Comparison of machine learning approaches for near-fall-detection with motion sensors

Abstract

Introduction: Falls are one of the most common causes of emergency hospital visits in older people. Early recognition of an increased fall risk, which can be indicated by the occurrence of near-falls, is important to initiate interventions.

Methods: In a study with 87 subjects we simulated near-fall events on a perturbation treadmill and recorded them with inertial measurement units (IMU) at seven different positions. We investigated different machine learning models for the near-fall detection including support vector machines, AdaBoost, convolutional neural networks, and bidirectional long short-term memory networks. Additionally, we analyzed the influence of the sensor position on the classification results.

Results: The best results showed a DeepConvLSTM with an F1 score of 0.954 (precision 0.969, recall 0.942) at the sensor position "left wrist."

Discussion: Since these results were obtained in the laboratory, the next step is to evaluate the suitability of the classifiers in the field.

Keywords: CNN; IMU; fall risk; machine learning; mobile health; near-fall; perturbation.

Figure 1

Perturbation treadmill (left) and IMU sensors and their position (right). The positions of the APDM sensors (green circle) are foot (left/right), wrist(left/right), sternum, and lumbar. The position of the activPAL sensor (blue box) is the right upper thigh.

Figure 2

Overview of the methods used to preprocess, train and classify the sensor datasets.

Figure 3

Creating a perturbation window around the maximum of the norm, which represents the maximum reaction to a perturbation. Green: Treadmill perturbation, blue: Maximum, red: Selected 2s perturbation window.

Figure 4

Confusion matrix of DeepConv2DLSTM model and optimal hyperparameters based on data of a multi-sensor system (left) and confusion matrix of the majority vote classification of a multi-sensor system (right).