Learning machines in Internet-delivered psychological treatment

Magnus Boman,Erik Forsell,Fehmi Ben Abdesslem,Viktor Kaldo,Nils Isacsson,Daniel Gillblad,Olof Görnerup,Magnus Sahlgren

doi:10.1007/s13748-019-00192-0

Abstract

A learning machine, in the form of a gating network that governs a finite number of different machine learning methods, is described at the conceptual level with examples of concrete prediction subtasks. A historical data set with data from over 5000 patients in Internet-based psychological treatment will be used to equip healthcare staff with decision support for questions pertaining to ongoing and future cases in clinical care for depression, social anxiety, and panic disorder. The organizational knowledge graph is used to inform the weight adjustment of the gating network and for routing subtasks to the different methods employed locally for prediction. The result is an operational model for assisting therapists in their clinical work, about to be subjected to validation in a clinical trial.

Highlights

Machine learning is here employed to help answer questions concerning prediction of outcome and engagement in psychological treatment
A number of machine learning methods are used in tandem, with their signal analyses fused and unified to produce decision support for the clinician
The goal of introducing a learning machine for Internetbased Cognitive Behavioral Therapy (ICBT) decision support will become an important part of an adaptive treatment strategy that has already been tested for one of the treatments [11], but could be further improved

Summary

Introduction

Machine learning is here employed to help answer questions concerning prediction of outcome and engagement in psychological treatment. Each machine learning method employed helps identify and amplify signals of bias, but due to the different nature of data points (e.g., standardized questionnaires, long texts, logs of system use) no machine learning method can be used to analyze every weak signal well enough to address the task of predicting future patient behavior and the clinical outcome of treatment. For this reason, a number of machine learning methods are used in tandem, with their signal analyses fused and unified to produce decision support for the clinician. Some of our initial experiments are described and discussed, before conclusions and some pointers to steps

Objectives

Methods

Findings

Discussion

Conclusion