A Medical Chatbot Companion for Cancer Patients: A Novel Approach to Enhance Patient Support, Education, and Data Sharing

 

Introduction

Cancer is a complex and devastating disease that affects millions of people worldwide. Cancer patients face many physical, emotional, and informational challenges throughout their journey, such as coping with symptoms, side effects, and uncertainty, managing treatment and care, and accessing reliable and personalised information. However, existing resources and services for cancer patients are often limited, fragmented, or inaccessible, which affects cancer care negatively. Therefore, there is a need for innovative and scalable solutions that can provide comprehensive and continuous support and education to cancer patients, regardless of their location, time, or situation.

 

Objective

This project aims to develop and evaluate a medical chatbot companion for cancer patients using artificial intelligence and machine learning. A chatbot is a computer program that can interact with human users through natural language, such as text or voice. A medical chatbot companion is a chatbot that can provide health-related information, guidance, and emotional support to patients based on their needs and preferences. The aim of this chatbot is to enhance patient support and education, improve patient outcomes and satisfaction, and reduce healthcare costs and burdens. Moreover, this chatbot also aims to facilitate data collection and sharing among different stakeholders, such as patients, doctors, researchers, etc., to improve the quality and availability of data for cancer care and research. In the initial phase, this project may specifically focus on breast cancer to ensure a targeted and detailed approach. Depending on the findings and requirements, this focus could later be expanded to include other cancer types.

 

Methods

The methods for this project consist of three main phases: design, development, and evaluation, with each phase integrating the needs and inputs of specific user groups, primarily cancer patients, their family members/caregivers, and healthcare professionals (oncologists, gynaecologists, radiologists).

 

Design Phase:

  • User Needs Analysis: Conduct targeted surveys, interviews, and focus groups with both cancer patients and their families/caregivers and healthcare professionals with a breast cancer focus (oncologists, gynaecologists, radiologists). The aim is to understand their specific needs, preferences, and challenges related to cancer care and information.
  • Stakeholder Collaboration: Engage healthcare professionals in workshops/interviews to identify key data points and insights that the chatbot should capture and communicate and what barriers need to be overcome for them to be willing and able to use the data. This collaboration will ensure the utility of the chatbot in the clinical setting and its relevance to patient care.
  • User Interface Design: Based on the insights gathered, design an intuitive and accessible user interface that caters to the diverse technological comfort levels of patients and their families/caregivers and aligns with the data requirements of healthcare professionals.

Development Phase:

    • The development phase involves implementing the chatbot using various AI and ML techniques. This may include the exploration and potential application of Large Language Models (LLMs) for advanced natural language understanding and generation. The use of LLMs will be contingent upon their suitability for the specific requirements of medical dialogue management and sentiment analysis in the context of cancer care.
    • Ensure the chatbot’s capability to handle specialised queries related to breast cancer and to provide personalised responses based on user interactions.
    • Incorporate features that facilitate effective data sharing between patients and healthcare providers, respecting patient consent and data privacy.

    While implementing the chatbot using AI and ML techniques, we will rigorously address technical challenges, such as the complexity of medical language interpretation and sensitive emotional support interactions, ensuring a robust and reliable system.

    Evaluation Phase:

    • Conduct an initial pilot phase with a selected group of cancer patients, their families/caregivers, and healthcare professionals. This phase serves to test the chatbot in a controlled, real-world setting.
    • Utilise usability tests, surveys, and interviews to gather feedback. Focus on understanding user experiences and identifying any immediate areas for improvement.
    • Evaluate the chatbot’s usability and gather user perceptions regarding its potential support for patient education, engagement and assistance to healthcare professionals.
    • Use insights from pilot testing and user feedback to refine the chatbot, ensuring it evolves to effectively meet user needs before considering broader applications.

    Co-Creation Approach:

    A key feature of this project is the use of a co-creation approach, which involves the participation and collaboration of the stakeholders (patients, family, caregivers, and doctors) in the design and development of the chatbot. This approach aims to ensure that the chatbot meets the needs and preferences of the users and that it is acceptable, trustworthy, and ethical. The co-creation process will include methods such as focus groups, workshops, prototyping, feedback, etc.

    Use of Patient-Reported Outcome Measures (PROMs):

    Another key feature of this project is the use of patient-reported outcome measures (PROMs), which are standardised questionnaires that capture the patient’s perspective on their health status, symptoms, quality of life, etc. PROMs are important for measuring the quality of care and outcomes of cancer patients and for informing clinical decisions and interventions. The chatbot will collect PROMs from the patients in a natural and convenient way and share them with the relevant doctors and researchers with the patient’s consent. The chatbot will also use the PROMs to provide personalised information and guidance to the patients based on their condition and progress.

    In our co-creation approach and while utilising PROMs, we will prioritise ethical considerations, including data privacy, patient consent, and secure handling of sensitive health information, adhering to the highest standards of ethical medical research.

    Results

    The expected results of this project are:

    • A prototype of a medical chatbot companion for cancer patients that can provide information, guidance, and emotional support to patients based on their queries, context, and profile.
    • A dataset of cancer-related questions and answers collected from various sources, such as interviews, cancer forums, websites, books, etc., that can be used to train and test the chatbot.
    • A set of evaluation metrics and criteria that can be used to assess the performance and quality of the chatbot. These will cover standard measures such as precision, recall, F1-score, response time, user satisfaction, and user engagement. Additionally, we will remain open to adapting or expanding these metrics to accommodate the specific technologies employed (e.g., LLMs) and insights gained from user interactions during the development process.
    • A set of co-creation methods and outcomes that can be used to demonstrate the involvement and collaboration of the stakeholders in the design and development of the chatbot and data sharing, as well as the benefits and challenges of this approach.
    • A set of PROMs and data-sharing methods and outcomes that can be used to demonstrate the collection and sharing of data among the stakeholders and the impact and implications of this process for cancer care and research.

     

    Conclusion

    This project proposes a novel approach to enhance patient support, education, and data sharing by developing and evaluating a medical chatbot companion for breast cancer patients using artificial intelligence and machine learning. This chatbot has the potential to improve patient outcomes and satisfaction by providing comprehensive and continuous support and education to patients, regardless of their location, time, or situation. It also has the potential to reduce healthcare costs and burdens by complementing or supplementing existing resources and services for cancer patients. Moreover, this chatbot has the potential to improve the quality and availability of data for cancer care and research by facilitating data collection and sharing among the stakeholders, with the patient’s consent and privacy protection. The success of this project depends on interdisciplinary collaboration, combining expertise from technology, medicine, ethics, data privacy, and patient advocacy to develop a comprehensive solution that considers the various aspects of cancer care. This project contributes to the field of medical chatbots by addressing a significant and unmet need for cancer care and research and by using a co-creation approach that involves the stakeholders in the design and development of the chatbot.

     

    While our initial focus is on breast cancer, the long-term vision is to expand the chatbot’s capabilities to encompass other types of cancer, ensuring the scalability and adaptability of the solution. This aligns with our goal of transforming cancer care through innovative technology.

     

    Contact details

    Viktoria Prantauer, TheChiefPatientOfficer.com, viktoria@thechiefpatientofficer.com, Berlin

    Feel free to reach out! I’m always looking for like-minded people to collaborate with. 

    Message me
    References

    [1] Bourgeois, A., Horrill, T.C., Mollison, A. et al. Barriers to cancer treatment and care for people experiencing structural vulnerability: a secondary analysis of ethnographic data. Int J Equity Health 22, 58 (2023). https://doi.org/10.1186/s12939-023-01860-3 

    [2] Choi, DW., Kim, S.J., Kim, D.J. et al. Does fragmented cancer care affect survival? Analysis of gastric cancer patients using national insurance claim data. BMC Health Serv Res 22, 1566 (2022). https://doi.org/10.1186/s12913-022-08988-y 

    [3] Garg R, Karim HMR. Healthcare research data sharing and academic journal: A challenging but fruitful initiative. Indian J Anaesth. 2023;67(9):763-766. https://doi.org/10.4103/ija.ija_797_23  

    [4] de Kok, J.W.T.M., de la Hoz, M.Á.A., de Jong, Y. et al. A guide to sharing open healthcare data under the General Data Protection Regulation. Sci Data 10, 404 (2023). https://doi.org/10.1038/s41597-023-02256-2

    [5] Chen S, Kann BH, Foote MB, et al. Use of Artificial Intelligence Chatbots for Cancer Treatment Information. JAMA Oncol. 2023;9(10):1459-1462. https://doi.org/10.1001/jamaoncol.2023.2954 

    [6] Xu L, Sanders L, Li K, Chow JCL. Chatbot for Health Care and Oncology Applications Using Artificial Intelligence and Machine Learning: Systematic Review. JMIR Cancer. 2021;7(4):e27850. Published 2021 Nov 29. https://doi.org/10.2196/27850 

    [7] Silveira A, Sequeira T, Gonçalves J, Lopes Ferreira P. Patient reported outcomes in oncology: changing perspectives-a systematic review. Health Qual Life Outcomes. 2022;20(1):82. Published 2022 May 21.https://doi.org/10.1186/s12955-022-01987-x 

    [8] LeBlanc, T., Abernethy, A. Patient-reported outcomes in cancer care — hearing the patient voice at greater volume. Nat Rev Clin Oncol 14, 763–772 (2017). https://doi.org/10.1038/nrclinonc.2017.153 

    [9] Swan, E.L., Peltier, J.W. and Dahl, A.J. (2023), “Artificial intelligence in healthcare: the value co-creation process and influence of other digital health transformations”, Journal of Research in Interactive Marketing, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/JRIM-09-2022-0293

    [10] Megaro, A. (2023), “Transparency in AI Systems for Value Co-creation in Healthcare”, Visvizi, A., Troisi, O. and Grimaldi, M. (Ed.) Big Data and Decision-Making: Applications and Uses in the Public and Private Sector (Emerald Studies in Politics and Technology), Emerald Publishing Limited, Leeds, pp. 93-105. https://doi.org/10.1108/978-1-80382-551-920231007

    [11] Fusco, F., Marsilio, M. and Guglielmetti, C. (2023), “Co-creation in healthcare: framing the outcomes and their determinants”, Journal of Service Management, Vol. 34 No. 6, pp. 1-26. https://doi.org/10.1108/JOSM-06-2021-0212  

    List of Past Projects

    AskMarie

    AskMarie is based on the vision to empower people to become the CEO of their health. It helps everyone to understand the science behind their health and well-being.

    Every day, around 600 million people search on Google for their health questions. To better understand their diagnosis or treatment or prepare themselves for a doctor’s appointment. But the results are not good because the answers need to be more trustworthy, and often the information conflicts. AskMarie is an AI-driven Online Health Information Seeking Tool. The basic idea is to use Large Language Models (like GPT3 or Bloom) to deliver people a summarized and easy-to-understand answers to their questions based on the Top 3 research papers. They can compare credible scientific knowledge easily ad-free, unbiased, and immediate.

    Together with Lilla Szulovsky I worked on this idea during the autumn Antler cohort 2022.

     

    White Paper Open Gender Data

    As part of the Open Heroines community, I work on an Open Gender Data Guide for a non-expert audience.

    In the Open Gender Data Lab, the need was identified to build bridges between open data experts and organisations dealing with gender issues without data knowledge. The belief is that these organisations want to use data for their causes and goals but are often unsure how to implement gender data projects.

    The guide aims to empower activists, NGO professionals, and civil servants without technical expertise by providing them with language, tools, and means to produce, analyse and use data from a feminist intersectional perspective.

    Contributors are welcome! Please feel free to reach out at any time.

    Viktoria1.0

    Viktoria1.0 is the first patient-led campaign from the Hippo AI Foundation.
    It started as an awareness campaign, which informed affected people what is possible if we unite our data and make it accessible to train AI models. People worldwide will be able to get a life-saving and accurate diagnosis. This is only possible if we, as a global society, ensure that the generated knowledge stays open and is not getting privatized.

    The young Hippo AI Foundation  empowered me to realise my vision, to provide everyone with the same level of care, I received as a patient. I collaborated with the Non-Profit-Organisation based in Berlin to launch viktoriaonezero.org. The project aims to become the most comprehensive open data set on breast cancer pathology. Viktoria1.0 aims to give humans worldwide access to an accurate diagnosis with open AI. The intense collaboration led me to join the Hippo AI Foundation as a co-founder.

    The Hippo AI Foundation brings people and data together to establish artificial medical intelligence as digital commons. It aims to make digital knowledge and medical insights gained through Artificial Intelligence from data altruism freely available to researchers and businesses. With the open-source-based approach, more data and AI building blocks are shared as digital commons. Through the collaboration of dedicated experts, they publish it with their copyleft license, which ensures that the data is available to the global community and its derivatives, such as the AI source code. Hippo AI aims to accelerate innovation and balance global inequities in healthcare through this approach.

    Newsletter

    News

    Projects

    About me

    Press & Media

    Work With Me