Unlocking amyotrophic lateral sclerosis diagnosis: How artificial intelligence is transforming early prediction

Amyotrophic lateral sclerosis (ALS) is one of the most disabling neurodegenerative conditions, characterized by progressive motor neuron degeneration that leads to paralysis and death, typically 3–5 years following onset. Despite decades of research, early diagnosis of ALS remains a major challenge. Symptoms often present subtly and overlap with other neuromuscular disorders, causing diagnostic delays averaging 10–16 months (Yuan et al., 2024). This lag further narrows the already limited therapeutic window and reduces the effectiveness of emerging disease-modifying treatments. While both in vitro and in vivo models have been valuable in elucidating ALS pathophysiology (Moreno-Jimenez et al., 2024), helping to define key etiological mechanisms such as mitochondrial impairment, protein misfolding, glutamate excitotoxicity, and neuroinflammation, the translation of these mechanistic insights into clinical diagnostics has proven remarkably difficult. The phenotypic and genetic complexity, and heterogeneity of ALS, still requires innovative strategies capable of processing and integrating high-dimensional and multi-layered large-scale datasets (Bono et al., 2025). This is where artificial intelligence (AI) is beginning to make a difference. In this perspective, we examine how AI, including machine learning (ML)–based approaches, is poised to transform ALS research, from preclinical models to human applications, with a particular emphasis on early diagnosis and disease prediction. We discuss the potential of AI-based advanced computational tools in extracting, analyzing, and interpreting biological data, including transcriptomic information, to improve early diagnosis and expand the intervention window for treatment of this disease. We also reflect on the ethical implications and translational challenges of applying such technologies in clinical practice.