Pedestrian counting has attracted much interest of the academic and industry communities for its widespread application in many real-world scenarios. While many recent studies have focused on computer vision-based solutions for the problem, the deployment of cameras brings up concerns about privacy invasion. This paper proposes a novel indoor pedestrian counting approach, based on footstep-induced structural vibration signals with piezoelectric sensors. The approach is privacy-protecting because no audio or video data is acquired. Our approach analyzes the space-differential features from the vibration signals caused by pedestrian footsteps and outputs the number of pedestrians. The proposed approach supports multiple pedestrians walking together with signal mixture. Moreover, it makes no requirement about the number of groups of walking people in the detection area. The experimental results show that the averaged F1-score of our approach is over 0.98, which is better than the vibration signal-based state-of-the-art methods.

Triple negative breast cancer (TNBC) is a highly aggressive type of breast cancer with poor prognosis that accounts approximately for 15% of breast cancer cases. The lack of targetable hormonal receptors or HER2/ErbB2 makes TNBC therapeutically challenging and existing drug treatments are often ineffective. TNBC is also a highly heterogeneous class of breast cancer, and different genetic approaches have not been so far successful in identifying clinically useful patient stratification strategies to date. In previous studies, we have discovered a correlation between targeted phospho-proteomics and mitotic inhibitor sensitivities in TNBC cells, suggesting the potential of using phospho-proteome profiles to identifying markers predictive of actionable therapeutic strategies. In the current study, we further explored the potential of phospho-proteomics for TNBC stratification and drug chemo-sensitivity prediction. We screened 25 TNBC cell lines against 517 oncology compounds, and in parallel ran mass spectrometry-based global phospho-proteome profiling of the same cell lines, resulting in 4283 confident phospho-peptide quantifications. Using multitask multiple kernel learning, we integrated the phospho-peptide data with other publicly available omics data for the 25 TNBC cell lines (genomics, transcriptomics, metabolomics, and methylomics), and developed a multi-omics machine learning model to quantify the predictive contribution of each omics profile to drug response prediction. When considering all the 517 compounds, we found that the phospho-proteomic and gene expression data were most informative for the prediction of compound’s chemo-sensitivity. This integrated approach has the potential to both identify panels of multi-omics markers for individual drug classes, as well as to stratify the heterogeneous TNBC cells based on their genomic, molecular and functional profiles. Citation Format: Prson Gautam, Xiangju Qin, Julia Vainonen, Srikar Nagelli, Johanna Lampe, Juha Klefström, Jukka Westermarck, Krister Wennerberg, Tero Aittokallio. Therapeutic stratification of triple negative breast cancer by integrating chemosensitivity & phospho-proteome profiles [abstract]. In: Proceedings of the 2019 San Antonio Breast Cancer Symposium; 2019 Dec 10-14; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2020;80(4 Suppl):Abstract nr P6-10-28.