Carolyn's Blog


Welcome to Carolyn's blog! My name is Carolyn Michener and I'm a sophomore in the speech, language, and hearing sciences program at Boston University. This is my second year working at the STEPP Lab. Since I'm studying speech science and not engineering, I often have to do personal research to understand the more technical aspects of other lab members' projects. I'm writing this blog to help explain these projects to other individuals with little technical experience, or to those who are interested in learning more about the STEPP Lab.
Defne Abur: Hearing Researcher, Ukulelist, and 'Amazing Race' Enthusiast
June 12, 2014
Defne Abur recently graduated from Smith College and is spending the summer at the STEPP Lab to learn more about speech science before eventually pursuing a Ph.D. in Speech and Hearing Sciences. Her work in the STEPP Lab involves working with clients who have vocal hyperfunction. Vocal hyperfunction is an overarching diagnosis for patients who misuse their vocal folds. Often, the vocal folds do not completely close (or adduct) during speech production so the voice sounds breathy due to the escaping air. In particular, Defne is creating a Just Noticeable Difference (JND) program that can be used with certain characteristics of vocal hyperfunction. In a JND experiment, the user compares two stimuli and determines if the stimuli are different or the same. The experiment determines when the user can no longer perceive a difference. Defne has been interested in Speech and Hearing Sciences since working on a project studying measures of power reflectance at Smith College. Ear canal based measures of power reflectance can be a noninvasive diagnostic for causes of conductive hearing loss. Defne was also involved in hearing research last summer when studying sound stimulation in cadaver ears. Her work involved looking at the sound produced by mechanics in a cadaver ear and simulating ear disorders to see how the sound would travel differently. In her spare time, Defne, whose spirit animal is an elephant, enjoys playing the ukulele, singing, and watching 'The Amazing Race'.
Supraja Anand: Speech-language Pathologist, Parkinson's Specialist, and 'Big Bang Theory' Fan
February 12, 2014
Supraja Anand is a Post-Doctoral Associate in the Speech, Language, and Hearing Sciences program at BU. Her research focuses on investigating speech processes in elderly clients and individuals with Parkinson's disease. Her current project specifically involves hypophonia. Hypophonia is better known as the very soft speech associated with Parkinson's disease, caused by reduced coordination of the vocal folds. When talking out loud to yourself at a normal volume, your vocal folds use coordinated movements to build and release the subglottal pressure that allows for speech production. If you want to shout to your friend across the room, your vocal folds need to have more precise control because they need to build even more subglottal pressure to permit the increase in volume. Individuals with Parkinson's disease may often have problems controlling their vocal folds, and therefore can have difficulty building the subglottal pressure needed to speak at normal or loud volumes. Their vocal folds may also not completely close, which allows air to escape, creating a breathy or raspy quality to the voice. Supraja chose to pursue a career in speech pathology because she was not interested in the traditional medical or engineering routes, and found that speech therapy was a unique opportunity to help people. She particularly found her experiences with Parkinson's research at the University of Florida to be very rewarding. Supraja hopes to one day become a university professor, and conduct research while training future speech pathologists. In her spare time, Supraja, whose spirit animal is a tiger, enjoys watching 'Friends' and 'The Big Bang Theory' and hopes to one day take a tour of Europe.
Stephanie Lien: Biomedical Engineer, Signal Processing Expert, and Baker of Fine Cakes
December 10, 2013
Stephanie Lien is a third year Ph.D. student in the Boston University biomedical engineering program. Her work primarily focuses on developing an RFF acoustic measurement (see post from October 1, 2013) for vocal hyperfunction. Stephanie chose this project because she thinks it's interesting how RFF changes before and after voice therapy. RFF is known to be lower in individuals with vocal hyperfunction, and return to normal after successful voice therapy. However its values do not change after surgery. Muscle tension dysphonia (see post from November 19, 2013) is a type of vocal hyperfunction, except whereas there is no apparent organic cause of muscle tension dysphonia, vocal hyperfunction can be associated with nodules and polyps, which are essentially calluses and blisters on the vocal folds respectively. Check out a video of vocal folds with polyps here. If the nodules and polyps are too massive to be cured by voice therapy alone, the individual may undergo surgery to remove the structure in addition to participating in voice therapy to ensure the nodules and polyps don't return. Stephanie plans to optimize the RFF recording protocol for individuals with vocal hyperfunction, then validate and automate the measurements. Stephanie chose to pursue a degree in biomedical engineering because she enjoys the combination of engineering and medicine to solve health problems. Her favorite class at John Hopkins University, where she earned her BS in biomedical engineering, was an instrumentation class where projects included designing and building devices for individuals with various disabilities. She hopes to use her expertise and passion for engineering to become a research engineer after graduation. Stephanie, whose Patronus is a puffin, enjoys watching The Big Bang Theory and wants to travel to Jeju, South Korea.
Liz Heller Murray: Speech-language Pathologist, Voice Enthusiast, and Food Truck Connoisseur
November 19, 2013
Liz Heller Murray is a Ph.D. student in the Boston University Speech, Language, and Hearing Sciences program. Her current research includes examining why some patients with normal looking vocal folds develop excess muscle tension in their larynx, a condition called muscle tension dysphonia. People with muscle tension dysphonia often sound strained or hoarse and raspy, as they are inappropriately using their muscles causing an altered quality to the voice. To determine how muscle tension dysphonia affects the vocal folds, it is helpful to talk about how normal vocal folds function. The vocal folds are two small tissues that abduct (move apart) to permit breathing and adduct (come together) to produce voicing. The muscles and cartilages of the larynx, or voice box, cause the vocal folds to move in this fashion. Check out a video of video stroboscopy of healthy vocal folds here. In muscle tension dysphonia, there is excess tension of the laryngeal muscles that causes the voice to sound so strained. You can see the extreme tension in this strobscopy. In the video, the vocal folds are partially obscured from view by the over-tensing muscles. However, when the individual breathes in and the vocal folds are abducted, we can clearly see that the vocal folds look normal, meaning that the physical, organic characteristics of the vocal folds have not changed. Research in muscle tension dysphonia is therefore very important because little is known about what causes the laryngeal muscles to become so tense without an organic cause. Liz chose to study speech-pathology because she believes that the voice is a very important component in our lives, and finds that helping clients properly use their voice is extremely rewarding. After completing her Ph.D., Liz hopes to find a job as a speech-language pathologist and researcher in a voice center. In her spare time, Liz, whose Patronus is a cat, enjoys watching How I Met Your Mother and hopes to one day travel to Bora Bora.
Velopharyngeal Dysfunction: When Nasality of Speech Affects Intelligibility 
October 29, 2013
The velopharyngeal dysfunction (VPD) project is another one of my favorites because, like the RFF of Vocal Effort project (see post from October 1, 2013), it's a combination of the lab's engineering expertise and my interest in speech disorders. VPD is essentially nasalization of speech, particularly in children, and can make speech difficult to understand. The phonemes /m/, /n/, and /ng/ (as in sing) are considered to be nasal because air escapes through the nose, not through the mouth, when you say them. To feel this yourself, place your finger between your nose and upper lip and say "mom" and "dad". When you say "mom" you should be able to feel air leaving through your nose, whereas when you say "dad" air escapes through the mouth. During production of nasal sounds, the velum (aka the soft palate) opens to allow air and sound into the nasal passageway. In individuals with VPD, the velar port remains open even during the production of non-nasal sounds, causing hypernasality of speech. Check out some speech samples of hypernasal speech here. The VPD project includes asking subjects to wear a headset designed by lab members that has a microphone to pick up acoustic signals and a small accelerometer placed on the nose to read nasal vibrations. We collected data from healthy and disordered children using this headset to compare nasal patterns between the two populations. In the future, we hope to expand this project to help kids with VPD to learn to produce more intelligible speech.
Meet Meredith Cler: Student Researcher, BCI Extraordinaire, and "Doctor Who" Fan 
October 16, 2013
Meredith Cler is a Ph.D. student in the Boston University Computational Science program working on a rotation project in the STEPP Lab. Meredith's current project involves using facial surface electromyography (sEMG) to enable individuals with high spinal cord injuries to communicate via a computer based keyboard. sEMG involves the use of electrodes placed on the skin that read the electrical conduction associated with muscle contractions. EMG is very similar to the more commonly known electroencephalography (EEG), except whereas EEG measures brain activity, EMG measures muscle activity. Facial sEMG is particularly pertinent for a population of individuals with high spinal cord injuries because their facial musculature is likely intact, and muscle activation is easily detected. Meredith chose to study computational neuroscience because she wanted to use her undergraduate computer science skills for something different. She received a graduate certificate in cognitive sciences, and found that computational neuroscience was a balance between her two passions. Meredith has also studied brain-computer interfaces (BCIs) in Germany during 2012. The study of BCIs is a facet of engineering that creates a direct pathway between the brain and a technological device, usually to increase quality of life. Her hands-on experience with patients who use BCIs affirmed her choice to study computational neuroscience. After earning her Ph.D., Meredith hopes to pursue a research career, preferably in industry. In her spare time, Meredith, who identifies as a member of the Ravenclaw House, enjoys watching Doctor Who, traveling, and playing with her rabbits, T-Rex and Admiral Grace Hopper.              
How Does Vocal Effort Affect Relative Fundamental Frequency? 
October 1, 2013
This is one of my favorite projects in the lab because it directly involves the speech science behind voice disorders and engineering. ur goal is to verify if there is a link between vocal effort and the relative fundamental frequency (RFF) of an individual's voice. RFF is loosely defined as the fundamental frequency of the cycles before and after the production of a voiceless consonant. You can read more about RFF and voice disorders here. Confused? Try placing your hand against your throat and say "aaaaffffaaaa". You should be able to feel the vibrations of your throat during the vowel production but your throat should feel comparatively at rest during the consonant production. This is because when a sound is "voiced" the vocal folds within the voice box vibrate, but when a sound is "voiceless" the vocal folds do not vibrate. Each time your al folds open and close during a voiced sound a vocal cycle is produced. By taking the frequency of the vocal cycles before and after the voiceless consonant, we can find the RFF. We asked the subjects who are participating in the study to speak at different vocal efforts so we can determine if it changes the RFF. The subjects spoke in a normal voice, an easy voice, and a strained voice. The normal voice is the same level of effort you use when talking to a friend or a colleague. An easy voice is a voice that requires less effort to produce it. If you sing along to your favorite song on the radio, you're using an easy voice. A strained voice is produced by increasing the effort it takes to produce speech. One way to speak in a strained voice is to raise your shoulders and tense the muscles in your neck while speaking. Try reading a couple sentences of this post while speaking in this fashion. If you sounded like Batman meets Gollum, you've successfully produced a strained voice! We think that determining how the RFF changes between each type of vocal effort could become an effective tool in diagnosing voice disorders.