An experimental protocol for video recording with a smartphone for use in a pre-trained neural network

Parkinson's disease is a neurodegenerative disease that has, as one of its main symptoms, a change in the response of the individual's muscle activity. It is the second most common neurodegenerative disease, the first being Alzheimer's disease. Due to motor alterations being one of its main symptoms, the proposed protocol is intended to capture movements from a smartphone camera for further analysis in object recognition software (DeepLabCut) and finally extract important data that can help in the detection of the disease or its stage classification. The protocol was designed to be simple and of easy replicability; to do so, a simple test was used for the volunteer to exercise (drawing of spiral formats and sinusoidal waves), and to record it, a support was developed for the smartphone and for the sheet, thus avoiding many measurements and bad positioning according to the replication. A neural network pre-trained by the authors has been used in the recognition software, from which data will be extracted when the protocol is applied. This is an experimental protocol that has not yet undergone a pilot test, so points of improvement can be found. However, it's a protocol that can enable quick and practical tests for Parkinson's disease diagnosis and quantification.