A neuromuscular clinician's primer on machine learning

Area of Science:

• Clinical Neurology
• Medical Informatics
• Artificial Intelligence in Medicine

Background:

• Artificial intelligence (AI) and machine learning (ML) are increasingly integrated into medical management and research.
• Generative AI, like ChatGPT3, has heightened public awareness of AI's potential for automating tasks.
• Neurology risks lagging behind other specialties in adopting AI/ML technologies for clinical practice.

Purpose of the Study:

• To provide a practical primer on machine learning (ML) fundamentals for clinicians.
• To educate neurologists on ML applications in neuromuscular and electrodiagnostic medicine.
• To address limitations, ethical concerns, and future directions of AI/ML in neurology.

Main Methods:

• Review of current AI/ML applications in neuromuscular disease diagnosis, monitoring, prognosis, and treatment.
• Discussion of ML in specific diagnostic modalities: nerve and muscle ultrasound, MRI, electrical impedance myography, nerve conduction studies, and electromyography.
• Exploration of AI/ML in clinical cohort studies.

Main Results:

• AI/ML applications are assisting in various aspects of patient care for neuromuscular diseases.
• These applications are currently largely confined to the research domain.
• Neurologists need to understand these technologies to avoid falling behind other medical specialties.

Conclusions:

• AI/ML will inevitably change clinical practice in neurology; the focus should be on *when* and *how*.
• The successful integration of AI/ML will be measured by improvements in patient outcomes.
• Understanding the basics, applications, and challenges of AI/ML is crucial for future neurologists.