Social Media Discussions Predict Mental Health Consultations on College Campuses

The mental health of college students is a growing concern, and gauging the mental health needs of college students is difficult to assess in real-time and in scale. To address this gap, researchers and practitioners have encouraged the use of passive technologies. Social media is one such "passive sensor" that has shown potential as a viable "passive sensor" of mental health. However, the construct validity and in-practice reliability of computational assessments of mental health constructs with social media data remain largely unexplored. Towards this goal, we study how assessing the mental health of college students using social media data correspond with ground-truth data of on-campus mental health consultations. For a large U.S. public university, we obtained ground-truth data of on-campus mental health consultations between 2011–2016, and collected 66,000 posts from the university’s Reddit community. We adopted machine learning and natural language methodologies to measure symptomatic mental health expressions of depression, anxiety, stress, suicidal ideation, and psychosis on the social media data. Seasonal auto-regressive integrated moving average (SARIMA) models of forecasting on-campus mental health consultations showed that incorporating social media data led to predictions with r = 0.86 and SMAPE = 13.30, outperforming models without social media data by 41%. Our language analyses revealed that social media discussions during high mental health consultations months consisted of discussions on academics and career, whereas months of low mental health consultations saliently show expressions of positive affect, collective identity, and socialization. This study reveals that social media data can improve our understanding of college students’ mental health, particularly their mental health treatment needs.

Twin Meander Coil

Energy-efficient and unconstrained wearable sensing platforms are essential for ubiquitous healthcare and activity monitoring applications. This paper presents Twin Meander Coil for wirelessly connecting battery-free on-body sensors to a textile-based reader knitted into clothing. This connection is based on passive inductive telemetry (PIT), wherein an external reader coil collects data from passive sensor coils via the magnetic field. In contrast to standard active sensing techniques, PIT does not require the reader to power up the sensors. Thus, the reader can be fabricated using a lossy conductive thread and industrial knitting machines. Furthermore, the sensors can superimpose information such as ID, touch, rotation, and pressure on its frequency response. However, conventional PIT technology needs a strong coupling between the reader and the sensor, requiring the reader to be small to the same extent as the sensors' size. Thus, applying this technology to body-scale sensing systems is challenging. To enable body-scale readout, Twin Meander Coil enhances the sensitivity of PIT technology by dividing the body-scale meander-shaped reader coils into two parts and integrating them so that they support the readout of each other. To demonstrate its feasibility, we built a prototype with a knitting machine, evaluated its sensing ability, and demonstrated several applications.

Examining Cloud Macrophysical Changes over the Pacific for 2007–2017 Using CALIPSO, CloudSat, and MODIS Observations

Cloud macrophysical changes over the Pacific from 2007 to 2017 are examined by combining CALIOP and CloudSat (CALCS) active-sensor measurements, and these are compared with MODIS passive-sensor observations. Both CALCS and MODIS capture well-known features of cloud changes over the Pacific associated with meteorological conditions during El Niño-Southern Oscillation (ENSO) events. For example, mid (cloud tops at 3–10 km) and high (cloud tops at 10–18 km) cloud amounts increase with relative humidity (RH) anomalies. However, a better correlation is obtained between CALCS cloud volume and RH anomalies, confirming more accurate CALCS cloud boundaries than MODIS. Both CALCS and MODIS show that low cloud (cloud tops at 0–3 km) amounts increase with EIS and decrease with SST over the eastern Pacific, consistent with earlier studies. It is also further shown that the low cloud amounts do not increase with positive EIS anomalies if SST anomalies are positive. While similar features are found between CALCS and MODIS low cloud anomalies, differences also exist. First, compared to CALCS, MODIS shows stronger anti-correlation between low and mid/high cloud anomalies over the central and western Pacific, which is largely due to the limitation in detecting overlapping clouds from passive MODIS measurements. Second, compared to CALCS, MODIS shows smaller impacts of mid and high clouds on the low troposphere (< 3 km). The differences are due to the underestimation of MODIS cloud layer thicknesses of mid and high clouds.

Low-SNR Asymptotics of Cross-Receiver Mutual Information in Gaussian Channels

<div>In this paper, we analyze the low-SNR behavior of the cross-receiver mutual information (CMI) between two received signals corrupted by uncorrelated, additive Gaussian noise. This framework has use in distributed, passive sensor applications, such as passive radar and collaborative opportunistic navigation. For Gaussian and BPSK signaling, the CMI can be expressed in terms of the effective SNR between the receivers. On-off keying (OOK), while not optimal in terms of spectral efficiency for a single-receiver channel, is shown to have greater CMI than Gaussian or BPSK signaling. This is in spite of the fact that, given the same received SNRs, all three source distributions have the same linear correlation coefficient. This indicates that for OOK sources, effective SNR and correlation coefficient are not meaningful descriptors for passive receivers.<br></div><div><br></div><div>Full-length version of conference paper submission.</div>

Low-SNR Asymptotics of Cross-Receiver Mutual Information in Gaussian Channels

<div>In this paper, we analyze the low-SNR behavior of the cross-receiver mutual information (CMI) between two received signals corrupted by uncorrelated, additive Gaussian noise. This framework has use in distributed, passive sensor applications, such as passive radar and collaborative opportunistic navigation. For Gaussian and BPSK signaling, the CMI can be expressed in terms of the effective SNR between the receivers. On-off keying (OOK), while not optimal in terms of spectral efficiency for a single-receiver channel, is shown to have greater CMI than Gaussian or BPSK signaling. This is in spite of the fact that, given the same received SNRs, all three source distributions have the same linear correlation coefficient. This indicates that for OOK sources, effective SNR and correlation coefficient are not meaningful descriptors for passive receivers.<br></div><div><br></div><div>Full-length version of conference paper submission.</div>