Oral Presentation Clinical Oncology Society of Australia Annual Scientific Meeting 2022

Emotional transformations during Victoria’s Cancer Council 131120 information and support service calls detected using Artificial Intelligence (#17)

Daswin De Silva 1 , Sajani Ranasinghe 2 , Gihan Gamage 2 , Harsha Moraliyage 2 , Nishan Mills 2 , Jessica Bucholc 3 4 , Katherine Lane 3 , Angela Cahill 3 , Victoria White 5 , Nikki McCaffrey 3 4
  1. Centre for Data Analytics and Cognition, La Trobe University, Melbourne, Victoria, Australia
  2. Centre for Data Analytics & Cognition, La Trobe University, Melbourne, Victoria, Australia
  3. Cancer Council Victoria, Melbourne, Victoria, Australia
  4. Deakin Health Economics, Deakin University, Burwood, Victoria, Australia
  5. School of Psychology, Deakin University, Burwood, Victoria, Australia

Aims Charities such as Cancer Council Australia provide telephone information and support services to address unmet information and psychological support needs. Customised, novel Artificial Intelligence (AI) algorithms and Natural Language Processing/Understanding (NLP/U) techniques can analyse large volumes of call recordings to provide insights into the multifactorial context and rationale for contacting the service, interactions and outcomes. This analysis aimed to explore callers’ emotions during interactions with Victoria’s Cancer Council 131120 service.  

Methods A custom AI framework composed of five modules was formulated to analyse the call recordings: speech-to-text pipeline for call recording transcription; acoustics pipeline to detect variations in tonality and pitch; unified data model to integrate structured and unstructured data; AI-generated emotion profile and transition extraction; and NLP/U extraction for lexicon and learning-based insights. The call transcriptions were split into five equal segments and the emotion profile extracted. The transition across each emotion was computed to determine the change in emotion as the call progressed. The framework was applied to a sample of 18,336 call recordings (Jan18-Dec21).

Results Most calls were from people living with or surviving cancer (42%), carers (30%), and the general public (27%). Overall, 44% calls lasted more than 15mins (54% ≤15mins, 2% <1min). Trust and joy had the largest increase during calls (84% and 73% respectively) while the four negative emotions, sadness, fear, anger, disgust, all decreased during the calls. Sadness expressed by the nurse increased just as the intensity of trust expressed by the caller increased.

Conclusions The findings suggest contacting Victoria’s Cancer Council 131120 service reduces callers’ feelings of sadness, fear, anger and disgust and promotes positive emotions such as trust and joy. This study illustrates the technical capability and practical value of AI and NLP/U techniques to provide insights into the psychological and information support needs of people living with cancer.