Machine learning based digital assessment of mild cognitive impairment using mouse trajectories during the trail making test
4 affiliations
- 1Laboratorio de Inteligencia Artificial Aplicada, Instituto de Ciencias de la Computación, Facultad de Ciencias Exactas y Naturales, Universidad de Buenos Aires - CONICET
- 2Departamento de Neurología, Hospital Italiano de Buenos Aires
- 3Latin American Brain Health Institute (BrainLat), Universidad Adolfo Ibañez
- 4University of Gothenburg
Abstract
One of the objectives of digital neuropsychology is to apply computational methods to improve the accuracy of traditional assessments. The Trail Making Test (TMT) is one of the most popular neuropsychological tests for executive functions assessment. Participants have to connect a sequence of items in order without lifting the pencil. In Part A (TMT-A) the items are identified with Arabic numerals. In Part B (TMT-B) they are identified with numbers and capital letters, and participants have to connect them in alternate order (1-A-2-B-…). We implemented a computerized TMT (cTMT) that preserves its original structure and records high-resolution mouse trajectories. Seventy-four older adults (41 with mild cognitive impairment and 33 healthy controls) completed the cTMT and a standard diagnostic battery. We also developed NeuroTask, a Python library that extracts features from cursor time series, including reaction times, speed and acceleration metrics, trajectory deviations, and state-based measures. We compared demographic, digital, and non-digital models using nested cross-validation and non-parametric permutation tests. Demographic models provided only modest discrimination (AUC = 0.56). Digital mouse features improved performance (AUC = 0.67), and combining them with demographics reached an AUC of 0.70, which approached the performance of the neuropsychological battery used to define the diagnosis (AUC = 0.76). In complementary regression analyses with digital plus demographic features, we obtained significant predictions for four out of seven target scores: MMSE, Digit Symbol, TMT-A and TMT-B. These results indicate that fine-grained mouse-movement features from the cTMT provide useful information for classifying mild cognitive impairment and for predicting multiple neuropsychological scores.
Related
- Curve walking gait analysis using depth camera and machine learning for detecting cognitive impairment
- Explainable machine learning identifies predictors for postural instability and cognitive impairment in early Parkinsons Disease
- Association between PCSK9 and post-stroke cognitive impairment: a retrospective cohort study and predictive models establishment based on machine learning.
- Modelling the risk of conversion from mild cognitive impairment to Alzheimer's disease using machine learning.
- Identifying Mild Cognitive Impairment from Routinely Collected Physical, Functional and Psychosocial Measures in Community-Dwelling Older Adults: A Comparative Machine-Learning and Regression Modelling Study in the Northern Japanese ORANGE Registry.
BibTeX
@article{juantorena2026machine,
title = {{Machine learning based digital assessment of mild cognitive impairment using mouse trajectories during the trail making test}},
author = {Gustavo E. Juantorena and Gianluca Capelo and Betsabe D. Leon Vallejos and Agustin Ibáñez and Agustin Petroni and Waleska Berrios and María Cecilia Fernández and Juan E. Kamienkowski},
journal = {Scientific Reports},
year = {2026},
month = jul,
doi = {10.1038/s41598-026-62955-9},
url = {https://doi.org/10.1038/s41598-026-62955-9},
note = {CC BY},
}
RIS
TY - JOUR
TI - Machine learning based digital assessment of mild cognitive impairment using mouse trajectories during the trail making test
AU - Gustavo E. Juantorena
AU - Gianluca Capelo
AU - Betsabe D. Leon Vallejos
AU - Agustin Ibáñez
AU - Agustin Petroni
AU - Waleska Berrios
AU - María Cecilia Fernández
AU - Juan E. Kamienkowski
PY - 2026
DA - 2026/07/01
JO - Scientific Reports
DO - 10.1038/s41598-026-62955-9
UR - https://doi.org/10.1038/s41598-026-62955-9
AB - One of the objectives of digital neuropsychology is to apply computational methods to improve the accuracy of traditional assessments. The Trail Making Test (TMT) is one of the most popular neuropsychological tests for executive functions assessment. Participants have to connect a sequence of items in order without lifting the pencil. In Part A (TMT-A) the items are identified with Arabic numerals. In Part B (TMT-B) they are identified with numbers and capital letters, and participants have to connect them in alternate order (1-A-2-B-…). We implemented a computerized TMT (cTMT) that preserves its original structure and records high-resolution mouse trajectories. Seventy-four older adults (41 with mild cognitive impairment and 33 healthy controls) completed the cTMT and a standard diagnostic battery. We also developed NeuroTask, a Python library that extracts features from cursor time series, including reaction times, speed and acceleration metrics, trajectory deviations, and state-based measures. We compared demographic, digital, and non-digital models using nested cross-validation and non-parametric permutation tests. Demographic models provided only modest discrimination (AUC = 0.56). Digital mouse features improved performance (AUC = 0.67), and combining them with demographics reached an AUC of 0.70, which approached the performance of the neuropsychological battery used to define the diagnosis (AUC = 0.76). In complementary regression analyses with digital plus demographic features, we obtained significant predictions for four out of seven target scores: MMSE, Digit Symbol, TMT-A and TMT-B. These results indicate that fine-grained mouse-movement features from the cTMT provide useful information for classifying mild cognitive impairment and for predicting multiple neuropsychological scores.
SN - 2045-2322
ER -