A New Artificial Intelligence-Based Model for Amyotrophic Lateral Sclerosis Prediction

Table 8

The conducted assessment results.

References	Deployed technology	Accuracy	F-score	Dice
[6]	Standard feedforward neural network (FFNN), a convolutional neural network (CNN), and a recurrent neural network (RNN)	N/A	N/A	N/A
[7]	Unsupervised uniform manifold approximation and projection (UMAP) modeling, semisupervised (neural network UMAP) modeling, and supervised (ensemble learning based on LightGBM) modeling	93.24%	N/A	N/A
[9]	Comorbidities and associated indicators using electronic medical records (EMRs)	83.7%	73.95%	N/A
[12]	Induced pluripotent stem cells (iPSCs)	N/A	N/A	N/A
[13]	A multilayer neural network	96.63%	N/A	N/A
[15]	Stacked autoencoders	88%	N/A	N/A
[18]	A neural network	68.8%	N/A	N/A
[26]	XGBoost (XGB) and SHAP	70.7%	N/A	N/A
The presented approach	New UNET architecture	85.21%	86.05%	85.88%