Clinical & Research Disclaimer
Critical regulatory disclosures regarding the investigational status of quantum variational classifiers and statistical benchmarking in clinical medicine.
QuantumX is an exploratory computational platform developed for academic research, algorithm validation, and comparative benchmarking between classical machine learning ensembles and parameterized quantum circuits. It is not an FDA-cleared diagnostic device.
1. No Medical Advice or Treatment Recommendations
The risk probabilities, classification predictions (e.g., Benign vs. Malignant), and gate ablation saliency maps generated by QuantumX do not constitute formal medical diagnoses, clinical prognoses, or personalized therapeutic prescriptions. Healthcare providers must evaluate patient conditions using established standard-of-care clinical tests, biopsies, and radiographic imaging.
2. Probabilistic & Noisy Intermediate-Scale Quantum (NISQ) Realities
Quantum machine learning algorithms executed on current NISQ physical processors are inherently susceptible to environmental decoherence, thermal relaxation and two-qubit gate crosstalk. Although QuantumX incorporates Zero-Noise Extrapolation (ZNE) and M3 readout error mitigation:
- Hardware shot variance may cause minor fluctuations in expectation values.
- All comparative claims are rigorously subjected to McNemar's test and Wilcoxon signed-rank tests with p < 0.05 thresholding to prevent premature claims of quantum advantage.
3. Clinician-in-the-Loop Requirement
Any translational research finding or biomarker attribution derived from QuantumX's QXplain engine must be independently reviewed by a qualified medical oncologist, pathologist, or geneticist prior to publication or clinical trial consideration.
All QuantumX diagnostic executions produce immutable, SHA-256 hashed provenance receipts linking the exact input tensor, VQC state, and gate ablation metrics.