Crowdsourcing Disease Biomarker Discovery Research: The IP4IC Study

J Urol. 2018 May;199(5):1344-1350. doi: 10.1016/j.juro.2017.09.167. Epub 2017 Dec 7.


Purpose: Biomarker discovery is limited by readily assessable, cost efficient human samples available in large numbers that represent the entire heterogeneity of the disease. We developed a novel, active participation crowdsourcing method to determine BP-RS (Bladder Permeability Defect Risk Score). It is based on noninvasive urinary cytokines to discriminate patients with interstitial cystitis/bladder pain syndrome who had Hunner lesions from controls and patients with interstitial cystitis/bladder pain syndrome but without Hunner lesions.

Materials and methods: We performed a national crowdsourcing study in cooperation with the Interstitial Cystitis Association. Patients answered demographic, symptom severity and urinary frequency questionnaires on a HIPAA (Health Insurance Portability and Accountability Act) compliant website. Urine samples were collected at home, stabilized with a preservative and sent to Beaumont Hospital for analysis. The expression of 3 urinary cytokines was used in a machine learning algorithm to develop BP-RS.

Results: The IP4IC study collected a total of 448 urine samples, representing 153 patients (147 females and 6 males) with interstitial cystitis/bladder pain syndrome, of whom 54 (50 females and 4 males) had Hunner lesions. A total of 159 female and 136 male controls also participated, who were age matched. A defined BP-RS was calculated to predict interstitial cystitis/bladder pain syndrome with Hunner lesions or a bladder permeability defect etiology with 89% validity.

Conclusions: In this novel participation crowdsourcing study we obtained a large number of urine samples from 46 states, which were collected at home, shipped and stored at room temperature. Using a machine learning algorithm we developed BP-RS to quantify the risk of interstitial cystitis/bladder pain syndrome with Hunner lesions, which is indicative of a bladder permeability defect etiology. To our knowledge BP-RS is the first validated urine biomarker assay for interstitial cystitis/bladder pain syndrome and one of the first biomarker assays to be developed using crowdsourcing.

Keywords: computational biology; crowdsourcing; cystitis; interstitial; machine learning; urinary bladder.

Publication types

  • Validation Study

MeSH terms

  • Biomarkers / urine
  • Biomedical Research / methods*
  • Case-Control Studies
  • Crowdsourcing / methods*
  • Cystitis, Interstitial / complications
  • Cystitis, Interstitial / diagnosis*
  • Cystitis, Interstitial / urine
  • Female
  • Humans
  • Machine Learning
  • Male
  • Middle Aged
  • Pelvic Pain / diagnosis*
  • Pelvic Pain / etiology
  • Pelvic Pain / urine
  • Postal Service
  • Severity of Illness Index
  • Social Media
  • Specimen Handling / methods*
  • Surveys and Questionnaires / statistics & numerical data


  • Biomarkers