Mammography, despite its widespread use, suffers from drawbacks including ionizing radiation exposure and decreased accuracy in dense breast tissue, limiting its effectiveness in key populations such as younger women. Alternative modalities such as thermography and bioimpedance spectroscopy have been explored for decades as adjunct or complementary screening options.
Thermography uses thermal sensors to detect abnormal heat patterns caused by increased metabolic activity in tumors, though its clinical adoption has been hindered by historically variable sensitivity and specificity.
Bioimpedance techniques measure the electrical properties of tissues, reflecting changes in cellular composition and membrane permeability characteristic of malignancies. However, challenges persist in obtaining reliable bioimpedance readings in vivo due to technical issues, including electrode contact stability and signal noise.
PHI-BRA Device & Study Design
The PHI-BRA Smart Bra utilizes a flexible sensor cap embedded with two types of sensors: 60 temperature sensors (MAX30205MTA+T) arranged across 12 branches, and 60 bioimpedance electrodes configured similarly. The temperature sensors are designed to capture spatially resolved thermal information over the breast surface, detecting localized variations that may indicate abnormal vascular activity linked to tumor growth.
The bioimpedance system measures tissue resistance and reactance across three frequencies (4, 40, and 80 kHz), chosen to probe extracellular matrix properties, membrane permeability, and intracellular composition, respectively. Steel electrodes embedded in the sensor cap are connected through multiplexers (ADG731) to capture these electrical properties. The device incorporates onboard analog-to-digital conversion and digital communication modules, ensuring signal integrity.
To facilitate clinical use, the sensor array is sized to accommodate common bra sizes with cups ranging from B to E. The system is designed to operate continuously for up to 8 days on battery power, making it easy to use in clinical and home environments. Measures were taken to ensure biocompatibility, including the use of stainless-steel electrodes and biocompatible PCB coatings, and electromagnetic compatibility was achieved through integrated resistor terminations.
The study was conducted as a prospective, two-phase clinical trial involving a calibration cohort (n=15) to develop diagnostic models and an analysis cohort (n=26) to blind-test the selected models. Participants included women with imaging-confirmed breast lesions and control subjects without lesions. Sensor data were collected under semi-standardized conditions using a sports bra to secure the device, with participants in a semi-seated position during an approximately 10-minute measurement session.
Diagnostic Performance & Insights
The thermal sensor array demonstrated robust technical feasibility in capturing physiologically relevant breast temperature data. Using temperature data alone, the logistic regression model yielded an area under the ROC curve (AUC) of approximately 80.8%, with sensitivity of approximately 85% and specificity of approximately 77% at the optimal discrimination threshold.
Bioimpedance measurements, while conceptually valuable, showed less consistent diagnostic performance. Challenges primarily arose from the stability of electrode-to-skin contact, which affected, in particular, low-frequency measurements critical for probing cellular membrane integrity. Despite this, the specificity of the bioimpedance model remained relatively high, suggesting that improved sensor interface design could augment diagnostic accuracy.
Combining temperature and bioimpedance data improved discrimination during the calibration phase but was not included in the final analysis due to the bioimpedance data’s moderate reliability. The study underscored that improving electrode contact through more secure, comfortable sensor integration is necessary to harness the full potential of electrical impedance measurements in wearable devices.
Notably, the sensor system operated without adverse events, indicating good biocompatibility and safety of the materials and sensor design. However, environmental variability - such as room temperature differences - during sensor data acquisition could have influenced thermal readings, highlighting the need to incorporate ambient temperature control or compensation mechanisms in future work.
Study limitations included a small sample size, single-center recruitment, and an atypical 1:1 ratio of lesion to non-lesion participants, which limited extrapolation of predictive value. The authors recommend larger multisite trials with standardized measurement protocols, improved sensor mechanics for stable contact, and integration of environmental sensing to better validate and optimize the technology.
Download the PDF of this page here
Implications and Future Directions
The PHI-BRA Smart Bra prototype successfully integrated arrays of temperature sensors and bioimpedance electrodes into a wearable form factor feasible for clinical use. The study validates the fundamental sensor technology and combined sensing approach, supporting further development of wearable multimodal sensor systems for breast cancer detection.
Ultimately, wearable sensor platforms like the PHI-BRA could transform breast monitoring by enabling safe, non-invasive, and frequent assessments, potentially improving early detection, especially in populations underserved by traditional mammography. The technology’s repeatability and patient comfort present promising avenues for longitudinal breast health management outside conventional clinical settings.
Journal Reference
Belmont A-S, Moreno M-V, et al. (2026). Assessing the Diagnostic Performance of a Smart Bra Using Temperature and Bioimpedance for Breast Cancer Detection: A First-in-Human Study. Sensors 26(9):2869. DOI: 10.3390/s26092869, https://www.mdpi.com/1424-8220/26/9/2869