Amyotrophic Lateral Sclerosis (ALS) is a neurological disorder that affects motor neurons, the nerve cells in the brain and spinal cord that control voluntary muscle movement and breathing accoridng to the National Institute of Neurological Disorders and Stroke.
On this disease the motor neurons degenerate and die, which means they stop sending messages to the muscles. The muscles get weaker until they atrophy. Patients with Amyotrophic Lateral Sclerosis lose their ability to control voluntary movements such as chewing, walking, talking, breathing and others, progressive over time.
It is a fatal neurodegenerative disease, there is no cure yet, however there are some drugs to prolong the survival of the patients managing the symptoms.
The article "Classification of ALS patients based on acoustic analysis of sustained vowel phonations" proposes an approach to voice assessment for automatic system that separates people with and without ALS; they focus on analysing of sustain phonation of vowels /a/ and /i/ to classify those patients. On thei studies they used the LASSO (Least Absolute Shrinkage and Selection Operator) algorithm obtaining 99.7% accuracy with 99.3% sensitivity and 99.9% specificity.
This project analyses the same data made available by the publishers of the article, by using instead a Multi-layer Perceptron classifier algorithm, and compare the results obtained.
Kaggle: https://www.kaggle.com/datasets/daniilkrasnoproshin/amyotrophic-lateral-sclerosis-als/data
ARXIV: https://arxiv.org/abs/2012.07347
ScienceDirect: Classification of ALS patients based on acoustic analysis of sustained vowel phonations. https://www.sciencedirect.com/science/article/abs/pii/S1746809420304614
ScienceDirect: Least Absolute Shrinkage and Selection Operator. https://www.sciencedirect.com/topics/engineering/least-absolute-shrinkage-and-selection-operator
National Institute of Neurological Disorders and Stroke: Amyotrophic Lateral Sclerosis (ALS). https://www.ninds.nih.gov/health-information/disorders/amyotrophic-lateral-sclerosis-als