A Narrative Review of the Role of Artificial Intelligence in Supporting the Mental Health of Patients with Breast Cancer

Introduction

Breast cancer is the most commonly diagnosed cancer among women worldwide, accounting for approximately 2.3 million new cases annually. ¹ While advances in screening, diagnosis, and treatment have significantly improved survival rates, breast cancer remains a major source of psychological distress for patients. Many individuals experience depression, anxiety, and post-traumatic stress disorder (PTSD) as they navigate the challenges of diagnosis, treatment, and survivorship. ² The emotional burden is further compounded by concerns about body image, treatment side effects, and uncertainty about the future, making mental health support a critical component of comprehensive breast cancer care. Despite the high prevalence of psychological distress, access to timely and personalized mental health interventions remains limited due to barriers such as stigma, provider shortages, and logistical constraints. ³

Artificial intelligence (AI) has the potential to transform oncology by enhancing early detection, treatment planning, and patient monitoring through advanced data analytics and machine learning models. ⁴ More recently, AI has shown promise in mental health care, offering innovative solutions for early distress detection, personalized therapeutic interventions, and augmented psychological support. AI-driven technologies, including natural language processing (NLP), chatbots, and predictive modeling, have the potential to bridge critical gaps in psycho-oncology by identifying at-risk patients, tailoring interventions, and expanding access to care. ⁵ As the field continues to evolve, understanding how AI can be ethically and effectively integrated into mental health support for patients with breast cancer is essential.

The objective of this article is to explore the potential of AI in enhancing mental health care for patients with breast cancer specifically, and patients with cancer more generally, examining current applications, benefits, limitations, and future directions. By analyzing existing research and emerging AI-based interventions, this article aims to provide insights into how AI can be leveraged to improve psychological well-being, increase access to mental health services, and support long-term survivorship in breast cancer care.

Psychological Impact of Breast Cancer

A breast cancer diagnosis is a life-altering event that brings not only physical challenges but also profound psychological distress. Studies indicate that up to 50% of patients with breast cancer experience clinically significant anxiety and depression during or after treatment. ² The uncertainty surrounding the disease, the side effects of treatment, and concerns about recurrence contribute to heightened emotional distress. While some patients experience adjustment-related stress that diminishes over time, others develop persistent psychological difficulties that can significantly impair quality of life. Even after achieving remission, many individuals continue to struggle with long-term anxiety, depression, and post-traumatic stress symptoms, highlighting the enduring psychological burden of breast cancer survivorship. ³

Despite the high prevalence of mental health concerns, access to appropriate psychological care remains limited for many patients with breast cancer. ⁶ Psycho-oncology services, though beneficial, are not always readily available, particularly for individuals in rural or underserved areas. Social stigma surrounding mental health issues may also prevent some patients from seeking support, as they may feel pressure to remain strong or prioritize their physical recovery over their emotional well-being. Additionally, the structure of traditional oncology care often leaves little room for dedicated mental health interventions, as oncologists and health care teams are primarily focused on medical treatment. Time constraints, competing priorities, and a lack of integrated mental health services contribute to unmet psychological needs, leaving many patients to manage their distress alone. ⁶

AI-driven tools have the potential to bridge critical gaps in accessibility, personalization, and real-time psychological monitoring for this patient population. Examples are provided in Table 1 and include how machine learning models may be able to predict which patients are at higher risk for psychological distress, allowing for earlier intervention. AI-powered chatbots and digital platforms of the future may provide immediate, stigma-free emotional support, while wearable devices and biometric monitoring could offer continuous, real-time assessments of mood and stress levels for patients with breast cancer of the future. By integrating AI-driven mental health solutions into oncology care, it may be possible to expand access, enhance personalization, and ultimately improve psychological outcomes for patients with breast cancer.

AI Applications in the Mental Health of Patients with Breast Cancer

As seen in Table 1, the potential for AI to offer new tools for early detection, personalized interventions, and long-term psychological monitoring in patients with breast cancer is growing. Given the high prevalence of depression, anxiety, and emotional distress among individuals facing a breast cancer diagnosis, AI-driven solutions have the potential to enhance access, efficiency, and effectiveness of psycho-oncology services. AI applications in mental health support can be categorized into early detection and screening, personalized interventions, and long-term monitoring, each contributing to a more comprehensive and proactive approach to psychological care.

AI for Early Detection and Screening of Psychological Distress

NLP is an advanced AI technique that enables computers to analyze, interpret, and derive meaning from human language. In the context of mental health care, AI platforms utilize NLP to assess patient-reported narratives, detect emotional distress, and identify signs of depression or anxiety. ⁷ These tools can be integrated into electronic health records, online support communities, or mobile applications, where they analyze textual data to flag at-risk patients and recommend appropriate interventions used to help inform clinical assessment and intervention by trained clinicians.

For instance, as reported in Table 1, machine learning models trained on large datasets of medical notes and patient communications have been shown to effectively predict psychological distress. ¹⁶ AI can also analyze social media posts, chat interactions, or therapy transcripts to detect linguistic patterns indicative of emotional distress, allowing for early psychological intervention before symptoms worsen. ¹⁷ Such automated assessments reduce reliance on self-report surveys, which can be burdensome for patients and subject to response bias. Beyond text analysis, AI can also be leveraged for predictive modeling of psychological distress. Machine learning algorithms can analyze patient demographics, tumor characteristics, treatment regimens, genetic predispositions, and prior mental health history to forecast which patients are at highest risk for anxiety, depression, or PTSD. ¹⁸ For example, as seen in Table 1, AI models trained on oncology databases can recognize patterns in treatment trajectories that correlate with heightened distress levels, allowing clinicians to offer targeted early interventions. ⁸ Additionally, predictive AI tools can assist oncologists and mental health providers in determining which patients may require more intensive psychological support throughout treatment and survivorship. ⁸

AI-Driven Personalized Interventions

AI-powered mental health chatbots have demonstrated effectiveness in delivering cognitive behavioral therapy (CBT)-based interventions to users experiencing anxiety and depression. ⁹ These conversational agents engage in empathetic dialogue, helping patients reframe negative thoughts, practice mindfulness, and develop coping strategies. They differ from similar platforms such as digital or virtual assistants in that they provide interventions and not merely aid in tasks. For patients with breast cancer, AI-driven chatbots could provide on-demand emotional support, reducing the burden on human therapists and ensuring that patients have access to psychological care outside of clinical settings. Some oncology-specific chatbots are being developed to offer tailored mental health resources, including education on cancer-related distress, coping mechanisms, and relaxation techniques. ¹⁰ These chatbots may also provide self-guided interventions, ensuring continuous psychological support even when human therapists are unavailable.

Telepsychiatry has become an essential component of modern mental health care, particularly for cancer patients who face logistical challenges in attending in-person therapy sessions. AI-driven enhancements, such as sentiment analysis and voice recognition, can analyze speech patterns, tone, and word choice during virtual therapy sessions to detect emotional distress in real-time. ¹¹ This information can help therapists adjust their approach dynamically, providing more personalized and effective care. Furthermore, AI-assisted mindfulness and meditation applications have been shown to reduce stress and anxiety in cancer patients. ¹² By incorporating adaptive AI algorithms, these apps can tailor relaxation exercises and guided meditations based on individual user preferences, real-time stress levels, and engagement patterns.

AI in Monitoring and Long-Term Mental Health Support

Recent advancements in wearable technology have enabled AI to continuously monitor physiological indicators of emotional distress, such as heart rate variability, sleep patterns, and activity levels. ¹³ Smartwatches and biosensors embedded in fitness trackers can collect real-time biometric data, allowing AI algorithms to identify early signs of stress, anxiety, or depressive episodes. For example, machine learning models can analyze heart rate fluctuations and sleep disruptions to detect when a patient with breast cancer is experiencing heightened distress, triggering an automated intervention, such as a relaxation prompt or a check-in from a mental health professional. ¹⁴ AI can often detect and differentiate patterns suggestive of mental versus physical distress using multimodal, contextualized data streams—but its accuracy depends on the quality, integration, and interpretation of that data. As reported in Table 1, these AI-driven monitoring systems offer a noninvasive way to track mental health trends over time, providing valuable insights for both patients and clinicians.

AI is also being used to develop personalized therapy recommendation systems, which analyze patient behavior, emotional responses, and past interactions to suggest the most effective interventions. For instance, AI-powered recommender systems can analyze patient engagement with various mental health resources—such as CBT, mindfulness exercises, or peer support groups—to determine which approaches yield the best outcomes for an individual patient. ¹⁵ Moreover, adaptive AI models can dynamically adjust therapy intensity based on real-time patient feedback and physiological data, ensuring that individuals receive the appropriate level of psychological support at any given time. These systems have the potential to optimize mental health care delivery, reduce the burden on human therapists, and improve long-term emotional well-being for breast cancer survivors.

AI-driven solutions offer significant potential in improving mental health care for patients with breast cancer by facilitating early detection, personalized interventions, and continuous psychological monitoring. NLP and predictive modeling allow for early identification of at-risk patients, AI-powered chatbots and telepsychiatry tools enhance access to psychological support, and wearable AI technologies provide real-time monitoring and adaptive therapy recommendations. While challenges remain, such as ensuring AI ethics, data privacy, and clinical integration, the ongoing advancements in AI-driven mental health interventions hold promise for enhancing the emotional resilience and psychological well-being of individuals facing breast cancer.

Benefits of AI in Psycho-oncology

The integration of AI in psycho-oncology offers significant advantages, particularly in improving accessibility, personalization, efficiency, and cost-effectiveness of mental health care for patients with breast cancer. Given the high prevalence of anxiety, depression, and post-treatment psychological distress, AI-driven interventions can help bridge existing gaps and enhance the quality of care.

One of the most compelling benefits of AI in psycho-oncology is its ability to extend mental health services to underserved populations. Traditional psychological care for cancer patients is often limited by geographic, financial, and logistical barriers, particularly in rural or low-resource settings. ¹⁹ As presented in Table 1, AI-based mental health solutions, including chatbots, telepsychiatry, and predictive analytics, enable wider access to support systems, ensuring that more patients receive timely interventions.

AI-driven tools provide a level of personalization that traditional mental health interventions often lack. ¹⁸ By leveraging machine learning algorithms and real-time patient data, AI can tailor interventions based on an individual’s unique psychological profile, treatment history, and emotional responses. ¹⁸ Personalized recommendations, such as adaptive therapy modules or mindfulness interventions, increase treatment adherence and efficacy.

AI has the potential to alleviate the workload of oncologists and mental health providers by automating mental health screenings, symptom monitoring, and preliminary psychological assessments. ²⁰ AI-powered tools can detect distress through patient-reported outcomes, voice analysis, and biometric indicators, allowing clinicians to focus on high-risk patients who require more intensive intervention. This targeted approach optimizes clinician time and resources while ensuring that psychological care is integrated into the cancer treatment continuum.

AI-driven mental health tools provide a cost-effective solution for addressing the large-scale psychological needs of patients with breast cancer. ²¹ By automating routine assessments and self-guided therapy, AI can reduce health care costs associated with traditional psycho-oncology services. ²¹ Furthermore, these tools have the potential to be scaled up, potentially making psychological support accessible to a larger patient population without requiring significant expansion of mental health workforce resources.

Challenges and Ethical Considerations

While AI holds promise for psycho-oncology, its implementation comes with significant challenges and ethical considerations. Ensuring privacy, fairness, clinical integration, and patient trust are critical for the responsible use of AI in mental health care. Given the sensitive nature of mental health data, protecting patient information is paramount. AI systems that collect and analyze psychological distress indicators must adhere to strict data privacy regulations, such as Health Insurance Portability and Accountability Act and General Data Protection Regulation. ²² Without proper safeguards, AI-based mental health applications could be vulnerable to data breaches, unauthorized access, or misuse of personal health information. Establishing robust encryption, anonymization techniques, and clear consent policies is essential for ensuring patient safety.

AI models are only as good as the data they are trained on, and bias in training datasets can lead to disparities in mental health care delivery. Studies have shown that AI algorithms may underperform for minority, low-income, or underrepresented patient populations, leading to inequities in access and care recommendations. ²³ Addressing this challenge requires the development of more inclusive AI models, trained on diverse datasets that reflect the full spectrum of patients with breast cancer. Additionally, continuous auditing of AI predictions can help mitigate potential biases in mental health screenings and interventions.

A significant barrier to AI adoption in psycho-oncology is the hesitancy of health care providers to trust and integrate AI recommendations into clinical workflows. Many oncologists and mental health professionals remain skeptical about AI’s ability to provide accurate, empathetic, and context-aware psychological assessments. This skepticism is not unfounded: current AI models, while powerful, are often trained on limited or nonrepresentative datasets that may not capture the full diversity of psychological experiences across cancer populations. ²⁴ As a result, there is a risk of false positives or missed signals, especially in cases where symptoms of psychological distress overlap with physical side effects of treatment (e.g., fatigue, appetite changes, or sleep disruption). AI systems may also misinterpret cultural, linguistic, or personal expressions of distress, particularly when NLP models are not adequately fine-tuned. For example, AI tools have shown promise in breast cancer care, where they can detect distress through physiological data, like heart rate fluctuations and sleep disturbances. These advancements are important given the well-documented psychological burden experienced by patients with breast cancer. However, expanding these models to serve patients with other cancers—such as lung, colorectal, or hematological malignancies—poses new challenges. These groups may experience different treatment regimens, disease trajectories, and emotional responses, which necessitate refined AI models capable of context-sensitive predictions. Moreover, psychological states are deeply nuanced and often influenced by interpersonal, existential, and environmental factors that remain difficult to quantify or embed into algorithms. Without access to rich contextual data—including patient narratives, clinical judgment, and therapeutic rapport—AI tools may over-rely on surface-level indicators, such as biometrics or self-report checklists, leading to reduced specificity and empathy in care recommendations. Successful integration will require not only clinician training and user-friendly interfaces but also the development of hybrid care models that prioritize AI as an augmentative tool—supporting but never replacing the human capacity for attunement, ethical reasoning, and relational depth.

Patients may have concerns about AI replacing human empathy in mental health care. Psychological support, particularly in the context of cancer, relies heavily on compassionate communication and therapeutic relationships, which AI alone cannot fully replicate. ²⁵ A human-AI hybrid model, in which AI enhances clinical decision-making and provides supplementary support rather than replacing human interactions, is likely to increase patient trust and engagement. This highlights the critical need for AI to function within a robust clinical decision support framework. Rather than viewing AI as a replacement for clinicians, it should be positioned as a collaborative partner that assists in synthesizing complex data, flagging potential concerns, and augmenting human insight. In this model, clinicians retain the final interpretive authority, integrating AI-driven insights with their nuanced understanding of each patient’s psychosocial context. Successful integration will therefore require not only clinician training and user-friendly interfaces but also a reimagining of care workflows to support human-AI collaboration. This includes establishing protocols for clinical oversight, building trust through transparent algorithms, and ensuring that AI remains accountable to human ethical standards. By strengthening this human-AI partnership, psycho-oncology can move toward more precise, responsive, and compassionate mental health support for patients across the cancer care continuum.

Future Directions and Research Priorities

The future of AI in psycho-oncology depends on advancing its capabilities, improving transparency, and addressing ethical concerns. Several key areas for future research and development include the following:

Emotion recognition technology represents a promising frontier in AI-driven mental health support for patients with breast cancer. Future AI applications could incorporate facial recognition, voice analysis, and biometric tracking to provide real-time, objective assessments of emotional distress. ²⁵ Unlike self-reported distress measures, which may be influenced by response bias or underreporting, AI-driven emotion recognition can detect subtle, involuntary cues that indicate anxiety, depression, or psychological distress. Facial expression analysis, for instance, can track micro-expressions, muscle movement, and eye activity, which have been correlated with emotional states. ²⁶

Similarly, voice analysis can detect changes in tone, speech patterns, and pitch variability, which may indicate distress or mood fluctuations. ²⁷ AI-powered sentiment analysis tools can be integrated into telehealth platforms or virtual therapy sessions to assess emotional well-being and provide real-time feedback to clinicians or mental health chatbots. Additionally, biometric tracking—using smartwatches or wearable sensors—can monitor heart rate variability, skin conductivity, and breathing patterns, all of which are physiological indicators of stress and emotional dysregulation. ²⁸ By integrating these multimodal emotional recognition techniques, AI can enhance early detection of psychological distress in patients with breast cancer, enabling timely interventions and personalized support. Future developments should focus on refining accuracy, reducing bias in emotion AI models, and ensuring that these technologies are ethically deployed, respecting patient privacy and autonomy. ²⁴

Multimodal AI interventions hold immense potential for enhancing mental health support in breast cancer care by combining multiple AI-driven technologies to create a holistic, interactive treatment experience. Traditional mental health interventions often rely on text-based assessments or verbal therapy sessions, but AI advancements allow for the integration of wearable sensors, virtual reality (VR), and NLP to provide a more immersive and adaptive intervention framework. ²⁹ For example, VR-based mindfulness programs have demonstrated efficacy in reducing stress, anxiety, and cancer-related fatigue. ³⁰ When combined with AI-driven emotion recognition, these VR environments can dynamically adjust based on real-time physiological data, such as heart rate variability and skin temperature, to tailor relaxation exercises or guided meditations to match the patient’s current emotional state. Similarly, wearable sensors can continuously track psychological well-being and trigger personalized digital interventions, such as suggesting breathing exercises when signs of stress are detected. ¹³

Furthermore, NLP-powered AI chatbots can enhance multimodal interventions by providing conversational therapy in conjunction with VR environments. A patient using a VR relaxation program could interact with an AI-powered assistant that offers real-time emotional support, mindfulness coaching, or CBT-based interventions. These AI-driven multimodal approaches increase accessibility, particularly for patients in remote areas or those who face barriers to traditional therapy, while also improving engagement and intervention effectiveness. Future research should focus on refining personalization algorithms, integrating real-time biofeedback, and ensuring equitable access to these advanced AI-driven mental health tools. ³¹ For AI to be widely accepted and effectively integrated into mental health care for patients with breast cancer, it must be explainable, transparent, and interpretable. The concept of explainable AI (XAI) refers to the ability of an AI system to provide clear, understandable, and justifiable reasoning for its predictions and recommendations. ³² In psycho-oncology, where mental health interventions require nuanced, patient-centered approaches, the ability of AI to explain why it makes specific decisions is crucial for both clinicians and patients.

Lack of transparency in AI, often referred to as the “black box” problem, can lead to skepticism and hesitation among oncologists and mental health providers in trusting AI-generated recommendations. ³³ If AI models cannot justify their predictions, clinicians may be reluctant to rely on them when making critical mental health decisions, such as identifying patients at high risk for depression or PTSD. To address this challenge, AI-driven mental health systems should utilize interpretable machine learning models, such as decision trees, attention-based neural networks, and feature attribution methods, to clarify how specific variables—such as treatment side effects, sleep patterns, or patient-reported distress—contribute to an AI-generated assessment. ³⁴

Beyond clinician trust, patient engagement also depends on AI transparency. Patients with breast cancer utilizing AI-powered mental health tools should receive clear, patient-friendly explanations for their distress scores, therapy recommendations, or chatbot responses. For instance, an AI system detecting heightened anxiety based on biometric data from a smartwatch should explain the physiological indicators involved, such as elevated heart rate and disrupted sleep patterns, and suggest actionable coping strategies, such as guided breathing exercises or therapy referrals. Such transparency empowers patients to make informed decisions about their mental health care and fosters confidence in AI-assisted interventions.

Strategies to enhance AI explainability in mental health include developing intuitive AI interfaces, integrating human-AI collaboration models, and establishing regulatory frameworks for XAI in psycho-oncology. ³⁵ Transparent AI dashboards with visual explanations and interactive feedback can help both patients and providers understand AI outputs. Additionally, AI should be positioned as a decision-support tool rather than a replacement for human clinicians, ensuring that its recommendations are subject to clinical judgment and ethical considerations. Regulatory bodies, including the U.S. Food and Drug Administration and the World Health Organization, are working toward ethical AI guidelines to address transparency, bias, and fairness, which will be essential for AI’s responsible deployment in psycho-oncology. ²¹

By prioritizing explainability and transparency, AI can enhance trust, optimize mental health interventions, and ultimately improve psychological outcomes for patients with breast cancer. Future research should focus on developing interpretable models, ensuring diverse and representative training datasets, and refining AI-human collaboration in mental health care to maximize AI’s potential while maintaining the human empathy essential to psycho-oncology.

Establishing comprehensive ethical guidelines for AI in psycho-oncology will be essential. Policies should address data privacy, bias mitigation, and accountability frameworks to ensure AI tools are safe, equitable, and clinically validated. ²¹

Conclusion

AI has the potential to revolutionize mental health care for patients with breast cancer, offering greater accessibility, personalized interventions, and real-time psychological support. AI-based tools can help identify distress early, tailor interventions to individual needs, and scale mental health services to reach more patients. However, challenges such as data privacy, algorithmic bias, clinical integration, and patient trust must be addressed to ensure responsible and ethical AI deployment in psycho-oncology.

Future research should focus on improving AI’s accuracy, transparency, and emotional intelligence, as well as establishing ethical frameworks to guide its integration into oncology care. Ultimately, a human-AI hybrid model, in which AI enhances clinical decision-making while preserving human empathy and therapeutic relationships, represents the ideal approach to advancing mental health support for patients with breast cancer.

Authors' Contributions

J.K.P.: Conceptualization, methodology, writing and reviewing. D.R.P.: Conceptualization, methodology, writing and reviewing. J.B.: Methodology, writing and reviewing.