Multiclass Diagnostic Improvement of Interstitial Cystitis (IC) and Overactive Bladder (OAB) via Machine Learning Pipeline on Surface-Enhanced Raman Spectroscopy (SERS).

Interstitial cystitis (IC) and overactive bladder (OAB) are chronic pelvic conditions with shared symptoms of urinary urgency, increased frequency, and nocturia with pain in the bladder. IC is characterized by bladder pain and inflammation without a clear etiology, whereas OAB involves detrusor overactivity in the absence of infection. Despite this symptomatic overlap, IC and OAB differ in underlying pathophysiology and require distinct treatment and medical care. A specific diagnostic strategy is currently lacking and highlights the need of a robust diagnostic method for effective treatment. Raman spectroscopy provides sensitive and nondestructive molecular fingerprints; when combined with artificial intelligence (AI)-driven analysis of spectral data, it offers a promising approach to overcoming the limitations of exclusion-based diagnoses. In this study, urinary samples were collected from healthy individuals (n = 117), IC patients (n = 19), and OAB patients (n = 45) for the acquisition of Raman spectra using Au-ZnO nanorod surface-enhanced Raman spectroscopy (SERS) chips. To obtain biological interpretable Raman spectra per group and high accuracy of classification performance, the linear models PCA-PLS-DA and PCA-LDA and tree-based nonlinear models XGBoost and LightGBM were applied and reached up to 92% accuracy with interpretable and pathologic relative Raman features for discriminating among healthy control, IC, and OAB. This approach suggests that urine-based SERS Raman spectra combined with machine learning could serve as a promising diagnostic platform to support disease-specific clinical decision.