BACKGROUND

HW is a 37 year old male with severe quadriplegic spastic cerebral palsy.  He has a lot of involuntary movements of his arms and legs, and cannot isolate his finger movement.  His best control has been with his head, which he has used to operate head switch arrays, single switches, and head pointers.  During his last year at his previous facility, he used several low technology communication methods.  These included a language board with more than 200 symbols, an ETran board to number encode messages, and a head pointer to operate keyboards. Shortly after coming to Woods Services in May 1985, head pointer use was discontinued due to neck strain. HW continues to use eye movements developed with the Etran board.  Communication involves his looking upper left for 0, middle left for 1 and 2, lower left for 3. 4,5, and 6 are along the bottom, and 7,8,9 are up the right side.  Digit choice is then confirmed, then HW and his communication partner repeat this process twice to build a three digit number corresponding to one of 268 utterances that he has memorized, and printed on his lap tray.  While this system has the advantage of always be available, and not needing batteries, not everyone knows how to use this communication technique. Additionally when he meets new people they must learn his system.

For this reason, several attempts have been made to move HW to an electronic communication system. Unfortunately the numbering system does not correspond to his frequency of usage or the numeric codes used on the Phonic Ear HC120 speech output communicator.  In many ways, the path HW took trialing electronic communication devices illustrates how much has changed in the field and how much has stayed the same.  While it is nice to have a better sounding voice in a smaller, lighter package with more battery life as AAC devices improve, technology will always have its quirks, users should be accepting of its limitations and trained to work around problems as they arise. This latest effort at using a Speech Generating Device was prompted by his psychologist, who does not know HW”s language system.  Indeed it would be very beneficial for a client to be able to talk to his psychologist without needing an interpreter.

Past communication efforts at Woods Services have included programming a Freestyle speech output communicator with his vocabulary.  To use this system he would single switch scan through five pages of vocabulary, then through blocks of 6 to 12 utterances, and then single square scanning to make the final selection. Page backgrounds had different colors, and because of the group scanning process, numbers were not scanned sequentially.

A trial communication system by the Occupational Therapy department used the AutoCorrect feature found in earlier versions of Microsoft Word.  Using Clicker 4 as an onscreen keyboard, three digit numbers would be entered into a Word document and then AutoCorrected into his utterance.  Clicker 4’s text-to-speech feature was then used to pronounce the text. Staff was unable to locate the AutoCorrect file to replicate the setup, and the original file was lost when computers were exchanged.  Macros to expand his digits into utterances were also tried but were hindered by antivirus protection.

From reviewing these attempts to incorporate electronic communication devices into HW’s daily life, a few points became apparent:

1. HW is very comfortable with his current eye gaze system, and a substitute or adjunct system has a greater chance of being used if it incorporates elements from his current method. 
   
2. When using a low tech eye gaze communication system, HW has the undivided attention of the communication partner, with a speech generating device, the communication partner can be doing other tasks as HW composes a message. 
   
3. Like many people, HW has certain people with which he likes to interact, a specialized communication technique may be part of enforcing his preferences.

Design goals for HW’s communication system included:

1. Single switch scanning -- while he operates power mobility by directional head switches, he did not want to directionally control a cursor in a two-dimensional matrix.
   
2. Use of a QWERTY keyboard layout, while far from optimal for single switch scanning.  From using On-Screen Keyboard and Cross Scanner, he was very familiar with this keyboard.
   
3. Keep as much of the three digit and vocabulary as possible

METHODOLOGY

Realizing that HW’s numbering system was a form of Minspeak, Speech Generating Devices able to use this technique were researched.  The Alphatalker and Deltatalker communicators by  Prentke Romich were considered and available.  The Deltalker was selected because of its Liquid Crystal Display, giving feedback as to which digits have been selected.  Initially the Deltalker was programmed as a 32 location device on a 8 x 4 grid. 

This required some folding of the digit sequence and a 3 x 4 number subgrid was designed.  While this could be accessed faster than the linear array of the standard QWERTY keyboard, in order to minimize change, the Deltalker was also programmed as a 128 location device on a 16 x 8 grid. While certainly not optimum for a single switch scanner, HW has years of practice with QWERTY keyboards.

To setup the Delta talker for his vocabulary, digits 0 –9 were made into icons, then each utterance was made into a three icon sequence.  A cell was programmed to switch the Delta talker from Minspeak to spell mode.  This allows HW to enter words not in his vocabulary on a letter by letter basis.  The icon prediction feature of the Delta talker was turned on, in Minspeak mode this limits scanning to two rows, including numbers and functions such as Speak Sentence and Delete Last Icon.  When spell mode is selected, the entire QWERTY keyboard is scanned.

The Speech Generating Device described here is part of a power mobility, communication, computer access, television remote control and lap tray system further described in a paper submitted for RESNA 2009 (1)

RESULTS

The communication system has been trialed with HW. An analysis of the video recording taken during two trial sessions shows overlap in the time it takes HW to convey a phrase to his communication partner between using his eye gaze technique and using the Speech Generating Device programmed with his vocabulary.  Therefore even with minimal practice using the device, HW can equal the communication rate of his eye gaze system.  It is expected that his communication will improve with more familiarity with the device.

DISCUSSION

The communication system being implemented is certainly not optimum for an individual with HW’s cognitive and physical abilities, however he has very set preferences, hopefully use of this system will encourage him to be transitioned to another communication system incorporating his number system, spelling, and other selection techniques which will allow him to communicate faster and easier.