NDDNet: a deep learning model for predicting neurodegenerative diseases from gait pattern
Date
2023-03-27Author
Faisal, Md Ahasan AtickChowdhury, Muhammad E.H.
Mahbub, Zaid Bin
Pedersen, Shona
Ahmed, Mosabber Uddin
Khandakar, Amith
Alhatou, Mohammed
Nabil, Mohammad
Ara, Iffat
Bhuiyan, Enamul Haque
Mahmud, Sakib
AbdulMoniem, Mohammed
...show more authors ...show less authors
Metadata
Show full item recordAbstract
Neurodegenerative diseases damage neuromuscular tissues and deteriorate motor neurons which affects the motor capacity of the patient. Particularly the walking gait is greatly influenced by the deterioration process. Early detection of anomalous gait patterns caused by neurodegenerative diseases can help the patient to prevent associated risks. Previous studies in this domain relied on either features extracted from gait parameters or the Ground Reaction Force (GRF) signal. In this work, we aim to combine both GRF signals and extracted features to provide a better analysis of walking gait patterns. For this, we designed NDDNet, a novel neural network architecture to process both of these data simultaneously to detect 3 different Neurodegenerative Diseases (NDDs). We have done several experiments on the data collected from 64 participants and got 96.75% accuracy on average in detecting 3 types of NDDs. The proposed method might provide a way to get the most out of the data in hand while working with GRF signals and help diagnose patients with an anomalous gait more effectively.
Collections
- Electrical Engineering [2579 items ]
- Medicine Research [1379 items ]