ANTRAM Join early access

AGNI EVIDENCE LIBRARY / REVIEWED 31 AUGUST 2026

The science we use.
The limits we keep.

ANTRAM is investigating whether repeated abdominal acoustics plus lightweight daily context can create a more useful personal record. These sources establish plausibility and research methods. They do not validate AGNI.

Abstract visualization of investigational abdominal acoustic signals

HOW WE READ EVIDENCE

Plausibility is a starting point, not a product claim.

  1. Independent evidenceUse peer-reviewed research and official guidance where available.
  2. Device-specific proofNever transfer another sensor's performance to AGNI.
  3. Prospective validationPredefine endpoints, reference methods, and limitations before testing.

SELECTED PAPERS & CREDITS

Research informing the roadmap

Citations use a compact APA-style format. Links lead to PubMed, the publisher, or an official source.

01 / SYMPTOM DIARIES

Wright-McNaughton, M., et al. (2019). Measuring diet intake and gastrointestinal symptoms in irritable bowel syndrome: Validation of the Food and Symptom Times diary.

Clinical and Translational Gastroenterology, 10(12), e00103. https://doi.org/10.14309/ctg.0000000000000103

Informs: contemporaneous meal and symptom recording. Does not prove: AGNI's wording, hardware, or algorithms.
View on PubMed
02 / RECALL QUALITY

Lackner, J. M., et al. (2014). The accuracy of patient-reported measures for GI symptoms: A comparison of real time and retrospective reports.

Neurogastroenterology & Motility, 26(12), 1802-1811.

Informs: the need to reduce reliance on retrospective memory. Does not prove: passive sensing solves recall bias.
View on PubMed
03 / ACOUSTIC EVIDENCE

Huang, et al. (2024). Diagnostic accuracy of computerized bowel sound analysis for irritable bowel syndrome: A systematic review and meta-analysis.

Chinese Medical Sciences Journal. PMID 38594814.

Informs: acoustic analysis is scientifically active. Limitation: only 4 studies met inclusion criteria; validity and applicability remain uncertain.
View on PubMed
04 / SENSOR QUALIFICATION

Mansour, T., et al. (2024). SonicGuard Sensor: A multichannel acoustic sensor for long-term monitoring of abdominal sounds examined through a qualification study.

Sensors, 24(6), 1843. https://doi.org/10.3390/s24061843

Informs: bench, phantom, comparator, and small human acquisition methods. Does not prove: AGNI signal quality or clinical utility.
View on PubMed
05 / FIELD REVIEW

Inderjeeth, A.-J., et al. (2018). The potential of computerised analysis of bowel sounds for diagnosis of gastrointestinal conditions: A systematic review.

Systematic Reviews. PMID 30115115.

Informs: the range of reported bowel-sound features and use cases. Limitation: heterogeneous acquisition and analysis methods.
View on PubMed
06 / STOOL FORM

Chumpitazi, B. P., et al. (2016). Bristol Stool Form Scale reliability and agreement decreases when determining Rome III stool form designations.

Neurogastroenterology & Motility. PMID 26690980.

Informs: stool form as a useful descriptive domain. Limitation: disagreement can occur around category boundaries.
View on PubMed
07 / WEARABLE MATERIALS

Dang, T. B., et al. (2025). Flexible, wearable mechano-acoustic sensors for body sound monitoring applications.

Nanoscale, 17(16), 9652-9685. https://doi.org/10.1039/D4NR05145A

Informs: flexible sensor architecture and body-sound monitoring opportunities. Does not prove: an abdominal wellness outcome.
View on PubMed
08 / CLAIM BOUNDARY

U.S. Food and Drug Administration. (2026). General Wellness: Policy for Low Risk Devices.

Guidance document issued January 2026.

Informs: disciplined separation of wellness information from diagnosis or treatment claims. Limitation: not a product-specific determination and not Indian legal advice.
View FDA guidance

WHAT AGNI MUST PROVE

Capture. Repeat. Compare. Validate.

01Bench & tissue-path tests02Artefact & repeatability03Human feasibility & comfort04Prospective usefulness

Any future sensor-derived metric or algorithm requires its own protocol, reference standard, predefined endpoint, error analysis, and external validation.