Project

Predicting individual treatment response for injection therapy in low back pain

Short Summary

Low back pain is among the leading causes of disability worldwide. A major problem is the unsatisfactory treatment success rate, which is likely attributable to the one-size-fits-all treatment strategy. To date, it difficult to discern patients who respond to injection therapy to alleviate lower back pain from non-responders. In our project, we will use machine learning to analyze this multi-modal data with the aim to improve outcome prediction and optimize treatment selection for patients with low back pain. The findings of our project will inform clinical decision making in a data-driven way.

Goals

The multifaceted nature of low back pain demands holistic strategies that can tailor medical treatment to the functional-anatomical properties and clinical profiles of each individual patient. We propose to leverage multi-modal health data and harness the power of machine learning to predict individual treatment responses for injection therapy in low back pain patients. In contrast to humans, machine learning algorithms are naturally suited to create actionable decisions while having to integrate heterogeneous sources of data.

Significance

The fundamental importance of the proposed project stems from the fact that even in the era of “modern” medicine, low back pain still represents a major medical challenge and a global public health issue. It continues to be treated as a single disease entity despite its heterogeneous nature. Consequently, treatment selection is based upon trial and error, rendering the therapeutic process highly dissatisfying for patients and doctors alike. Moreover, with its tremendous impact on patients’ wellbeing and ability to work, low back pain makes a significant contribution to the economic burden of disease and rising health costs. Our study aims to systematically assess patients with multi-modal measures, incl. lumbar MRI and gait analysis, as a basis for machine learning. The goal is to predict from these structural and functional data whether a given patient will benefit from injection therapy. Our project will lay the foundation for personalized treatment selection for patients suffering from low back pain.

Background

Affecting 80–85% of people of all age groups over their lifetime, low back pain is among the leading causes of disability worldwide. Beyond their immediate health consequences, the societal and economic costs are of epidemic proportion and thus, make low back pain a global public health priority. Despite its highly heterogeneous nature, low back pain continues to be treated as a single disease entity. Treatment selection is based upon trial and error, rendering the therapeutic process highly dissatisfying for patients and doctors alike.

iDoc

Co-Investigators

Dr. Zina-Mary Manjaly
Prof. Ender Konukoglu

Consortium

Schulthess Clinic

Status

In Progress

7th Call, iDoc