Start date: October 2021, January 2022, April 2022, July 2022
Full-time/part-time availability: Full-time (3 years)
Primary Supervisor: Sijung Hu, Email: [Email Address Removed] - http://www.lboro.ac.uk/departments/meme/staff/sijung-hu
Secondary Supervisor: Francois Nadal, https://www.lboro.ac.uk/departments/meme/staff/francois-nadal/
Modulated breathing could be used for an augmentative and alternative communication (AAC, doi.org/10.3390/bios8020048). The proposed PhD programme involves the use of direct analogue modulation of breath to encode a message, based on a “language” generated in an interactive manner between the device/system and its user. Loughborough’s AAC breath decoding and interpretation research has led to new approaches of breath-to-speech developments, using non-intrusive technology to create a dynamic air pressure detection system (DAPDS) for non-intrusive breathing detection. Standard AAC systems tend to rely on purposeful movement of a user’s muscles, and eyes which makes them limited in scope and versatility (doi.org/10.3390/s19081911). Hence, this PhD project aims to research how to comprehensively record human breathing patterns, process the recorded signals, analyse the modulation of different breath patterns and converting such patterns into meaningful words or sentences. The work is based on consolidating advanced supervised machine learning (sML) to deliver 1) a wearable, user-oriented prototype with effective frequency bandwidth and 2) pressure range for effective communication establishment.
Excellence in Science and Technology
The Smart-DAPDS will be one of the outcomes derived from solid fundamental research of human breathing flow dynamics and ML for the decoding and interpretation of breathing patterns modulations (www.lboro.ac.uk/departments/meme/research/research-projects/augmentativeandalternativecommunication/). The Smart-DAPDS will be used for a standardised testing procedure to demonstrate its reliability through the sML architecture. The Smart-DAPDS prototype offers both clinicians, AAC specialists and individuals the ability to gain a more in-depth knowledge of breathing modulation processes, as the DAPDS design makes it easy to collect data due to its ultra-light weight, comfortable, unobtrusive, and personal nature.
Due to the infinite numbers of possible user-generated breathing modulations, it is unlikely to be possible to build up a specific processing procedure for every person. Hence, we are looking for an engineering architecture to establish a ML algorithm capable of extracting breath features, adapt its way of extracting data, and interpret and predicate then speak out the words/sentences, as the proposed AAC would be, overall, more efficient, and faster consequently.
Scope of PhD Programme
The programme will consolidate the Smart-DAPDS with a sML architecture for the design of a system prototype, with potential application in and rehabilitation. The novelties of the project will include:
1) An intelligent breath detection system together with a self-calibrated ML architecture, to be consolidated with the Smart-DAPDS system.
2) Heterogeneous design of AAC to provide enhanced functionalities of the DAPDS system with optimal performance during use.
3) A heterogeneous AAC system with embedded sML architecture will be verified through biomedical engineering realisation and validated through the selective subject test, to gain a further insight into nature of breathing patterns.
This interdisciplinary research cuts across human breathing flow dynamics, ML architecture, healthcare and AAC applications, to lead a new research paradigm for real-time augmented communication.
Applicants should have or expect to achieve a 2:1 honours degree or better, or equivalent, in Engineering (Electronic Engineering, Biomedical Engineering), physical sciences (Physics), and computing programming.
English language requirements:
Applicants must meet the minimum English language requirements. Further details are available on the International website.
How to apply:
All applications should be made online. Under programme name, select ‘Electronic and Electrical Engineering’. Please quote reference number: UF-SH-2021.