The following is the summary of  “A prospective observational study for a Federated Artificial Intelligence solution for moniToring mental Health status after cancer treatment (FAITH): study protocol” published in the December 2022 issue of Psychiatry by Lemos, et al.


A number of stages in the course of cancer are associated with elevated rates of depression. Yet, despite having higher prevalence rates than the general population, it is often not reported or goes unrecognized. In addition, depression’s somatic symptoms are relatively uncommon in the oncological setting and should not be disregarded as a generic symptom of cancer. After therapy ends, patients often have inconsistent contact with the healthcare system, making it even more difficult to monitor their mental health. Using a federated machine learning (ML) technique, the FAITH project aims to anticipate the onset of depression symptoms in cancer survivors from a distance, with the ultimate goal of protecting their anonymity.

Remotely, FAITH will analyze depression indicators and forecast downward tendencies. Nutrition, sleep, activity, and speech are all biomarkers that can be evaluated in diverse ways, partly by wearable devices. Patients with a history of breast or lung cancer will be recruited for the trial between 1 and 5 years after their initial cancer has been treated. Depression and quality of life among cancer survivors will be evaluated over the course of a 12-month prospective observational cohort study that will include monthly assessments. The severity of depressive symptoms at 3, 6, 9, and 12 months as determined by the Hamilton Depression Rating Scale (Ham-D), is the primary objective. Baseline and monthly self-reports of anxiety, depression, and quality of life using the Hospital Anxiety and Depression Scale (HADS) and EORTC questionnaires are secondary objectives. 

FAITH aims to further create a conceptual federated learning framework based on the study’s predictive models, allowing for constructing machine learning models to predict and monitor depression without requiring direct access to user-specific data. There is a need for increased objectivity in psychiatric evaluation. Indicators of depression and anxiety, as well as biofeedback tools, can be gleaned from wearable devices. If the FAITH app proves useful, oncology clinics will have access to a cutting-edge tool for screening patients for depression.

Source: bmcpsychiatry.biomedcentral.com/articles/10.1186/s12888-022-04446-5