# Breath-Dx: eNose Disease Screening Benchmark With Realistic Sensor Drift

> Source: <https://explainx.ai/blog/breath-dx-enose-benchmark-sensor-drift-clinical-screening-2026?utm_source=rss&utm_medium=feed&utm_campaign=blog>
> Published: 2026-11-18 00:00:00+00:00

**Recommendation:** Split eNose and other sensor health data by calendar time instead of randomly, and log drift curves over time alongside aggregate AUC, since random CV overstated AUC by 0.05 to 0.12 across all ten baselines.
More in the playbook "Evals and testing for LLM apps": <https://wpnews.pro/playbooks/evals-testing>

**Breath-Dx** is a new **open benchmark** for **electronic-nose (eNose) disease screening** that forces models to survive **sensor drift**, not just **random cross-validation**. Authors released the **[S-OH dataset on Hugging Face](https://huggingface.co/datasets/ivanpodd/S-OH?ref=explainx)**, **[code on GitHub](https://github.com/ivanpodd/Breath-Dx?ref=explainx)**, and **[arXiv paper 2611.15348](https://arxiv.org/abs/2611.15348?ref=explainx)** so teams can reproduce **ten baseline families** on **1,234 patients**.

If you build **health ML**, the headline is simple: **every method they tested lost 0.05–0.12 ROC AUC** when evaluation respected **time**, exposing **lab scores that would fail in clinics**.

## TL;DR

| Item | Detail | 
|---|---|
| Data | **S-OH** :**1,234 patients** ,**9** diagnostic groups,**17-channel** series | 
| Methods | **10** models across tabular, CNN, GNN, and TS architectures | 
| Drift effect | **−0.05 to −0.12 AUC** vs random CV for**all** families | 
| Peak temporal AUC | **~0.74–0.76** on some**lung** conditions | 
| Release | **HF dataset** ,**GitHub** splits + eval protocols | 
| Award | Authors cite **IEEE ICDM 2026 Applied Track runner-up** | 

## Why eNose research needed a harder benchmark

**Electronic noses** promise **portable, non-invasive screening** from **exhaled breath**. Most papers use **small private cohorts** and **random splits**, which **hide drift**: sensors age, environments shift, and **batch effects** correlate with **time**.

Breath-Dx standardizes **temporal train/test partitions** on **public** data so **ROC AUC** reflects **deployment reality**, the same discipline **[InclusionMed](https://explainx.ai/blog/inclusionmed-healthcare-ai-benchmark-index-2026)** pushes for **healthcare AI indices**.

## What the S-OH dataset contains

**Scent of Health (S-OH)** is described in the paper as the **largest public clinical eNose collection** to date:

- **1,234 patients**
- **Nine diagnostic groups**
- **17-channel multivariate time series** per participant
- **Explicit temporal splits** accounting for**sensor drift**

Download via **[ivanpodd/S-OH](https://huggingface.co/datasets/ivanpodd/S-OH?ref=explainx)** and read the **dataset card** for **collection protocol** and **label definitions** before **fine-tuning frontier models**.

## Methods and results under temporal validation

Breath-Dx evaluates **four families**:

1. **Feature-based tabular learners**
2. **Convolutional models**
3. **Graph-based models**
4. **Time series architectures**

Under **temporal validation**:

- **All ten methods** lose**0.05–0.12 AUC** versus**random CV**
- **Recurrent models** show the**most consistent** performance across diseases
- **Specialized TS models** achieve**highest peaks** on some tasks
- **Best temporal AUC ~0.74–0.76** for**some lung conditions** ; others**near chance**

That spread matters for **product roadmaps**: **screening UX** must communicate **uncertainty per condition**, not a single **"AI accuracy"** banner.

Monitoring deployment drift when sensor baselines shift over calendar time

## Demographics and confounding

Authors emphasize **demographic confounders**:

- Breath signals add **independent value** for**some diseases**
- For **others** , signal adds**little beyond age**

Benchmark **protocols** in the GitHub repo document how to **report** stratified metrics. Copy that pattern when you ship **internal health agents** audited like **[GPT-6 healthcare pilots](https://explainx.ai/blog/openai-computer-use-healthcare-gpt-6-astra-cvs-november-2026)**—** subgroup tables** are not optional.

## How to run the benchmark

1. Clone **[github.com/ivanpodd/Breath-Dx](https://github.com/ivanpodd/Breath-Dx?ref=explainx)** .
2. Pull **[ivanpodd/S-OH](https://huggingface.co/datasets/ivanpodd/S-OH?ref=explainx)** with the**published temporal splits** .
3. Reproduce **baselines** before swapping in**custom transformers** .
4. Log **drift curves** over**calendar time** , not only**aggregate AUC** .

For **time-series craft**, cross-read **[multi-branch transformer KV work](https://explainx.ai/blog/multi-branch-transformer-joint-kv-mbt-uiuc-2026)** when you design **long-context** breath windows—different domain, same **leakage** pitfalls.

## Implications for AI-for-health builders

**Do not ship random-CV numbers** in investor decks if **temporal splits** collapse performance. Breath-Dx gives **citeable** evidence that **drift is universal** in their study, not an **edge case**.

**Regulatory path stays long.** Benchmark success **does not** clear **FDA/EMA** pathways; treat Breath-Dx as **R&D hygiene**, like **[Sakana JFinBench](https://explainx.ai/blog/sakana-jfinbench-japanese-financial-certification-benchmark-2026)** for **finance agents**.

**Open data helps audits.** Public splits enable **third-party replication**, reducing **vendor-only** claims—aligned with **[Meta Brain2Qwerty](https://explainx.ai/blog/meta-brain2qwerty-v2-non-invasive-brain-to-text-decoder-2026)** style **reproducibility** expectations.

## Conference recognition

In the [Hugging Face paper thread](https://huggingface.co/papers/2611.15348?ref=explainx), authors note **IEEE ICDM 2026 Best Paper Runner-Up (Applied Track)**. That signals **peer review interest**, not **clinical endorsement**.

## Limitations to state plainly

- **Nine groups** still**exclude** many diseases.
- **eNose hardware** varies;**S-OH** protocols may not transfer to**consumer gadgets** .
- **Temporal splits** punish**drift** but**do not simulate** full**prospective trials** .

## Comparison with other health ML evaluation habits

Many **Kaggle-style** medical competitions still publish **random patient splits**. Breath-Dx argues that practice is **especially toxic** for **eNose** because **sensor aging** correlates with **calendar time**. The **0.05–0.12 AUC** penalty is not a **rounding error**; for **screening** at **fixed sensitivity**, it can flip **cost-effectiveness** models used in **hospital procurement**.

If you previously cited **non-temporal AUC** in decks, rerun tables with **Breath-Dx splits** before **regulatory meetings**. Align messaging with **[InclusionMed](https://explainx.ai/blog/inclusionmed-healthcare-ai-benchmark-index-2026)** gaps on **demographic reporting**.

## Integration with modern time-series foundation models

Teams experimenting with **Chronos**, **Moirai**, or **custom transformers** should treat Breath-Dx as a **cheap sanity check** before **million-parameter** finetunes. **Pretraining on generic series** does **not** immunize models against **drift**; the paper shows **universal degradation** under **temporal validation**.

Start with **released recurrent baselines** to establish **floor** metrics, then swap in **your architecture** while holding **splits fixed**. Log **calibration curves**—** AUC alone** hides **overconfidence** when **prevalence** shifts.

## Data card fields to read before training

On **[ivanpodd/S-OH](https://huggingface.co/datasets/ivanpodd/S-OH?ref=explainx)**, verify **label definitions** per **diagnostic group**, **acquisition dates** tied to each **patient row**, **sensor firmware** revisions if documented, and **missing channel** handling rules.

Mis-reading **label granularity** (staging versus **binary screening**) is a common **silent bug** in **health RAG** pipelines, similar to [JFinBench](https://explainx.ai/blog/sakana-jfinbench-japanese-financial-certification-benchmark-2026)**credential-level** mistakes in **finance agents**.

## What clinicians will ask in pilot meetings

Prepare answers for: **(1)** what happens when the sensor is **recalibrated mid-study**—point to **temporal split** design; **(2)** whether breath signal is **independent of age**—cite **demographic confounder** analysis; **(3)** which diseases are **near chance** even in lab data—quote **per-condition AUC** bands from the paper.

## Open science citation hygiene

When you write about Breath-Dx, cite **arXiv 2611.15348**, the **GitHub repository**, and the **Hugging Face dataset card**—not aggregator rewrites. That supports **reproducibility** norms we use in **[Casey Handmer’s physics literature audit](https://explainx.ai/blog/casey-handmer-ai-physics-literature-audit-158k-papers-2026)** and other **open benchmark** coverage.

Source citation discipline for clinical AI benchmark posts that link datasets and code releases

## Model family details builders should reproduce

Breath-Dx spans **feature engineering pipelines** and **end-to-end neural models**. Tabular baselines extract **hand-crafted statistics** from **17-channel** traces—mean, slope, spectral bins—then feed **gradient boosted trees** or **logistic heads**. Convolutional baselines treat each breath sample as a **2D map** (time × channel). Graph models connect **sensor nodes** with **learned edges**, useful when **cross-channel correlation** shifts under drift.

Time-series transformers and **State-space** variants chase **long-range dependencies** across **entire breath profiles**. The paper’s point is not “** transformers always win**”; it is “** every family overfits random splits**.” When you port **Hy4-class multimodal models** or **small SLMs** to S-OH, keep the **temporal split** as your **primary metric**, with random CV labeled **diagnostic only**.

## Designing a drift monitoring stack for eNose products

Production eNose programs need **three layers** Breath-Dx implies but does not ship as software:

1. **Reference gas standards** on a schedule tied to**calendar time** , not sample count alone.
2. **Population shift dashboards** comparing**embedding centroids** week over week.
3. **Human override** when**AUC drops** below**pre-registered** thresholds.

Map layer 3 to **[Americans’ negative AI sentiment polls](https://explainx.ai/blog/americans-negative-ai-sentiment-five-polls-owid-2026)**: consumers tolerate **assistive screening** only when **false positives** are **explained** and **appealable**.

## Regulatory and ethics framing

**EU AI Act** high-risk framing may apply if breath models inform **triage** without clinician sign-off. Breath-Dx does **not** solve **classification** under law; it supplies **evidence** that **uncalibrated lab AUC** misleads.

Document **informed consent** for **biometric breath** collection, especially when **workplace wellness** programs also collect **camera or wearable** telemetry that employees may read as **surveillance**.

## Extending the benchmark responsibly

Authors invite **reproducible research**. If you contribute new models:

- Publish **seeded splits** unchanged from GitHub.
- Report **per-disease** tables, not one**macro AUC** .
- Disclose **hardware calibration** if you**re-record** supplemental patients.

Avoid **merging S-OH** with **private cohorts** without **domain adaptation** ablations—drift will reappear.

## Pairing Breath-Dx with agentic health workflows

If you wire **eNose scores** into **LLM triage agents**, treat model output as **features**, not **diagnoses**. Agents should surface **confidence**, **collection timestamp**, and **sensor batch ID** in every summary—metadata Breath-Dx implies but agents often drop. Read **[AWS responsible agent roadmaps](https://explainx.ai/blog/aws-responsible-ai-agentic-ai-phased-roadmap-bedrock-2026)** before automating **patient-facing** messages from breath classifiers.

Hospital IT should segment **research laptops** running Breath-Dx baselines from **EHR-connected** production VLANs, the same **blast-radius** discipline we recommend after **[Java4Shell](https://explainx.ai/blog/java4shell-gpt-6-cyber-cve-2026-114938-oracle-patch-november-2026)** patching sprints.

## Future cohort and hardware directions

Authors point toward **multi-site** sensors and **prospective drift correction** algorithms. Until those land, treat Breath-Dx as the **reference implementation** for **honest eNose metrics**, not as proof that any **single architecture** is **clinic-ready**.

When you present results to **hospital innovation boards**, lead with **temporal AUC** and **per-disease** tables. Add a **one-slide drift story** explaining why **random CV** overstated past pilots. That framing prevents **procurement** teams from comparing Breath-Dx numbers to **vendor brochures** trained on **leaky splits**.

## Workshop path for health ML teams

Structured **[AI workshops](https://explainx.ai/workshops)** help teams adopt **leakage-safe validation** before **clinical pilots**. Pair Breath-Dx exercises with **[InclusionMed](https://explainx.ai/blog/inclusionmed-healthcare-ai-benchmark-index-2026)** gap analysis for **equity** reporting.
