Mild Cognitive Impairment Detection Using Machine Learning Models Trained on Data Collected from Serious Games

Abstract

Mild cognitive impairment (MCI) is an indicative precursor of Alzheimer’s disease and its early detection is critical to restrain further cognitive deterioration through preventive measures. In this context, the capacity of serious games combined with machine learning for MCI detection is examined. In particular, a custom methodology is proposed, which consists of a series of steps to train and evaluate classification models that could discriminate healthy from cognitive impaired individuals on the basis of game performance and other subjective data. Such data were collected during a pilot evaluation study of a gaming platform, called COGNIPLAT, with 10 seniors. An exploratory analysis of the data is performed to assess feature selection, model overfitting, optimization techniques and classification performance using several machine learning algorithms and standard evaluation metrics. A production level model is also trained to deal with the issue of data leakage while delivering a high detection performance (92.14% accuracy, 93.4% sensitivity and 90% specificity) based on the Gaussian Naive Bayes classifier. This preliminary study provides initial evidence that serious games combined with machine learning methods could potentially serve as a complementary or an alternative tool to the traditional cognitive screening processes.

Visualization and Interaction Technologies in Serious and Exergames for Cognitive Assessment and Training: A Survey on Available Solutions and Their Validation

“…In the first case, the user is required to handle a device and apply a physical pressure to a surface to trigger events. Whereas, in the second case, [2], [4], [5], [7], [10], [11], [18], [22], [24], [25], [28], [30], [34], [45], [51], [53], [60]- [121] 79 Mobile devices [13], [16], [19], [26], [31], [36], [37], [39], [41], [43], [44], [59],…”

“…As participants could suffer from visual and hearing problems or have troubles with atten- tion, visual scenes have to be as simple as possible and search space should be reduced. During gameplay, simple and immediate data as a countdown timer, the number of remained chances to guess an answer, the current score or achieved goals are often shown to incentive the subject to answer as quickly and accurately as possible [3], [14], [17], [19], [20], [22], [24], [26], [28]- [31], [33], [38]- [41], [44], [49], [50], [52], [53], [71], [72], [79], [83], [96], [97], [99], [100], [106], [109], [115], [120], [124], [125], [130]- [132], [137], [138], [142], [150], [161], [164], [179], [191], [209], [211], [212], [231],…”

“…Touchscreen surface [3]- [5], [7], [13], [16], [19], [21], [26], [31]- [33], [36], [37], [39], [41], [43], [44], [48], [50], [58], [59], [85], [118], [121]- [127], [129]- [145], [160]- [168], [199], [225], [232] 60…”

“…[77] EEG [2], [53], [75], [93], [110] Treadmill/pressure plate/cycle ergometer [22], [88], [94] [13], [19], [26], [31], [36], [37], [39], [41], [43], [44], [122]- [124], [126], [127], [3], [21], [32], [33], [48], [144], [160], [162]- [168] Touchful and body tracking [38], [40] VR Semi-Immersive Driving simulator Driving tool [170], [171]…”

“…Access to a higher level is often bound to the achievement of a certain score, accuracy or number of consecutive correct trials or to the acquirement of certain motor and cognitive abilities [2]- [24] and, sometimes, inadequate performance can lead to a negative level adjustment [25]- [42]. Alternative solutions include the use of AI techniques, as case-based algorithms [43], ontology based models [44], or open learner models [45], or giving the therapists [46]- [48] and the patients [17], [49], [50] the permission to intervene in the difficulty level selection. Otherwise, difficulty adjustments can be done real-time, during gameplay, by modifying game parameters according to players performance [51]- [53].…”