Reminiscence therapy is an approach used in the field of gerontology, the field of study of aging and its consequences. It is a way of integrating the past into the present by bringing back distant emotions and memories. This is one of the solutions proposed to people suffering from Alzheimer’s disease. Spanish researchers have developed an AI-based application to facilitate this kind of therapy.
A study on reminiscence therapy
In 2018, a study was conducted at Bangor University in Wales by Bob Woods, Laura O’Philbin, Emma Farrell, Aimee Spector and Martin Orrell. They published their findings on reminiscence therapy. They report that the effects of reminiscence interventions are inconsistent, often diminished, and can differ considerably across settings and modalities. However, the brain seems to be stimulated if the individual thinks about or evokes their own existence through experiences and memories they have already had, even if they do not remember them at first glance.
A research project on the implementation of reminiscence therapy was subsequently launched by Francisco Mugica, Anass Benali and Àngela Nebot, researchers in artificial intelligence at the IDEAI centre of the Polytechnic University of Catalonia, Natalia Albino Pires, an expert in philology and traditional Iberian cultures, and Dr. Sara Domenech, of the “Salut i Envelliment” Foundation of the Autonomous University of Barcelona.
At present, in order for a therapy of this type to work, three to four sessions per year must be carried out with the help of specialised personnel to prepare the necessary material and then carry out the follow-up of the therapeutic session. One of the difficulties lies in collecting the specific material for these therapies in order to use them properly in EPHAD for example. To address this difficulty, researchers have developed an AI-based solution to facilitate this process.
An application using artificial intelligence to propose adapted memories
The aim of the application is to analyze the emotions of elderly people participating in therapy to see if a positive reminiscence occurs when they look at a visual medium or listen to an audio track. The tool will use various memories from free cultural references or it will be possible to include personal memories related to the patient.
When the program is first used with a patient, the algorithm in the media is blank. As it interacts with the elderly person, it will analyze the memories and emotions that the person is giving off through the camera present on the device used. The artificial intelligence will be able to detect joy, sadness, indifference or anger and thus adapt the memories and cultural references according to the patient’s reactions.
Several personal data of the person are taken into account by the artificial intelligence in order to propose adapted memories: place of birth, place of residence during different periods of the patient’s life, date of birth, place of birth of parents, etc. A first test version of this software will be available during the summer and the researchers hope that a functional version will be available by the end of the year.
Translated from Alzheimer : Une application basée sur l’intelligence artificielle au service des thérapies par réminiscence
