Entity

Time filter

Source Type

MGH
Boston, MA, United States

News Article
Site: http://news.mit.edu/topic/mitmechanical-engineering-rss.xml

More than 13 million pain-blocking epidural procedures are performed every year in the United States. Although epidurals are generally regarded as safe, there are complications in up to 10 percent of cases, in which the needles are inserted too far or placed in the wrong tissue. A team of researchers from MIT and Massachusetts General Hospital hopes to improve those numbers with a new sensor that can be embedded into an epidural needle, helping anesthesia doctors guide the needle to the correct location. Currently, anesthesiologists must guide a four- to six-inch needle through multiple layers of tissue to reach the epidural space surrounding the spinal cord. They know when the needle has reached the right spot based on how the tissue’s resistance changes. However, some patients’ tissues vary from the usual pattern, which can make it more difficult to determine whether the needle is in the right place. “The needle is placed essentially blindly,” says T. Anthony Anderson, an anesthesiologist at MGH and an assistant professor at Harvard Medical School. “The needle can go too far or into the wrong tissue, which means the patient doesn’t get the positive effect that you want or is injured.” In most cases, these complications lead to reduced effectiveness of the pain-killing drug, or an excruciating post-procedure headache. In rare cases in which the needle goes too far or into a blood vessel, a stroke or spinal cord injury can occur. To improve the accuracy of epidural needle placement, Anderson teamed up with researchers at MIT’s Laser Biomedical Research Center, headed by Peter So, a professor of mechanical engineering and biological engineering. So and MIT research scientist Jeon Woong Kang designed and tested several types of optical sensors that could be placed at the tip of an epidural needle and determined that the best is one that relies on Raman spectroscopy. This technique, which uses light to measure energy shifts in molecular vibrations, offers detailed information about the chemical composition of tissue. In this case, the researchers measured the concentrations of albumin, actin, collagen, triolein, and phosphatidylcholine to accurately identify different tissue layers. This sensor, which the researchers described in the journal Anesthesiology, provides immediate feedback telling the anesthesiologist which tissue the needle is in. As an epidural needle is inserted, it passes through five layers — skin, fat, supraspinous ligament, interspinous ligament, and ligamentum flavum — before reaching the epidural space, which is the target. Beyond that space lies the dura mater, a stiff membrane that surrounds the spinal cord and cerebrospinal fluid. “The sensor is continuously measuring Raman spectroscopy signals, which tells you the chemical composition of the tissue. From the chemical composition you can identify all tissue layers, from skin to spinal cord,” Kang says. The team found that Raman spectroscopy could distinguish each of the eight tissue layers around the epidural space with 100 percent accuracy. Two other techniques that they tested, fluorescence and reflectance spectroscopy, could distinguish some layers but not all eight. The researchers have tested the sensor in pig tissue and now plan to do further animal studies before testing it in human patients. They also plan to reduce the diameter of the sensor slightly, from 2 millimeters, which is too large to fit in the most commonly used epidural needles, to 0.5 mm. Jeanine Wiener-Kronish, chief of anesthesia and critical care at MGH, says this type of sensor could greatly improve safety for epidurals, as well as other procedures involving needles. “The era of blind procedures is one we need to move away from, because we’re very interested in improving safety and quality,” says Wiener-Kronish, who was not involved in the research. “This sensor could allow us to take a fairly blind procedure and be able to get more information about where the needle is.” The researchers have started a company, Medisight Corp., to continue developing the technology, which they believe could also be applied to medical procedures, such as cancer biopsies or injecting drugs into the joints, which can be difficult to do accurately. This commercialization effort is supported by MIT entrepreneurship programs, including the MIT Translational Fellows Program, MIT Venture Mentoring Service, and MIT Innovation Initiative. The team also received support from the National Science Foundation in the form of a Small Business Technology Transfer program grant. In addition to So, Kang, and Anderson, authors of the paper include Tatyana Gubin, an MIT undergraduate, and Ramachandra Dasari, a principal research scientist in MIT’s Department of Chemistry.


News Article | August 31, 2016
Site: http://www.scientificcomputing.com/rss-feeds/all/rss.xml/all

There’s no human instinct more basic than speech, and yet, for many people, talking can be taxing. One in 14 working-age Americans suffer from voice disorders that are often associated with abnormal vocal behaviors — some of which can cause damage to vocal cord tissue and lead to the formation of nodules or polyps that interfere with normal speech production. Unfortunately, many behaviorally-based voice disorders are not well understood. In particular, patients with muscle tension dysphonia (MTD) often experience deteriorating voice quality and vocal fatigue (“tired voice”) in the absence of any clear vocal cord damage or other medical problems, which makes the condition both hard to diagnose and hard to treat. But a team from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) and Massachusetts General Hospital (MGH) believes that better understanding of conditions like MTD is possible through machine learning. Using accelerometer data collected from a wearable device developed by researchers at the MGH Voice Center, researchers demonstrated that they can detect differences between subjects with MTD and matched controls. The same methods also showed that, after receiving voice therapy, MTD subjects exhibited behavior that was more similar to that of the controls. “We believe this approach could help detect disorders that are exacerbated by vocal misuse, and help to empirically measure the impact of voice therapy,” says MIT PhD student Marzyeh Ghassemi, who is first author on a related paper that she presented at last week’s Machine Learning in Health Care (MLHC) conference in Los Angeles. “Our long-term goal is for such a system to be used to alert patients when they are using their voices in ways that could lead to problems.” The paper’s co-authors include John Guttag, MIT professor of electrical engineering and computer science; Zeeshan Syed, CEO of the machine-learning startup Health[at]Scale; and physicians Robert Hillman, Daryush Mehta and Jarrad H. Van Stan of Massachusetts General Hospital. How it works Existing approaches to applying machine learning to physiological signals often involve supervised learning, in which researchers painstakingly label data and provide desired outputs. Besides being time-consuming, such methods currently can’t actually help classify utterances as normal or abnormal, because there is currently not a good understanding of the correlations between accelerometer data and voice misuse. Because the CSAIL team did not know when vocal misuse was occurring, they opted to use unsupervised learning, where data is unlabeled at the instance level. “People with vocal disorders aren’t always misusing their voices, and people without disorders also occasionally misuse their voices,” says Ghassemi. “The difficult task here was to build a learning algorithm that can determine what sort of vocal cord movements are prominent in subjects with a disorder.” The study was broken into two groups: patients that had been diagnosed with voice disorders, and a control group of individuals without disorders. Each group went about their daily activities while wearing accelerometers on their necks that captured the motions of their vocal folds. Researchers then looked at the two groups’ data, analyzing more than 110 million “glottal pulses” that each represent one opening and closing of the vocal folds. By comparing clusters of pulses, the team could detect significant differences between patients and controls. The team also found that after voice therapy the distribution of patients’ glottal pulses were more similar to those of the controls. According to Guttag, this is the first such study to use machine learning to provide objective evidence of the positive effects of voice therapy. “When a patient comes in for therapy, you might only be able to analyze their voice for 20 or 30 minutes to see what they’re doing incorrectly and have them practice better techniques,” says Susan Thibeault, a professor at the department of surgery at the University of Wisconsin School of Medicine and Public Health who was not involved in the research. “As soon as they leave, we don’t really know how well they’re doing, and so it’s exciting to think that we could eventually give patients wearable devices that use round-the-clock data to provide more immediate feedback.” One long-term goal of the work is to be able to use the data not just to improve the lives of those with voice disorders, but to potentially help diagnose specific disorders. The team also hopes to further explore the underlying reason why certain kinds of vocal pulses are more common in patients than in controls. “Ultimately we hope this work will lead to smartphone-based biofeedback,” says Hillman. “That sort of technology can help with the most challenging aspect of voice therapy: getting patients to actually employ the healthier vocal behaviors that they learned in therapy in their everyday lives.”


News Article
Site: http://phys.org/biology-news/

Type 3 secretion systems (T3SSs) are used by several pathogenic bacterial species to deliver effector proteins, responsible for the effects of infection, into the target host cells. An MGH research team has discovered that interaction between structural proteins within the host cell called intermediate filaments and IpaC, a bacterial protein that forms the pore through the host cell membrane, is required for T3SSs to dock onto the pores and secrete effector proteins into the host cell. Credit: Goldberg Laboratory, Division of Infectious Diseases, Massachusetts General Hospital Researchers from the Massachusetts General Hospital (MGH) Division of Infectious Diseases are investigating the mechanism by which several important pathogenic species of bacteria deliver proteins into the cells of the organisms they are infecting. In a paper receiving advance online publication in Nature Microbiology, the team describes determining a key step in how the diarrheal pathogen Shigella injects proteins into target host cells. Their findings may apply to additional bacterial species, including Salmonella and those responsible for typhoid fever, bubonic plague and many hospital-acquired pneumonias. "Many bacterial pathogens establish infection by delivering effector proteins - which are responsible for many of the effects of infection - into cells. Therefore, delivery of effector proteins is a significant mediator of the impact of infection on human tissue," says Marcia Goldberg, MD, director of Research in the MGH Division of Infectious Diseases, senior and corresponding author of the report. "The most common mechanism used by a major group of bacteria is a specialized apparatus known as a type 3 secretion system, by which pathogens sitting outside the cell first make a pore in the cell membrane and then deliver effector proteins into the cell. Our data show that interaction between one of the pore proteins and another protein within the cell is required for this process to succeed." Also called injectisomes, type 3 secretion systems (T3SSs) consist of a base within the pathogen and a needle-like structure that detects contact with the target host cell and through which bacterial proteins associated with the system are secreted. Upon contact with human cells, T3SSs secrete two proteins that make up the pore in the target cell's membrane. After the pore is formed, the T3SS docks onto the pore to deliver bacterial proteins into the cell. Exactly how the pore is formed and the process by which the T3SS docks onto the pore have not been clearly understood, and the MGH team set out to investigate the mechanism behind the docking process. In a series of cellular experiments, they first determined that Shigella infection requires the presence of cytoskeletal proteins called intermediate filaments - in this case a protein called vimentin - within host cells. They then showed that vimentin was required for efficient delivery of effector proteins via the T3SS and that vimentin directly interacts with IpaC, one of the proteins that make up the pores in the host cell membrane. While vimentin is not involved in the formation of pores, the researchers showed it is required for efficient docking of Shigella T3SSs onto the pores in the host cell membrane and that docking induces the secretion of effector proteins into the host cell. While these experiments focused on the role of vimentin, the investigators also showed that all intermediate filaments normally present in cells infected by Shigella are required for infection by the pathogen. Additional experiments with strains of Salmonella and Yersinia, the families that include the typhoid fever and plague bacteria, confirmed that intermediate filaments were also required for the T3SSs used by those pathogens, suggesting similar mechanisms may apply to all pathogens using T3SSs. "We know that type 3 secretion systems are critical for the virulence of the bacterial pathogens that use them and their inactivation renders those pathogens non-virulent," says Brian Russo, PhD, a research fellow on Goldberg's team and lead author of the report. "Fundamental understanding of how the type 3 secretion system functions could lead to the discovery of novel therapies for these very important infectious diseases." Explore further: Researchers see activity of bacterial effector protein in molecular detail More information: Brian C. Russo et al. Intermediate filaments enable pathogen docking to trigger type 3 effector translocation, Nature Microbiology (2016). DOI: 10.1038/NMICROBIOL.2016.25


News Article
Site: http://phys.org/technology-news/

Unfortunately, many behaviorally-based voice disorders are not well understood. In particular, patients with muscle tension dysphonia (MTD) often experience deteriorating voice quality and vocal fatigue ("tired voice") in the absence of any clear vocal cord damage or other medical problems, which makes the condition both hard to diagnose and hard to treat. But a team from MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL) and Massachusetts General Hospital (MGH) believes that better understanding of conditions like MTD is possible through machine learning. Using accelerometer data collected from a wearable device developed by researchers at the MGH Voice Center, researchers demonstrated that they can detect differences between subjects with MTD and matched controls. The same methods also showed that, after receiving voice therapy, MTD subjects exhibited behavior that was more similar to that of the controls. "We believe this approach could help detect disorders that are exacerbated by vocal misuse, and help to empirically measure the impact of voice therapy," says MIT PhD student Marzyeh Ghassemi, who is first author on a related paper that she presented at last week's Machine Learning in Health Care (MLHC) conference in Los Angeles. "Our long-term goal is for such a system to be used to alert patients when they are using their voices in ways that could lead to problems." The paper's co-authors include MIT professor John Guttag; Zeeshan Syed, CEO of the machine-learning start-up Health[at]Scale; and Drs. Robert Hillman, Daryush Mehta and Jarrad H. Van Stan of Massachusetts General Hospital. Existing approaches to applying machine learning to physiological signals often involve supervised learning, in which researchers painstakingly label data and provide desired outputs. Besides being time-consuming, such methods currently can't actually help classify utterances as normal or abnormal, because there is currently not a good understanding of the correlations between accelerometer data and voice misuse. Because the CSAIL team did not know when vocal misuse was occurring, they opted to use unsupervised learning, where data is unlabeled at the instance level. "People with vocal disorders aren't always misusing their voices, and people without disorders also occasionally misuse their voices," says Ghassemi. "The difficult task here was to build a learning algorithm that can determine what sort of vocal cord movements are prominent in subjects with a disorder." The study was broken into two groups: patients that had been diagnosed with voice disorders, and a control group of individuals without disorders. Each group went about their daily activities while wearing accelerometers on their necks that captured the motions of their vocal folds. Researchers then looked at the two groups' data, analyzing more than 110 million "glottal pulses" that each represent one opening and closing of the vocal folds. By comparing clusters of pulses, the team could detect significant differences between patients and controls. The team also found that after voice therapy the distribution of patients' glottal pulses were more similar to those of the controls. According to Guttag, this is the first such study to use machine learning to provide objective evidence of the positive effects of voice therapy. "When a patient comes in for therapy, you might only be able to analyze their voice for 20 or 30 minutes to see what they're doing incorrectly and have them practice better techniques," says Dr. Susan Thibeault, a professor at the department of surgery at the University of Wisconsin School of Medicine and Public Health who was not involved in the research. "As soon as they leave, we don't really know how well they're doing, and so it's exciting to think that we could eventually give patients wearable devices that use round-the-clock data to provide more immediate feedback." One long-term goal of the work is to be able to use the data not just to improve the lives of those with voice disorders, but to potentially help diagnose specific disorders. The team also hopes to further explore the underlying reason why certain kinds of vocal pulses are more common in patients than in controls. "Ultimately we hope this work will lead to smartphone-based biofeedback," says Hillman. "That sort of technology can help with the most challenging aspect of voice therapy: getting patients to actually employ the healthier vocal behaviors that they learned in therapy in their everyday lives." Explore further: Voice prostheses can help patients regain their lost voice


Carruthers M.N.,Harvard University | Khosroshahi A.,Emory University | Augustin T.,North Shore Hospital | Deshpande V.,MGH | Stone J.H.,Harvard University
Annals of the Rheumatic Diseases | Year: 2015

Objectives: We evaluated the sensitivity, specificity and positive and negative predictive values of elevated serum IgG4 concentrations for the diagnosis of IgG4-RD. Methods: Between 2001 and 2011, 190 unique patients had elevated serum IgG4 measurements. We reviewed electronic medical records to determine the indication for IgG4 measurement and underlying clinical diagnosis. Additionally, we reviewed the records of 190 other randomly selected patients from a pool of 3360 with normal results, to evaluate test characteristics of the IgG4 measurement. Results: Among 380 patients analysed, 72 had either probable or definite IgG4-RD. Sixty-five of the 72 IgG4-RD patients had elevated serum IgG4 concentrations (mean: 405 mg/dL; range 140-2000 mg/dL), for a sensitivity of 90%. Among the 308 subjects without IgG4-RD, 125 had elevated IgG4 (mean: 234 mg/dL; range 135-1180 mg/dL) and 183 had normal IgG4 concentrations, for a specificity of 60%. The negative predictive value of a serum IgG4 assay was 96%, but the positive predictive value only 34%. Analysis of the serum IgG4/total IgG ratio did not improve these test characteristics. Doubling the cutoff for IgG4 improved specificity (91%) but decreased sensitivity to 35%. Discussion: Multiple non-IgG4-RD conditions are associated with elevated serum IgG4, leading to poor specificity and low positive predictive value for this test. A substantial subset of patients with biopsy-proven IgG4-RD do not have elevated serum IgG4. Neither doubling the cutoff for serum IgG4 nor examining the serum IgG4/IgG ratio improves the overall test characteristics for the diagnosis of IgG4-RD. © 2015, BMJ Publishing Group. All rights reserved. Source

Discover hidden collaborations