Joint prediction of cocaine craving and euphoria using structured prediction energy networks
Joint prediction of cocaine craving and euphoria using structured prediction energy networks
复制标题
使用结构化预测能量网络联合预测可卡因渴望和欣快感
DOI：
10.1145/3469266.3469881
复制
发表时间：
2021
期刊：
影响因子：
--
通讯作者：
Rahman, Tauhidur
中科院分区：
文献类型：
--
作者：
Gullapalli, Bhanu Teja;Angarita, Gustavo A;Ganesan, Deepak;Rahman, Tauhidur
近年来，可穿戴和移动健康传感技术已经发展到跟踪药物使用和监测不同的成瘾相关状态，包括渴望和欣快感。这些状态是相互依存和相关的，这在文献中有很好的记载。然而，最先进的数字生物标志物技术相互独立地模拟这些状态，因此在进行预测时无法使用固有关系。在我们目前的工作中，我们展示了如何使用结构化预测能量网络（SPENs）来捕捉自我报告的渴望、欣快感和潜在的生理生物标志物之间的相关性和依赖性。更具体地说，我们使用SPENs来共同预测可卡因渴望和欣快感的自我报告视觉模拟量表（VAS）评分，这些评分来自可穿戴胸带捕获的心脏信号。提出的基于spn的模型可以提高VAS渴望和VAS欣快度的预测性能，标准化均方根误差分别为4.6%和5.4%。
In recent years, wearable and mobile health sensing technologies have been developed to track drug usage and monitor different addiction-related states, including craving and euphoria. These states are interdependent and correlated, which is well documented in the literature. However, the state of the art digital biomarker technologies model these states independent of each other and thus fail to use the inherent relationship while making predictions. In our current work, we demonstrate how structured prediction energy networks (SPENs) can be used to capture the correlation and dependencies between self-reported craving, euphoria, and the underlying physiological biomarkers. More specifically, we use SPENs to jointly predict self-reported visual analog scale (VAS) ratings of cocaine craving and euphoria from cardiac signals captured from a wearable chest band. The proposed SPEN-based model can improve the performance of both VAS craving and VAS euphoria prediction by a Normalized Root Mean Square Error of respectively 4.6\% and 5.4\%.
使用可穿戴传感器识别吸烟事件中的抽吸现象的 CNN-LSTM 神经网络
影响因子：
4.6
作者：
V. Senyurek;Masudul H. Imtiaz;Prajakta Belsare;Stephen T. Tiffany;E. Sazonov
通讯作者：
E. Sazonov
DOI：
10.1145/3328917
发表时间：
2019-06
影响因子：
--
作者：
Bhanuteja Gullapalli;A. Natarajan;G. Angarita;R. Malison;Deepak Ganesan;Tauhidur Rahman
通讯作者：
Bhanuteja Gullapalli;A. Natarajan;G. Angarita;R. Malison;Deepak Ganesan;Tauhidur Rahman