The Prediction of Recurrence Lumbar Disc Herniation At L5-S1 Level Through Machine Learning Models Based on Endoscopic Discectomy Via the Interlaminar Approach

Summary

What Was the Study About? This study focused on improving the care of patients with a specific type of back problem called lumbar disc herniation at the L5-S1 level. Doctors often treat this condition with a minimally invasive surgery known as percutaneous endoscopic interlaminar discectomy (PEID). However, sometimes the herniation (the damaged disc) can come back after surgery. The goal of this study was to develop computer models that help predict which patients might experience a recurrence of their herniated disc.

Who Participated? The study reviewed the medical records of 309 patients who had undergone the PEID surgery. Out of these, 33 patients experienced a recurrence of their herniation, while 276 patients did not.

What Did the Researchers Do?

Data Collection:

They gathered information from each patient before the surgery, including clinical details (like body weight and any health conditions such as diabetes) and imaging studies (like X-rays, CT scans, or MRIs) that show the condition of the spine.

Identifying Key Risk Factors:

Using a statistical method called LASSO regression, the researchers identified eight important factors that could influence whether the herniation might come back. These included factors such as body mass index (BMI), a measure related to disc height (posterior disc height index), signs of spinal canal narrowing, how long the patient had symptoms before surgery, and other health conditions.

Developing Prediction Models:

They then used several machine learning techniques (advanced computer methods that learn from data) to build prediction models. Two of the best-performing models were based on methods called Random Forest and Extreme Gradient Boosting (XGB).

What Were the Main Findings?

Key Predictors: Higher BMI and changes in the disc (as measured by the posterior disc height index) were found to be the strongest predictors of a herniation coming back after surgery. Other factors, like spinal canal narrowing and longer duration of symptoms before surgery, also played significant roles.

Practical Implication: These models can help doctors identify which patients are at higher risk for recurrence. With this information, they can adjust treatment plans and follow-up care to better manage and potentially reduce the risk of the herniation coming back.

Why Is This Important? For patients and their families, this study offers hope for more personalized and effective treatment plans, reducing the chances of needing additional surgeries in the future. For healthcare providers, the findings provide useful tools to improve decision-making before surgery, ensuring better long-term outcomes for patients with L5-S1 lumbar disc herniation.

In summary, this research uses modern computer methods to predict the risk of recurrent disc herniation after a common minimally invasive back surgery, aiming to enhance patient care and improve surgical outcomes.

Conditions

• Recurrent Lumbar Disc Herniation

Interventions

DIAGNOSTIC_TEST

VAS Point and Imaging Examination

This intervention uses a machine learning model to predict the risk of recurrent lumbar disc herniation (rLDH) in patients who have had percutaneous endoscopic interlaminar discectomy (PEID) at the L5-S1 level. The model combines clinical data (e.g., BMI, disease duration, diabetes) and imaging metrics (e.g., posterior disc height index, spinal canal stenosis) to create a personalized risk score, unlike traditional methods that rely on clinical judgment or imaging alone. Key Features: Data-Driven Approach: Developed using data from 309 patients for real-world relevance. Advanced Variable Selection: Identifies eight key predictors using LASSO regression. Multiple Machine Learning Techniques: Uses algorithms like support vector machine, random forest, and extreme gradient boosting. Optimized for Clinical Decision-Making: Assists surgeons in personalizing treatment plans to reduce recurrence risk.

Sponsors & Collaborators

• Nantong First People's Hospital

  collaborator OTHER

• Jinyu Chen

  lead OTHER

Eligibility

Sex

ALL

Healthy Volunteers

No

Timeline & Regulatory

Start

2020-01-01

Primary Completion

2024-05-31

Completion

2024-11-01

Countries

• China

Study Locations