Evidence status

MyZenCheck Validation Methodology and Results

No public diagnostic-accuracy claim

MyZenCheck has not published an independently reviewed clinical-validation protocol or sufficient public data to support a diagnostic-accuracy claim.

What is currently established

  • The deployed code orchestrates multiple computer-vision components and a text-generation service.
  • The user flow includes an initial image-quality check.
  • The output is framed as educational TCM wellness information, not a medical diagnosis.
  • External researchers have studied tongue-image standardization, feature recognition, and practitioner reliability.

What is not publicly established

  • diagnostic accuracy, clinical utility, sensitivity, specificity, or patient outcomes;
  • the provenance, consent status, deduplication, demographic composition, or label quality of a model-development dataset;
  • an independent test cohort, preregistered protocol, evaluator independence, inter-rater analysis, confidence intervals, or confusion matrices;
  • that a model confidence score represents probability of a correct health conclusion.

Publication gate for future metrics

  1. Freeze and identify the exact model and dataset version.
  2. Define each task, label, denominator, exclusion, and primary endpoint before evaluation.
  3. Describe consent, provenance, deduplication, subgroup composition, and missing data.
  4. Use an independent held-out cohort and identify qualified evaluators and conflicts.
  5. Publish per-class metrics, uncertainty, calibration, confusion matrices, failure examples, and subgroup performance.
  6. Separate component image-classification metrics from practitioner agreement and from clinical outcomes.
  7. Record limitations, adverse-use risks, revision history, and a downloadable protocol appendix.

External research—not MyZenCheck validation

The references below provide context about research methods and reliability. Their findings cannot be transferred to this product without direct evaluation.

. The classification of tongue colors with standardized acquisition and ICC profile correction in Traditional Chinese Medicine. BioMed Research International. ;2016. . Tooth-Marked Tongue Recognition Using Multiple Instance Learning and CNN Features. IEEE Transactions on Cybernetics. ;49(2):380-387. . Traditional Chinese Medicine Tongue Inspection: An Examination of the Inter- and Intrapractitioner Reliability for Specific Tongue Characteristics. Journal of Alternative and Complementary Medicine. ;14(5):527-536.

Practical interpretation

Use MyZenCheck only as an optional educational description of visible image characteristics and TCM terminology. Do not use it to delay care, select treatment, stop medication, or rule out disease.

Read the user-facing limitations guide