A software framework for adaptive and interoperable internet-delivered psychological treatments

Abstract

BACKGROUND Statistics unveil the predominance of mental and neurological disorders globally. Handling these mental and neurological disorders is economically, physically and emotionally challenging. Proper healthcare treatments would have been an ideal solution for these people suffering from these disorders. However, provided limited healthcare resources, an alternative solution is to use the Internet to provide psychological treatments. The use of Internet-Delivered Psychological Treatments (IDPT) can accelerate treatments for people globally at a lower cost. While such IDPT systems have been practised at volume, user adherence is low with high dropout rates. Low adherence in treatments is primarily due to the IDPT system’s inability to adapt treatments according to user needs, context and preferences.

OBJECTIVE This study is accomplished in collaboration with a large interdisciplinary project entitled as INTROMAT. INTROMAT brings together ICT researchers, ICT industries, health researchers, patients, clinicians, and patients next of kin to reach its vision. The project’s vision is to improve public mental health through innovative technologies. The main objective of this thesis, inclined to fulfil this vision, is to design, develop and evaluate an adaptive IDPT framework.

METHODS Based on the INTROMAT project’s problem domain and goals, we started with the study of state-of-the-art works, including systematic literature review, evaluation of available systems, analysis of the previous studies on the problem domain, and evaluation of past case studies. Based on these studies, we identified two primary gaps in the current IDPT systems, lack of adaptiveness and limited interoperability. Then, we used MDE and Domain-Driven Design (DDD) techniques to address these two gaps.

RESULTS We proposed a software framework for developing adaptive, reusable, and interoperable Internet-Delivered Psychological Treatments (IDPT), referred hereto as OpenIDPT Framework. The OpenIDPT Framework includes a) a Reference Model (RM), b) a Reference Architecture (RA), c) an Information Architecture (IA), and d) an open-source implementation of an adaptive IDPT system.The reference model reveals the adaptive elements (what to adapt), adaptive dimensions (on what basis to adapt), information architecture (how to structure content), and strategies (how to adapt) of an adaptive IDPT system. The Reference Architecture unveils the technical architecture of an adaptive IDPT system. The information architecture guides how to structure and organize the content for better discoverability and comprehensibility. To evaluate the proposed RA of adaptive IDPT systems, we implemented a prototype as an Open-Source Software. We refer to it as Open-Source Adaptive IDPT System (OSAIS). We used Design Science Research (DSR) evaluation methods to assess the efficacy of the proposed artefacts and their ability to address identified research gaps. Our preliminary results demonstrate that the proposed artefacts exhibit capabilities to use comprehensive user profiling techniques to adapt interventions using different rule-based engines, recommendation systems, and artificial intelligence (AI) based algorithms. As a Proof-of-Concept of AI-based algorithms, we present an adaptive strategy based on Natural Language Processing (NLP) techniques that analyze patientauthored text data and extract depression symptoms corresponding to a clinically established psychometric assessment questionnaire PHQ-9. The strategy utilizes the proposed novel Word Embedding (Depression2Vec) to extract depression symptoms from patient-authored text and adapts psychological treatments based on the absence or presence of depression symptoms. Furthermore, to obtain interoperability in the OpenIDPT Framework, we created an open-source Resource Server (RS) based on GraphQL. An RS is a web application that can read, write, update, and delete (CRUD) HL7 FHIR resources. HL7 FHIR is an open healthcare IT standard (analogous to data structure) for healthcare data exchange. GraphQL is a data query and manipulation language for Web-Service communications.

CONCLUSION This study demonstrates the feasibility of using an adaptive system to enhance user adherence. With the ubiquity of ambient intelligence and predictive algorithms, further study on how to combine these IoT technologies with the adaptive system is prudent and exciting.