, which might catapult our human evolution and human studies,\u201d Prof. Hensch said.\u00a0<\/p>\nDr. Kasai discovers translatable biomarkers to understand the pathophysiology of psychosis onset in adolescence<\/strong><\/span><\/p>\nLoss of the excitation and inhibition balance, or \u201ce\/i balance,\u201d in the brain is believed to play a role in schizophrenia, which often presents itself in adolescence. However, the exact pathophysiology that prompts the illness\u2019s onset largely remains a mystery. And Prof. Kasai has been trying to solve that problem for more than two decades.\u00a0<\/p>\n\u00a0In his early research at Harvard, he discovered that those who had just experienced their first psychosis episodes had less gray matter in their brains, compared to healthy individuals and patients of affective disorders — which include depression and bipolar. He also found that the gray matter volume \u201cprogressively decreased\u201d over 1.5 years thereafter. This gray matter volume reduction is associated with auditory hallucinations, according to Prof. Kasai.\u00a0<\/p>\nProf. Kasai\u2019s recent findings include progressive reductions in mismatched negativity (MMN) in patients suffering from schizophrenia. Reduced MMN means that one\u2019s ability to process auditory sensory information is impaired. His further research revealed that MMN had been already reduced — and the plasma glutamate level was elevated — in people with schizophrenia before they ever had their first episodes of psychosis. In addition, \u201c40Hz auditory steady state response\u201d (AASR) — the brain\u2019s response to auditory stimuli that is often used to test patients with schizophrenia — was also \u201csignificantly reduced\u201d in patients in the early stages of psychosis.\u00a0<\/p>\nTo prove all these factors are interrelated, Prof. Kasai used marmosets — which have a large frontal cortex and exhibit similar social behaviors as humans — for his studies. By using various imaging tools, including electrocorticography (ECoG) and 2-photon calcium imaging, he was able to locate exactly in which parts of the brain MMN and AASR occur, and also discovered, through cohort studies, that the globus pallidus volume increases in schizophrenia, including adolescents with psychotic symptoms. These findings not only indicated that increased globus pallidus volume can be a biomarker, but also enabled Prof. Kasai to hypothesize that \u201caccumulation of dopamine D-2 receptor dysfunction by social stress overload\u201d may prompt the development of psychosis. He is now conducting research based on that hypothesis by using mice.\u00a0<\/p>\nAll of these \u201ctranslatable\u201d biomarkers that are common among both humans and animals can help researchers develop early interventions for psychosis, he said.\u00a0<\/p>\nDr. Kawato applies the mechanism of learning in the human brain to computational psychiatry<\/strong><\/span><\/p>\nDr. Kawato\u2019s latest research began with a simple question: How do human brains learn so much from a small sample of information?\u00a0<\/p>\nFor example, his grandson learned how to get down from a couch and stand on the floor after trying it about 10 times when he was 9 months old, Dr. Kawato said. It would take humanoid robots thousands of trials and errors to master such motor skills.\u00a0<\/p>\nIn machine learning, when the complexity — or the number of parameters — for a learning model grows beyond a certain point, the margin of \u201cgeneralization error\u201d begins to increase, because a larger dataset contains more variables, which AI takes into account for generalization. In theory, Prof. Kawato said, when data size remains the same, the only way to archive lower margins of generalization error is to reduce the \u201cdegrees of freedom,\u201d or the complexity.\u00a0<\/p>\nIn order to see how the human brain circumvents this \u201ccurse of dimensionality\u201d to learn so much from so little experience, Dr. Kawato and his team conducted a reinforcement learning experiment in which they used imaging technology and AI to monitor the participants\u2019 brain activities. Because some people argue that humans may be genetically wired to master motor skills quickly, Prof. Kawato\u2019s team made sure to give the participants a learning task that couldn\u2019t be helped by humans\u2019 innate capabilities. The result showed that the more the participants learned, the higher their confidence level became in performing the task. In other words, the dorsolateral prefrontal cortex (dlPFC), an area responsible for cognition, knew that the new knowledge had already been attained in the basal ganglia. This means metacognition helps dimension reduction, he said.\u00a0<\/p>\nDr. Kawato and his team then built a large database of multi-disorder patients and successfully identified synchronized circuitry connections that can serve as biomarkers for autism spectrum disorder and depression. They developed \u201cDecoded Neurofeedback (DecNef),\u201d a process of correcting a patient\u2019s unwanted brain activities through the use of perceptual learning. For example, whenever a photo shown to a phobia patient triggers an fMRI signal pattern indicating fear in their brain, the research team would give them a small reward. Through this manipulation, or \u201cneural feedback,\u201d that region of the brain will eventually begin to show a different signal pattern in response to the image, and the patient will no longer feel fearful. The patient would be unaware of the purpose of the exercise throughout the process. This method is now being applied to PTSD therapy, Dr. Kawato said.\u00a0<\/p>\n\u201cPsychiatry has not made great progress in the last 30 years and needs help from neuroscience and AI,\u201d Dr. Kawato said. \u201cCoded neurofeedback for causal human neuroscience can also be used as a tool for data-driven and precision-medicine intervention.\u201d\u00a0<\/p>\nVirtual poster session, \u2018Flash Talk,\u2019 and networking session brings together the global science community\u00a0<\/strong><\/span><\/p>\nThis year\u2019s poster session was held virtually with about 30 scientists presenting their research work that included a wide range of topics, from COVID-19 to microgravity to climate change. The \u201cFlash Talk Audience Award\u201d went to the following three presenters who demonstrated presentation excellence, scientific novelty and excellence, and interdisciplinary impact.\u00a0<\/p>\n<ul>\n<li>Dr. Hayami Koga, Harvard T.H. Chan School of Public Health (\u201cWell-being and health\u201d)<\/li>\n<li>Dr. Man-Yin Tsang, the University of Toronto and Kobe University (\u201cFukushima Dai-Ichi nuclear power plants- a view from the ocean\u201d)<\/li>\n<li>Dr. Mayuka Nakajima, Harvard University SEAS (\u201cTopical application of choline and geranic acid (CAGE) based ionic liquid for the treatment of oral infectious disease\u201d)<\/li>\n<\/ul>\nIn the Networking Session that followed the Flash Talk, the attendees enjoyed group conversations.\u00a0<\/p>\n\u201cThe forum gave me the opportunity to meet other fellows in JSPS\u2019s Restart Research Abroad program. Getting to know them was an empowering experience for which I am grateful,\u201d said Dr. Kimino Fujimura, a neurobiology research fellow at Boston Children’s Hospital\/Harvard Medical School.\u00a0<\/p>\nMr. Kashin Sugishita, post-doc at the State University of New York at Buffalo, said the forum enabled him to learn about scientific research in a wide range of research fields.\u00a0<\/p>\n\u201cBeing able to connect and talk with researchers from all across the U.S. in a casual atmosphere while working from home was an especially\u00a0 stimulating experience,\u201d Mr. Sugishita said.\u00a0<\/p>\nBoth Dr. Fujimura and Mr. Sugishita participated in the Flash Talk, showcasing their research on the immune response in murine Zika Virus microcephaly and on the evolution of air transport networks, respectively.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"791\" height=\"1024\" src=\"http:\/\/www.jspsusa.org\/wp\/wp-content\/uploads\/2020\/11\/The-5th-Japan-US-Science-Forum-in-Boston-791x1024.png\" alt=\"\" class=\"wp-image-5218\" srcset=\"https:\/\/www.jspsusa.org\/wp\/wp-content\/uploads\/2020\/11\/The-5th-Japan-US-Science-Forum-in-Boston-791x1024.png 791w, https:\/\/www.jspsusa.org\/wp\/wp-content\/uploads\/2020\/11\/The-5th-Japan-US-Science-Forum-in-Boston-232x300.png 232w, https:\/\/www.jspsusa.org\/wp\/wp-content\/uploads\/2020\/11\/The-5th-Japan-US-Science-Forum-in-Boston-768x994.png 768w, https:\/\/www.jspsusa.org\/wp\/wp-content\/uploads\/2020\/11\/The-5th-Japan-US-Science-Forum-in-Boston-1187x1536.png 1187w, https:\/\/www.jspsusa.org\/wp\/wp-content\/uploads\/2020\/11\/The-5th-Japan-US-Science-Forum-in-Boston.png 1225w\" sizes=\"(max-width: 791px) 100vw, 791px\" \/><\/figure>\n","protected":false},"excerpt":{"rendered":"Virtual Special Program [Nov. 4, 2020] -DATE- Saturday, November 14, 2020, 5:00pm – 9:00pm (EST) \/ Sunday, November 15, 2020, 7:00am \u2013 11:00am (JST)-VENUE- Virtual (The link will be sent to the registrants prior to the forum.) YouTube link of event: BOSTON (Nov. 14, 2020) \u2013 Psychiatrists are faced with the difficult task of making …","protected":false},"author":1,"featured_media":5218,"parent":4889,"menu_order":11,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"_links":{"self":[{"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/pages\/5616"}],"collection":[{"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/comments?post=5616"}],"version-history":[{"count":13,"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/pages\/5616\/revisions"}],"predecessor-version":[{"id":6310,"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/pages\/5616\/revisions\/6310"}],"up":[{"embeddable":true,"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/pages\/4889"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/media\/5218"}],"wp:attachment":[{"href":"https:\/\/www.jspsusa.org\/wp\/wp-json\/wp\/v2\/media?parent=5616"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}