Diagnostic biomarkers in amyotrophic lateral sclerosis: Evaluation of TMS, PET and MRG findings

ABSTRACT Diagnostic Biomarkers in Amyotrophic Lateral Sclerosis: Evaluation of TMS, PET, and MRI Findings Introduction and Aim: Amyotrophic Lateral Sclerosis (ALS) is a progressive, fatal neurodegenerative disorder characterized by the degeneration of both upper and lower motor neurons. In the early stages of the disease, the absence of overt clinical signs of upper motor neuron (UMN) involvement and the limitations of current diagnostic criteria often lead to diagnostic delays. This study aimed to evaluate the diagnostic contributions of transcranial magnetic stimulation (TMS), magnetic resonance imaging (MRI), and positron emission tomography (PET) in ALS, and to explore the potential of multimodal biomarker integration. Methods: This prospective, cross-sectional, observational study included 74 patients diagnosed with definite ALS and 38 age- and sex-matched healthy controls, recruited between May 2024 and June 2025 at the Department of Neurology, Çukurova University Faculty of Medicine. ALS patients were stratified into early-stage (n=39) and middle-to-late-stage (n=35) groups based on the ALS Functional Rating Scale-Revised (ALSFRS-R) scores. TMS was applied bilaterally to the abductor pollicis brevis (APB) and first dorsal interosseous (FDI) muscles, and parameters including resting motor threshold (RMT), cortical and peripheral latencies, central motor conduction time (CMCT), motor evoked potential (MEP) amplitude, and short-interval intracortical inhibition (SICI) were recorded. From high-resolution MRI scans, a total of 555 radiomic features were extracted from five brain regions and significant variables were selected using the Boruta algorithm. A subset of patients underwent 18F-FDG PET imaging to evaluate motor cortex metabolism. All data were analyzed using LASSO regression and Random Forest algorithms. Results: Significant differences were observed in TMS parameters such as SICI, RMT, and CMCT in the ALS group. Among these, SICI at a 3 ms interstimulus interval demonstrated the highest diagnostic accuracy (AUC: 0.846). Radiomic analysis revealed notable changes in texture-based features, particularly increased entropy and reduced homogeneity. FDG-PET imaging identified hypometabolism in the motor cortex. The integrated model combining TMS and MRI data achieved high diagnostic performance in distinguishing early-stage ALS patients from healthy controls (AUC: 0.984). Conclusion: Both cortical excitability parameters assessed by TMS and radiomic MRI features offer promising biomarkers for early diagnosis and phenotypic differentiation in ALS. The integration of these modalities through machine learning algorithms provides objective diagnostic support, particularly in early-stage disease. Our study demonstrates that combining neurophysiological and imaging-based parameters presents a robust multimodal approach to ALS diagnosis. Keywords: Amyotrophic Lateral Sclerosis, Transcranial Magnetic Stimulation, SICI, Radiomic MRI, FDG-PET, Biomarkers, ALSFRS-R, Machine Learning