EXPERIMENTAL CHARACTERIZATION OF A SOUND LOCALIZATION SENSOR FOR EARLY DIAGNOSIS OF AUTISM
Mr Luigi Battista
Full Name:Luigi Battista
City: Potenza Country: Italy
Field of study: Engineering
Degree type : Five-year Degree
Year of graduation: 2009
Name: Università Campus Bio-Medico di Roma
Country: Italy Town: Roma
Name: Luigi Battista
Date of Birth: 10/10/1985
March 2010: Inscription to Engineer Professional Association of Province of Rome
December 2009: Engineering licence achieved at Università “Campus Bio-Medico” di Roma.
19 October 2009: Graduate Degree in Biomedical Engineering, (Final Mark 110/110 with honours), Università “Campus Bio-Medico” di Roma.
Degree thesis entitled "EXPERIMENTAL CHARACTERIZATION
OF AN ANGLE SENSOR OF SOUND PRODUCED BY A ONE FREQUENTIAL SOURCE"
24 October 2007: Bachelor’s Degree in Biomedical Engineering, (Final Mark 110/110 with honours), Università “Campus Bio-Medico” di Roma.
Degree thesis entitled "AN INNOVATIVE METHODOLOGY FOR DOSE REDUCTION INTO THE RADIOTHERAPIC TREATMENT OF SOFT TISSUE TUMORS"
July 2004: High School Diploma (Final Mark 98/100), Liceo Scientifico “Galileo Galilei” di Potenza.
26 Octorber 2009 - 31 December 2009: Instructor
support for didactic and research activities in Mechanical and Thermal Measurement at Mechanical and Industrial Engineering Department (DIMI), “Università degli Studi Roma3”
1 January 2010 - Today: Doctor of Philosophy Student in "Mechanical and Industrial Engineering", at Mechanical and Industrial Engineering Department (DIMI), “Università degli Studi Roma3”
17 May 2010: winner of "2009 Thesis Award", with thesis "EXPERIMENTAL CHARACTERIZATION OF AN ANGLE SENSOR OF SOUND PRODUCED BY A ONE FREQUENTIAL SOURCE", released by BIC Lazio (Business Innovation Centre)
Criterias to contestualize the project
: mechanical measurement, sensor project
character (other interested fields):
Environment & Territory:
Other: biomedical instrumentation
Project (block 1) - Objectives of the Project*
Autism Spectrum Disorder (ASD) is a behavioural disorder, with onset in childhood, which is characterized by deficits in three basic domains: social interaction, language and communication, and pattern of interests. There is a strong evidence that autistic children typically tend to avoid eye contact, do not pay more attention to human faces than they do to everyday objects and, often, seem not to react when called by their name (initially mistaken for deafness). Such behaviours, present already in very young children, occur in social situations and are therefore elusive and difficult to be assessed.
For this purpose, a wearable device for multimodal behavioral analysis, Audio-Visuo-Vestibular Cap (AVVC), was developed to assess how humans coordinate vision, hearing and orienting behavior during social situations in response to visual and sound stimuli. This sensorized cap is composed of three different kinds of sensors (Figure 1): i) a magneto/inertial sensor for measuring the orientation of the head (similarly to the vestibular system); ii) a mini camera, or ‘eye-cam’, for detecting the gaze direction; iii) a sound localization sensor, constituted by a pair of omni-directional microphones, to record sounds occurring around the child. This sound localization sensor provides an information about sound source direction estimating the delay between the time when sound from a single source reaches the near ear and when it reaches the far ear. This cue is technically referred to as the Interaural Time Difference (ITD).
The sensor shows an evident biomimetic approach: the auditory system of humans and of many other animals uses ITD in the localization of sounds.
In this work project, fulfilment, theoretical and experimental characterization of AVVC’s sound localization sensor has been realized.
(block 2) : Progress, Methodologies and tools used
for the Project*
I start off with theoretical study of sound localization sensor finding mathematical relationship between sound source direction and Interaural Time Difference (ITD): ITD occurs whenever the distance from the source of sound to the two ears is different, resulting in differences in the arrival times of the sound at the two ears (Figure 2). When the sound source is directly in front of the listener, there is no ITD (ITD is zero), because sound reaches two ears at the same time; when sound comes from listener’s right, sound arrives first at right ear and then at left ear with a certain delay (ITD is maximum).
Then prototype’s manufacture has been realized (Figure 3): sound signals, perceived by the two microphones, are picked up by a sound card and are sent to a PC to perform signal processing and to obtain sound source direction. Prototype’s cost is about 350 €: 70% of cost is attributable to microphones and 30% to sound card; signal processing has been achieved through an algorithm which has been developed ad hoc.
After prototype’s realization, experimental trials, with manikin and non-autistic children, were performed in order to verify the sensor’s mathematical model; moreover sensor accuracy has been evaluated, that is quantification of degree of closeness of measured sound source direction to its actual value. This analysis has carried out following guidelines suggested in the works published by ISO (International Organization for Standardization ): “Guide to the Expression of Uncertainty in Measurement” and its Supplement 1 “Evaluation of measurement data - Propagation of distributions using a Monte Carlo method”.
(block 3) - Results *
Center for Disease Control (2007) estimates that an average of 1 in 110 children in the USA have an Autism Spectrum Disorder (ASD). In Italy honourable Delfino Teresio et al have suggested a bill for autism prevention and treatment. The first 2-3 years of life before language development, represents an important temporal window for an early diagnosis of ASD, so it is important to
develop technological platforms and methods for early diagnosis of developmental disorders.
For these reasons, a sound localization sensor has been projected, manufactured and theoretically and experimentally characterized for the autism early diagnosis: this transducer, constituted by a pair of microphones placed at the opposite ends of a cap, provides an information about sound source direction measuring the interaural time difference (or ITD), that is the difference in arrival time of a sound between two microphones. Afterward a theoretical study, sensor has been manufactured and a signal processing algorithm for ITD measurement has been developed ad hoc. Sensor’s accuracy and resolution were theoretically evaluated and then were experimentally validated.
ITDs have been measured for different angles of sound: experimental data, obtained without any obstacles between two microphones, show a good fitting with the mathematical model (coefficient of determination r-square: 0.997); experimental data, obtained with a sphere between two microphones (sphere simulates child’s head), show a good fitting with the mathematical model (coefficient of determination r-square: 0.987).
Data emerging from sensor’s accuracy show how its is possible to employ this sensor, without performing substantial modification, also in other contests, such as industrial automation, video conferencing, surveillance, speech recognition and technical support for deaf people.