10.1.1. Background

Healthcare systems worldwide are under mounting pressure due to aging populations, an increasing prevalence of chronic diseases and rising demands for personalized care. Specialties such as oncology and physiotherapy are particularly affected, as they require ongoing monitoring, multidisciplinary coordination and long-term patient engagement (Raj et al. 2020). Traditional healthcare delivery models often struggle to maintain consistent communication and support for patients, especially between clinical visits. In this context, artificial intelligence (AI)-enabled chatbots and virtual assistants have emerged as promising tools to extend healthcare services beyond conventional settings. These technologies can enhance patient engagement, streamline workflows and provide scalable support throughout the care continuum.

10.1.2. Definitions and scope

To frame the discussion, it is important to define the key technologies explored in this chapter:

• Chatbots: AI-powered software applications designed to simulate human conversation via text or voice (Karn et al. 2024). They typically handle predefined queries and offer information, reminders or basic interactions in a user-friendly manner.
  
• Virtual assistants: more advanced systems that use natural language processing (NLP), machine learning (ML) and contextual awareness to perform complex tasks, interact intelligently with users and provide personalized guidance. These assistants can analyze user input, adapt to individual needs and integrate with other digital health platforms (Jimenez et al. 2023).

These tools can be deployed across various interfaces – including smartphones, web-based platforms, wearable devices and electronic health records (EHRs) – to support functions such as symptom tracking, patient education, rehabilitation coaching, appointment scheduling and mental health support.

10.1.3. Objectives

This chapter aims to:

• analyze the applications of AI-enabled chatbots and virtual assistants in oncology and physiotherapy;
  
• evaluate their effectiveness in improving both clinical outcomes and patient experience;
  
• discuss current limitations, ethical and privacy concerns, and future prospects for the integration of these tools into standard care practices.

10.2.1. Natural language processing (NLP)

Natural language processing (NLP) forms the backbone of conversational AI systems, enabling chatbots and virtual assistants to comprehend, process and generate human language (Babu and Akshara 2024). In healthcare applications, NLP allows these tools to interact meaningfully with patients and clinicians. Key advancements include:

• Named entity recognition (NER): facilitates the identification of specific medical terms such as symptoms, medications, diagnoses and anatomical references from patient input. This is critical for symptom triage and clinical decision support.
  
• Sentiment analysis: detects emotional cues within a patient’s responses, enabling the chatbot to adjust its tone, offer reassurance or escalate issues to human professionals when necessary – particularly important in oncology and mental health support (Lal et al. 2025).
  
• Context-aware dialogue management: maintains coherent, multi-turn conversations by recognizing context, tracking user intent and handling ambiguity. This feature allows for dynamic and patient-specific interaction, improving user engagement and the quality of care provided.

These NLP capabilities collectively enable chatbots to act not merely as scripted responders, but as adaptive conversational agents capable of delivering nuanced healthcare communication.

10.2.2. Machine learning and knowledge bases

Machine learning (ML) enhances the functionality and adaptability of AI-driven chatbots and virtual assistants (Inavolu 2024). Through supervised and reinforcement learning, these systems can refine their responses, personalize interactions and predict patient needs over time. Key elements include:

• Predictive analytics: ML models can assess risk factors, forecast treatment adherence or anticipate complications based on user behavior and input patterns.
  
• Knowledge bases and medical ontologies: integration with standardized resources such as SNOMED CT (Systematized Nomenclature of Medicine Clinical Terms), ICD-10 (International Classification of Diseases) and clinical guidelines ensures that chatbot responses are medically valid and aligned with evidence-based practice (Vuokko et al. 2023).
  
• Continuous learning: these systems can improve performance and expand functionality by analyzing data from user interactions, clinical feedback and real-world outcomes, making them increasingly reliable and relevant.

10.2.3. Integration with EHRs and mHealth devices

A critical enabler of personalized care is the integration of chatbots with electronic health records (EHRs) and mobile health (mHealth) devices. Such integration allows for:

• Personalized responses: access to patient history, comorbidities, medications and recent clinical data enables chatbots to tailor advice, reminders and interventions more precisely (Hindelang et al. 2024).
  
• Real-time monitoring: when linked with wearable devices or mobile sensors, chatbots can monitor vital signs, mobility metrics and rehabilitation progress. This data can inform feedback on exercise routines, detect anomalies or alert clinicians when thresholds are exceeded.
  
• Closed-loop systems: integration with EHRs facilitates bidirectional communication, allowing clinicians to input care plans and receive updates on patient progress from chatbot interactions, enhancing continuity and coordination of care (Chen et al. 2024).

10.3.1. Patient education and decision support

Oncology patients frequently face complex and emotionally charged information regarding their diagnosis, treatment options and potential outcomes. AI-enabled chatbots can play a critical role in improving health literacy and empowering patients to participate actively in their care (Anisha et al. 2024). These tools provide 24/7 access to understandable, consistent and personalized medical information:

• Cancer Chatbot offers structured, evidence-based responses to frequently asked questions on chemotherapy, radiation therapy, immunotherapy and prognostic expectations. Its interactive format allows patients to receive tailored information based on cancer type and stage.
  
• Virtual assistants such as IBM Watson for Oncology assist healthcare professionals by synthesizing massive datasets, including medical literature, clinical trial data and treatment guidelines, to provide evidence-based treatment recommendations. This aids oncologists in clinical decision-making while ensuring transparency and explainability in care delivery (Adeniran et al. 2024).

By improving both patient comprehension and clinician decision-making, these tools support shared decision-making and enhance the quality of oncology care.

10.3.2. Symptom monitoring and reporting

Managing symptoms during cancer treatment is crucial for patient safety and quality of life.

Conversational AI can facilitate real-time symptom tracking, prompt interventions and improve communication between patients and providers (Li et al. 2023):

• Vivibot: developed by Hopelab, this targets young adult cancer survivors and assesses emotional wellbeing using chatbot interactions. It also delivers coping strategies and resilience-building exercises.
  
• Belong.life: a comprehensive cancer navigation platform, this incorporates AI-driven chat features that allow users to report symptoms such as fatigue, nausea and pain. It enables timely alerts to healthcare teams and connects patients with community-based peer support.

Such systems reduce delays in addressing adverse effects and improve adherence to treatment regimens, contributing to better clinical outcomes and reduced hospital readmissions.

10.3.3. Mental health and psychosocial support

Cancer diagnosis and treatment are often accompanied by significant psychological distress, including anxiety, depression, isolation and fear of recurrence (Simonelli et al. 2017).

AI-based conversational agents are increasingly used to address these psychosocial needs:

• Chatbots can deliver cognitive–behavioral therapy (CBT) modules, guiding users through structured exercises to manage negative thoughts and emotions.
  
• They can also provide motivational dialogues, daily affirmations.