We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Adherence Forecasting for Guided Internet-Delivered Cognitive Behavioral Therapy: A Minimally Data-Sensitive Approach

Abstract: Internet-delivered psychological treatments (IDPT) are seen as an effective and scalable pathway to improving the accessibility of mental healthcare. Within this context, treatment adherence is an especially pertinent challenge to address due to the reduced interaction between healthcare professionals and patients, compared to more traditional interventions. In parallel, there are increasing regulations when using peoples' personal data, especially in the digital sphere. In such regulations, data minimization is often a core tenant such as within the General Data Protection Regulation (GDPR). Consequently, this work proposes a Self-Attention-based deep learning approach to perform automatic adherence forecasting, while only relying on minimally sensitive login/logout-timestamp data. This approach was tested on a dataset containing 342 patients undergoing Guided Internet-delivered Cognitive Behavioral Therapy (G-ICBT) treatment. Of these 342 patients, 101 (~30%) were considered as non-adherent (dropout) based on the adherence definition used in this work. The proposed Self-Attention Network achieved over 70% average balanced accuracy, after only 20 out of the 56 days (~1/3) of the treatment had elapsed. This study demonstrates that automatic adherence forecasting for G-ICBT, is achievable using only minimally sensitive data, thus facilitating the implementation of such tools within real-world IDPT platforms.
Comments: Ulysse C\^ot\'e-Allard and Minh H. Pham share first authorship. Submitted for peer-review. 11 pages, 7 figures, 2 tables. The dataset and code used in this work are available here: this https URL
Subjects: Machine Learning (cs.LG); Computers and Society (cs.CY)
Cite as: arXiv:2201.04967 [cs.LG]
  (or arXiv:2201.04967v2 [cs.LG] for this version)

Submission history

From: Ulysse Côté-Allard [view email]
[v1] Tue, 11 Jan 2022 13:55:57 GMT (327kb,D)
[v2] Fri, 13 May 2022 21:27:21 GMT (379kb,D)

Link back to: arXiv, form interface, contact.