An algorithm that analyzes the writing of neurological patients is able to provide information on their health status with remote monitoring, and can replace the outpatient assessment. This is what promises a study conducted by an interdisciplinary research team, coordinated by Antonio Suppa of the Department of Human Neuroscience of the Sapienza University of Rome, which has proposed an innovative telemedicine system based on the analysis of writing through machine learning algorithms. The results of the work were published in the journal ‘Frontiers in Aging Neuroscience’ and the research was carried out with the collaboration of the Departments of Information Engineering, Electronics and Telecommunications of Sapienza, of the Irccs Neuromed of Pozzilli (Isernia) and of the of neurology of the University of Cincinnati in Ohio.
Handwriting, the researchers explain, “is an acquired cognitive and motor task of particular complexity, which offers an interesting window of observation on the functions of the brain. For this reason, monitoring it offers useful biological information, especially in neurological patients: disorders of writing are in fact frequently observed in people with neurodegenerative diseases, including Parkinson’s disease (micrograph) and Alzheimer’s disease (agraphy) “. The new monitoring system is based on the accuracy of machine learning algorithms in detecting some writing ‘patterns’ attributable to the physiological aging of healthy subjects, and is described as an alternative to the usual outpatient clinical evaluation.
The researchers recruited 156 healthy and right-handed people and divided them into three age groups: 51 young people between 18 and 32, 40 adults between the ages of 37 and 57, and 63 people in advanced adulthood, i.e. 62 and 90 years. Each was asked to write their name and surname with a black ballpoint pen 10 times on a sheet of white paper and, subsequently, to photograph their writing sample with a smartphone and send it to the researchers. “The main scientific goal of our study – underlines Suppa – consists in the accuracy of the automatic analysis of the handwriting with artificial intelligence algorithms, able to objectify the progressive reduction in the amplitude of the characters due to physiological aging and, therefore, to attribute any sample of writing to a specific age range of the author “.
“Although previous research had already shown changes in writing dexterity linked to increasing age, approaches based on more complex analytical techniques such as machine learning were required to analyze a large amount of data in the field of telemedicine,” Suppa specifies. “The analysis of writing with artificial intelligence algorithms – adds Simone Scardapane, co-author of the study – was carried out thanks to the use of a convolutional neural network, that is an artificial network specialized for the processing of images and digital signals, able to automatically convert characters into parameters of interest “.
It is a simple, ecological, low cost and easy to use method in various fields. In fact, in addition to the significant implications in the neurological field, it can contribute, for example, to the historical dating of a specific document, thanks to the automatic assessment of the age of the person who wrote it. In particular, in the medical-legal field it could facilitate the dating of a will at the time of drafting or signing.
“Our hope – concludes Francesco Asci, co-author of the study – is that the analysis of writing remotely and using artificial intelligence algorithms may constitute an innovative aging biomarker in the future, with a significant impact in the field of disease diagnostics. neurodegenerative and in accordance with the methods of telemedicine “.
#Patient #writing #reveals #health #status #algorithm