Session Item

Radiotherapy (RT)-induced fatigue (RIF) is a clinical subtype of cancer treatment-related effect potentially impacting on patients’ Quality of Life (QoL). Wearable technologies, already available on the market, allow to monitor patients while accomplishing daily activities and to collect qualitative measurements of fatigue related biomedical parameters. Aim of our study was to infer biomedical parameters over RT treatment time able to evaluate RIF.

A cohort of 35 patients undergoing RT for breast cancer in the Department of Radiation Oncology (DRO) at a single institution was supplied with wearable Fitness Trackers (FT). Using FT, data (heart rate (HR), steps walked (STP) and sleep level (SLP)) was non-invasively collected. The gathered information is stored in the FT’s internal memory. Patients coming daily to the DRO for RT have data downloaded from FT’s memory via gateway devices (GW) placed in the waiting room being part of an architecture built ad hoc following the paradigm of pervasive computing. Collected data is sent to a central server and made available to medical staff through a web page. An in house script running on the server analyzes data to recognize repeated activity windows based on STP and HR.

Applying the described methodology, it was possible to recognize different repeated activity windows for each patient over the RT treatment days. Each patient generally shows from three to five repeated activity windows in a day with at least one in the morning and two in the afternoon. The repeatability of the recognized windows was evaluated in relation to durations and time placements. For each single patient and within her activity windows, inference has been made on HR and STP variation highlighting a negative trend for the median STP and an increase of HR median values. Using FT, we were not able to identify a trend for SLP.

Using smart wearable devices already available on the market it is possible to non-invasively collect qualitative biomedical data related to patient activity windows. By monitoring HR and STP median trends during such windows, medical staff can perform an evaluation of RIF based on objective data. As regard to SLP, more suitable devices are needed to get to a reliable study on sleep quality variation during RT time. Future developments of our research is focusing on ballistocardiograph data collected with the help of non-invasive pressure sensors under patient beddings.