AI Model May Accurately Predict Mental Health Outcomes in Patients With Cancer

Get Permission

A novel artificial intelligence (AI) model may accurately predict which patients with cancer may require mental health services during cancer care, according to a recent study published by Nunez et al in Communications Medicine.


Mental health has been shown to have a significant impact on treatment outcomes and quality of life among patients with cancer. On average, patients with worse depression and anxiety often experience poorer survival outcomes—likely because these patients have more difficulties following through with treatment recommendations and tolerating side effects.

Current estimates from the Canadian Association of Psychosocial Oncology suggest that approximately 15% of patients with cancer may require psychiatric services and 45% would benefit from visiting a counselor. However, barriers such as stigma, lack of awareness of services, and difficulties diagnosing mental health conditions can prevent patients from accessing psychosocial care.

“Battling cancer can be a harrowing experience, affecting not only [patients’] bodies but [their] minds and emotions,” stressed lead study author John-Jose Nunez, MD, MSc, FRCPC, a psychiatrist and clinical research fellow at the University of British Columbia Mood Disorders Centre and BC Cancer.

Study Methods and Results

In the recent study, an interdisciplinary team of researchers with expertise in medical oncology, psychiatry, and computer science developed the AI model. The model was then trained and tested using data from 59,800 patients across six cancer centers located in geographically diverse regions of British Columbia. To protect the patients’ privacy, their data were stored securely and presented anonymously. Unlike patient chart reviews by human research assistants, the new approach has the added benefit of maintaining complete confidentiality of patient records.

The AI model was designed to use language processing and advanced neural networks to analyze oncologists’ notes from the initial consultation appointments with their patients. Although these appointments are typically focused on the patients’ medical history and treatment options, the researchers found that the AI model was capable of detecting subtle clues hidden within the medical language to indicate which patients may benefit from early psychiatric or counseling interventions. The model predicted with greater than 70% accuracy whether patients would visit a psychiatrist or counselor within 1 year.

Additionally, the researchers made significant advances in AI interpretability, developing a new technique that allowed them to peer inside the model to see how it was making predictions. They identified common themes with certain key words and topics suggesting a greater need for mental health services: family history of cancer, patterns of alcohol or substance use, and certain types of aggressive cancers and treatment strategies.


“These findings demonstrate the tremendous potential of AI to essentially act as a personal assistant to oncologists, enhancing patient care by helping identify mental health needs sooner and ensuring more patients receive the support they need,” Dr. Nunez said.

The researchers hope to collaborate with oncologists and patients to explore how the novel AI model could be implemented to improve early access to psychiatric services.

“Perhaps the AI prompts the oncologist to talk with the patient about psychiatric services, or it could trigger an e-mail to patients with a tailored list of available services. There are a lot of possibilities, but it’s important that we work with health professionals and patients to ensure the AI is meeting [patients] where they need it,” Dr. Nunez emphasized.

“Neural models have been commonly called black boxes because it’s difficult to interpret how they arrive at an answer,” stated co–study author Raymond Ng, BSc (Hons), PhD, Professor in the Department of Computer Science at the University of British Columbia. “These cutting-edge techniques allow us to unravel these mysteries for multiple documents at a time, which could further empower researchers and clinicians to understand the complex intersections of oncology and mental health.”

The researchers plan to expand the novel AI model’s use beyond oncology to other fields of medicine where psychosocial factors may significantly impact patient outcomes. The findings could pave the way for a broader application of AI in health care, targeting early intervention across various medical disciplines.

Disclosure: For full disclosures of the study authors, visit

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.