iTALKbetter: an interactive tablet-based application to help aphasic patients speak again
Lead Research Organisation:
University College London
Department Name: Institute of Cognitive Neuroscience
Abstract
The aim of this PhD project is for a student to develop our UCL developed interactive, speech rehabilitation application iTALKbetter jointly with our industry partner SoftV from current proof-of-concept to a deployable treatment software App for these patients to improve their spoken language function.
PURPOSE: iTALKbetter is a speech therapy App for mobile devices that will improve aphasic patients language and communication function. A speech problem in "finding the right words" (anomia) is the hallmark of aphasia. Currently there are no digital treatments for aphasia that provide feedback on speech performance. Feedback is critical for word learning.
FEASIBILITY: This will be an excellent opportunity for the right candidate to gain valuable experience in an exciting project that intersects voice signal processing, computer science, clinical neuroscience and medical entrepreneurship. The current VR prototype software programme uses the freely available Sphinx recogniser, not for speech recognition but for speech verification. The software (a mobile App) will have three key components: diagnosis, naming therapy and outcome measurement. A key USP is the development of voice recognition (VR) software within a game format that learns from and interacts with aphasic patient's speech and supports personalisation of the therapy game according to patient's progress.
PURPOSE: iTALKbetter is a speech therapy App for mobile devices that will improve aphasic patients language and communication function. A speech problem in "finding the right words" (anomia) is the hallmark of aphasia. Currently there are no digital treatments for aphasia that provide feedback on speech performance. Feedback is critical for word learning.
FEASIBILITY: This will be an excellent opportunity for the right candidate to gain valuable experience in an exciting project that intersects voice signal processing, computer science, clinical neuroscience and medical entrepreneurship. The current VR prototype software programme uses the freely available Sphinx recogniser, not for speech recognition but for speech verification. The software (a mobile App) will have three key components: diagnosis, naming therapy and outcome measurement. A key USP is the development of voice recognition (VR) software within a game format that learns from and interacts with aphasic patient's speech and supports personalisation of the therapy game according to patient's progress.
People |
ORCID iD |
| Jennifer Crinion (Primary Supervisor) | |
| David Barbera (Student) |
| Title | Audio recordings of naming attempts by people with aphasia dataset |
| Description | This dataset consists of audio recordings of naming attempts by people with aphasia (PWA) during picture naming tasks. There will be a total of 30 of PWA that are going to be contributing to this dataset by participating in the clinical trial of iTALKbetter app which collects the audio recordings and stores them in a UCL server according to current GDPR guidelines. It is expected that at the end of the clinical trial the dataset will have more than 60000 recordings of 2000 different words. This dataset is still being collected but has already enough samples to be used in our research studies. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2019 |
| Provided To Others? | No |
| Impact | There haven't been any notable impacts yet other than it is already being used in one of our papers currently in preparation. |
| Title | Gotcha: an app to practice names of familiar faces for people with dementia |
| Description | Gotcha is an app to help dementia patients to maintain or reduce decay of their capacity to remember the names of familiar faces (family, friends or any people of interest). Development of automatic speech recognition ASR to automatically assess if the patient has named a certain face correctly has been finished and so the rest of the app developed by industrial partners SoftV, including back-end infrastructure for data collection (audio recordings and statistical data) adhering to current GDPR guidelines. The app Gotcha is currently in a clinical trial to scientifically validate its effectivity on people with dementia in maintaining or reducing any decay in naming familiar faces. This app and its clinical trial have been mainly funded by the NIHR but the development of the ASR technology used in the app has been funded by the MRC. |
| Type | Therapeutic Intervention - Medical Devices |
| Current Stage Of Development | Early clinical assessment |
| Year Development Stage Completed | 2019 |
| Development Status | Actively seeking support |
| Impact | There are no other notable impacts arising from the development of this medical product other than it has helped us to refine and improve inter-disciplinary collaboration between clinicians, industrial partners and speech experts. |
| Title | iTALKbetter: a naming therapy app for anomia treatment on chronic aphasic post-stroke survivors |
| Description | iTALKbetter is an app for anomia treatment on aphasic patients. The main objective is to provide mass practice on naming therapy tasks for people with aphasia (PWA) with word retrieval difficulties (anomia). The main feature is our in-house-made speech recognition (SR) module that delivers real-time feedback to the patients by assessing each naming attempt as 'correct' or 'incorrect' and that also is responsible in driving therapy reactively governing the administration of word items depending on the answers. Development is currently finished for both our SR module and the whole app developed in conjunction with our industrial partner SoftV including data collection (audio and statistics) in a client-server model that adheres to current GDPR guidelines. iTALKbetter is currently used on a clinical trial to assess any scientific evidence of the beneficial effect on speech recovery using this digitally-delivered mass practice neuro-intervention on PWA. The SR module development has been funded by the MRC and the development of the rest of the app and clinical trial by NIHR. |
| Type | Therapeutic Intervention - Medical Devices |
| Current Stage Of Development | Early clinical assessment |
| Year Development Stage Completed | 2019 |
| Development Status | Actively seeking support |
| Impact | There hasn't been any other notable impacts arising from the development of this medical intervention other than has helped us to refine a way of working with industrial partners, speech experts and clinicians. |