Single-label and multi-label classification for disease recognition with special consideration of comorbidities

Certain diseases require rapid treatment to avoid long-term consequences for patients. However, they may be difficult to recognize, especially if the symptoms are ambiguous and compatible with multiple possible diagnoses. Completing all necessary examinations often takes time, thereby prolonging patient suffering. Data-driven approaches, such as single-label classification (SLC) and multi-label classification (MLC), can help accelerate the diagnostic process and improve accuracy.

Comparing SLC and MLC allows us to investigate whether disease recognition benefits from considering comorbidities. In this context, we aim to provide a conceptual framing of (differences between) the two approaches in model formulation, decision spaces and handling of class imbalance. To empirically assess their performance, we conduct a case study applying SLC and MLC to data from chronic pain patients.

Our analysis yields an ambiguous picture of whether incorporating comorbidities improves disease recognition. The suitability of SLC and MLC is determined by multiple factors, notably the dependency structure among diseases and between diseases and covariates, as well as by data characteristics such as class imbalance. Consequently, it is of high importance to address each individual problem in an individual manner, especially when the implications can be significant, such as in patient referrals.

Our study highlights the importance of considering the specific characteristics of the data when selecting an appropriate classification approach for disease recognition and beyond. We illustrate how underlying assumptions, learning structures, and strategies for handling class imbalance shape predictive performance. Our results thus lead to the point that computational prediction tools should serve as decision support for treating physicians, but never replace their clinical judgment.