AI Shows Promise in Early Detection of Cervical Spondylotic Myelopathy

Cervical spondylotic myelopathy (CSM), a leading cause of spinal cord dysfunction in older adults, is often diagnosed late due to subtle initial symptoms. However, fresh research suggests artificial intelligence (AI) could significantly improve early detection, potentially leading to more effective treatment. A study published in npj Digital Medicine details an AI-based approach that may allow clinicians to identify CSM up to 30 months before a traditional clinical diagnosis.

Understanding Cervical Spondylotic Myelopathy

CSM arises from arthritis in the neck, causing compression of the spinal cord [1]. Symptoms can include neck pain, muscle weakness, and difficulty with walking. Delayed diagnosis is a significant concern, as earlier intervention often leads to better outcomes.

The AI-Powered Approach

Researchers at Washington University in St. Louis developed and tested seven different AI models using electronic health record (EHR) data from over 2 million individuals [1]. These models analyzed patterns in healthcare interactions – tests, diagnoses, and other medical history data – to identify patients at higher risk of developing CSM.

Foundation Models vs. Clinically Guided Models

The team compared two main types of AI models: large “foundation models” pretrained on vast clinical datasets, and smaller, specialized models built with specific clinical knowledge. Although the foundation models performed well in initial testing, the clinically guided models demonstrated better generalizability and consistent performance when tested across different healthcare systems [1].

“We were able to achieve at least comparable, if not superior, performance with a much, much simpler model by focusing on existing clinical knowledge while still using a deep learning model,” said Jacob Greenberg, MD, assistant professor of neurosurgery at WashU Medicine [1].

The Importance of Clinical Insight

The findings highlight the continued importance of clinical expertise in the age of AI. Chenyang Lu, director of the AI for Health Institute at WashU, emphasized that embedding clinical insight into AI solutions is crucial for developing robust and trustworthy tools [1]. A key challenge for AI in medicine is generalizability – a model that works well in one hospital may not perform as effectively in another.

Future Implications

This research suggests that AI has the potential to transform the diagnosis and treatment of CSM. By identifying at-risk patients earlier, clinicians may be able to intervene sooner, potentially slowing disease progression and improving quality of life. Further research and validation are needed to refine these models and integrate them into clinical practice.

Key Takeaways

• AI models can analyze electronic health records to predict the risk of developing cervical spondylotic myelopathy (CSM).
• Clinically guided AI models, incorporating existing medical knowledge, may generalize better across different healthcare systems than large foundation models.
• Early detection of CSM is crucial for improving treatment outcomes.
• Clinical expertise remains essential in the development and implementation of AI-based healthcare solutions.

Source: Yakdan, S., et al. (2026). Clinically-guided models or foundation models? predicting cervical spondylotic myelopathy from electronic health records. npj Digital Medicine.