A study in China validated a urine-based tuberculosis adherence test that uses patient-submitted smartphone images to check medication intake. Across 1,655 submitted photos, 72% were interpretable, with issues centered on image blur, cartridge artifacts, and unusual strip colors that are described as fixable. Operationally, the approach aims to reduce clinic burden and improve monitoring reach for adherence—often a bottleneck in TB control. While the interpretability rate still leaves room for improvement, the results provide a pathway for scaling mHealth adherence verification with standardized photo-quality guidance and device reliability checks.