Electronic nose to distinguish bladder cancer by urinary odour feature: A pilot study.

摘要: BACKGROUND No study has yet investigated the use of electronic nose (eNose) technology to reveal pattern recognition urological diseases, including bladder cancer. OBJECTIVE We sought determine diagnostic performance eNose in recognizing urinary odour patients with METHODS The is a commercially available model equipped two sensors. angle sensors (θ) depends on kinds chemical substances, thus defining θ as feature odour. Quantity number detected during measurement. Urine samples were from 36 untreated cancer, 29 urolithiasis, 10 tract infection (UTI), and 27 healthy volunteers. RESULTS Based ROC analysis quantity an optimal cut-off value for 49, 48, 55 was applied compare volunteer, urolithiasis UTI groups, respectively. There significantly differences between cancer other conditions using these specific points (p< 0.0001, respectively). resulting sensitivity 61.4%, 45.6%, 60.8%, specificity 52.8%, 68.4%, 90.2%, AUC 0.565, 0.548, 0.909, CONCLUSION small, portable, rapid, low cost, noninvasive instrument distinguishing benign conditions.