Abstract 173P
Background
Breast cancer (BC) is one of the leading causes of cancer death in women worldwide. Screening with mammography (MMG) is limited in low- and middle-income countries. The implementation of affordable and effective screening method is crucial. The Intelligent Breast Exam (iBE) has emerged as portable device with a glove-shaped using piezoelectricity. This study compares the performance of the iBE with MMG or breast magnetic resonance imaging (MRI) and Clinical Breast Examination (CBE).
Methods
The study included patients under surveillance in a Medical Oncology Unit. iBE was performed after each CBE and compared with BI-RADS classifications. MMG/MRI was classified as negative (BI-RADS ≤ 2) or positive (BI-RADS ≥ 3). Measures of accuracy and agreement between tests were calculated.
Results
Between September 2022 to 2023, 151 women were included, who underwent iBE, CBE and MMG/MRI. 48 were excluded due to damaged piezoelectric sensors. There were 29 positive results on CBE and 38 on MMG. 9 women already had a confirmed diagnosis (BI-RADS 6). CBE and MMG showed moderate agreement in categorization (ρ = 0.99). With a specificity for predicting a negative MMG of 90.8% and a negative predictive value of 79.7%. The final diagnosis of all 103 patients was assigned after MMG, CBE and additional biopsy in BIRADS 4 and 5. Benign findings were found in 65 patients, 3 cysts, 6 fibroadenomas, 8 microcalcifications, and 10 BC, excluding the 9 with a previous diagnosis. The performance of IBE was evaluated by comparing the breast with alterations to the control breast within each BI-RADS categorization.
Conclusions
The study aimed to evaluate the IBE performance in comparison with MMG and CBE, concluding that IBE does not detect breast alteration to date. As there is no correlation between iBE and MMG, we are unable to determine the sensitivity and specificity of iBE. Challenges, such as the absence of glove acceleration data, sensor quality, and low sampling rate, hindered iBE effectiveness determination. The improvement proposals emphasized the incorporation of accelerometer sensors, a signal conditioning to allow for the collection of compression and decompression data from the sensors and consideration of pressure stress. These improvements are crucial to optimizing the iBE's ability to detect changes in breast texture, enhancing the iBE's effectiveness in the early detection of BC.
Legal entity responsible for the study
Glooma.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.