Well Said: Toronto Speech Therapy | Providing Speech Therapy for Adults since 2012

View Original

The Role of Artificial Intelligence in Speech Therapy

Written by: Kendra Wormald / Treatment / December 20, 2023 / 10 minutes read

See this content in the original post
  1. What Is Artificial Intelligence

  2. Incorporating Artificial Intelligence into Speech Therapy

  3. Limitations of AI in Speech Therapy

As technology continues to evolve, the use of artificial intelligence (AI) in a professional capacity within the healthcare system is on the rise. As such, it is warranted to consider how artificial intelligence could be integrated into a speech therapist's practice.

See this content in the original post

AI has historically been around since 1950, accompanying it are various definitions of what it actually is. According to McKinsey and Company (2023), "Artificial intelligence (AI) is a machine’s ability to perform the cognitive functions we associate with human minds, such as perceiving, reasoning, learning, interacting with an environment, problem solving, and even exercising creativity". Alternatively defined with a healthcare perspective; AI can be defined as "... a statistical approach that harnesses the power of data to create tools that can enhance our clinical practice and promote the health and success of our clients "(Liss and Berisha, 2020). 


You may interact with AI more frequently than you think. Using Siri, Google Maps, Alexa and ChatGPT for example, use machine learning algorithms to answer questions and provide information in a moment's notice… or within one or two repetitions.

See this content in the original post

AI has been present in the speech and language field in some capacity for many years. For example, automatic speech recognition (ASR) is an AI tool that allows us to dictate session notes and book sessions into a calendar. Reducing administrative task time leads to more time available for direct client care.


Intelligent assistive technologies and augmentative and alternative communication (AAC) are algorithms and devices that compensate for communicative and cognitive impairments. Automatic Speech Recognition (ASR) systems can be trained to recognize a specific set of words that may be challenging for the average person to decipher. The system can then display what has been said visually (text) or verbally (synthesized speech). Most basic AI systems require the user to know how to spell or know the phrase they want to say. Advancements are focused on providing a bank of common phrases and words to provide a better ‘starting point’.

Voice banking is another way AI can support communication. This technology is often used with populations who may eventually lose their ability to verbally communicate, such as with some stages of ALS or Parkinson's Disease. The person will periodically record their speech using a web-based platform. These voice samples will be used to generate synthesized speech which more closely resembles the person's voice, compared to a regular text-to-speech software. Our vocal qualities are part of our identity, thus this system allows the person to maintain congruence between themselves and how they are perceived by others. 


Better Speech has developed a generative AI-powered Speech Therapist ("Jessica") that has continuously advancing, "...speech-recognition and language- processing abilities that can assess speech patterns, identify areas for improvement, and deliver targeted interventions," (Ortiz, 2023). The company also notes, "our software is designed to augment our service and serve as a practicing tool, not to replace our Speech Therapists". 

See this content in the original post

The use of AI may support more efficient, accurate and reliable assessment administration or interpretation and be able to suggest personalized treatment plans (Liss and Berisha, 2023). This support may lead to less human error and more accurate analysis of communication abilities and progress, which reduces the current problem of test reliability. AI can also detect changes in speech and communication and can provide real-time feedback within your home, which will support generalization of skills and continuity of care (Jaime Van Echo, Matt Zbrog, 2023).

Other AI systems such as Cognixion One, can detect signals from users’ brain waves associated with language and turn these thoughts into speech via a wearable device (Miller, 2023).


The AI Institute for Exceptional Education recently received a 5-year grant to generate an AI Screener that will, "... listen to and observe children in the classroom, collecting samples of children’s speech, facial expressions, gestures and other data. It will create weekly summaries of these interactions that catalog each child’s vocabulary, pronunciation, video snippets and more" (Penn State, 2023). Generated summaries may help teachers monitor speech and language abilities or if warranted, suggest a formal assessment be conducted by a registered Speech-Language Pathologist. A significant challenge currently facing is that there is more need than can be supported, leading to lengthy waitlists. Early intervention is a key principle in promoting faster progress and more dramatic results (Kay-Raining Bird, Cleave, Trudeau, Thordardottir, Sutton, and Thorpe 2012).



To develop your overall Communication Wellness skills, see our Masterclass. 


At WELL SAID, we are in a unique position to optimize the use of AI to support client care facilitated through the online video platform, Zoom. Zoom has recently increased the variety of AI features it offers such as intelligently processing recordings, summarizing meetings, providing feedback to an inquiry during the meetings, and providing feedback on your communication skills such as rate, filler words used and even 'patience' level.

See this content in the original post
  • It is crucial that personal health information is stored in a manner that coincides with regulatory standards. 

  • As great as AI is, it's not 100% accurate. For example, summaries it provides may not accurately reflect what had been discussed in a session 

  • Feedback provided by AI may not be considerate of diverse dialects, accents and speech styles

  • AI can only be as smart or effective as the quality of data you provide it

  • AI has a limited ability to express and feel empathy or compassion, two necessary qualities that support a holistic and person-centered approach 


AI systems often improve the more we use them, in both accuracy and ease of use. As noted by Liss and Berisha (2020), "SLPs will need to understand AI clinical tools, know their limitations, and employ them judiciously as they become available." The speech therapist must always inform the client that AI may be used and allow them to agree or disagree to the use of AI at any point in time. Although the future of AI is exhilarating and appealing, it is important to remember that AI tools are meant to complement human therapists rather than replace them.   

To speak with a psychotherapist or one of the speech-language pathologists at Well Said: Toronto Speech Therapy, schedule an initial consultation by clicking the link below or calling (647) 795-5277.

See this gallery in the original post