International Validation of Electronic Nose Technology for Fibrotic Interstitial Lung Diseases
Fibrotic interstitial lung diseases, often abbreviated as fILDs, are a group of chronic lung disorders in which scar tissue gradually builds up in the lung interstitium, the delicate tissue that supports the air sacs. Over time, this scarring makes it harder for oxygen to move into the bloodstream, leading to shortness of breath, dry cough, fatigue, and declining exercise tolerance. Because these symptoms are common to many other conditions, and because several fILD subtypes can look similar on imaging and clinical examination, diagnosis is often delayed and may require multiple tests, including bronchoscopy or lung biopsy in some cases.
This international multicenter study evaluated whether electronic nose technology, or eNose, could help distinguish between different types of interstitial lung disease using only exhaled breath. The findings suggest that breath profiling may become a practical, noninvasive tool to support diagnosis in specialized lung care settings.
Why Breath Analysis Matters
The idea behind an eNose is straightforward: diseases can change the pattern of volatile organic compounds, or VOCs, in a person’s breath. VOCs are small chemical molecules released by the body and influenced by inflammation, tissue damage, metabolism, and the lung microbiome. Instead of measuring each molecule individually, an eNose uses a sensor array to detect the overall “breath fingerprint.” That fingerprint can then be analyzed by pattern-recognition software.
In lung diseases, this approach is especially appealing because the airways and alveoli are directly exposed to exhaled breath. Prior single-center studies had already suggested that eNose patterns may help identify idiopathic pulmonary fibrosis and distinguish it from other lung diseases. However, before this technology could be considered clinically useful, it needed validation in different hospitals, countries, and patient populations.
How the Study Was Done
The investigators enrolled patients with interstitial lung disease who had been diagnosed through multidisciplinary team discussion, which is the current standard for complex ILD diagnosis. This means pulmonologists, radiologists, pathologists, and other specialists reviewed the clinical information together. All included patients also had pulmonary fibrosis visible on high-resolution chest computed tomography, or HRCT.
Patients were recruited from five international ILD expert centers. Breath samples were collected using the SpiroNose device, a type of portable eNose designed for exhaled breath analysis. The study team then compared breath profiles in two ways: first, they examined whether specific ILD subtypes could be separated from all other ILDs combined; second, they tested whether six different ILD subtypes could be distinguished from one another.
To reduce the risk of overfitting, the researchers trained their models on data from selected centers and then tested them externally in other centers. This is an important step because a diagnostic tool must work reliably outside the hospital where it was developed.
Main Findings
Breath profiles from 587 patients were analyzed. The results were encouraging. When ILD subtypes were compared with all other ILDs, the area under the curve, or AUC, ranged from 0.88 to 0.92 in the training set and from 0.75 to 0.95 in the external validation set. In diagnostic research, an AUC closer to 1.0 indicates better ability to separate disease groups, so these values suggest strong performance.
When the investigators looked at six individual ILD subtypes, the eNose also showed high discriminatory power. AUC values ranged from 0.95 to 0.98 in the training set and from 0.83 to 0.93 in the validation set. In practical terms, this means that breath patterns carried meaningful disease-specific information and that the technology remained useful when tested across centers.
These findings are especially notable because interstitial lung diseases are highly heterogeneous. They include conditions such as idiopathic pulmonary fibrosis, connective tissue disease-associated ILD, hypersensitivity pneumonitis, sarcoidosis with fibrotic change, and other forms of pulmonary fibrosis. Many of these disorders require different treatments and have different prognoses, so better early differentiation can influence patient management.
Clinical Meaning
The study supports the possibility that eNose technology could serve as a fast, noninvasive point-of-care test in the diagnostic workup of fibrotic ILD. A point-of-care tool is one that can be used close to the patient, potentially during a clinic visit, rather than requiring complex laboratory processing.
If integrated into routine care, an eNose test might help clinicians:
1. Identify patients who are more likely to have a particular ILD subtype.
2. Prioritize patients for further specialist testing.
3. Reduce diagnostic uncertainty.
4. Potentially limit the need for invasive procedures in selected patients.
That said, the eNose is not yet a replacement for current diagnostic standards. HRCT imaging, detailed clinical history, pulmonary function testing, serologic evaluation, exposure history, and multidisciplinary review remain essential. The eNose should be viewed as an adjunctive tool that may improve efficiency and confidence in diagnosis.
Why This Is Important for Patients
For patients, diagnostic delay in ILD can be frustrating and concerning. Symptoms often progress slowly, and the path to diagnosis may involve repeated appointments, imaging scans, breathing tests, and specialist referrals. A noninvasive breath test could one day shorten this path.
Because the study included multiple international expert centers, the results are more generalizable than those from a single hospital. This matters because breath profiles can be influenced by differences in equipment, patient selection, treatment patterns, and local practice. Demonstrating reproducible performance across centers is therefore a major step toward real-world use.
Limitations and Next Steps
Although the results are promising, several limitations should be kept in mind. First, the study was performed in expert centers, where patients typically have more complex disease and where diagnostic standards are high. Performance may differ in community hospitals or in earlier-stage disease.
Second, eNose technology is pattern-based. It does not identify the exact chemical compounds responsible for the signal, so the biological mechanisms behind the breath signatures remain only partially understood. Future studies may combine eNose screening with detailed chemical analysis of VOCs to better explain what the sensors are detecting.
Third, before broad clinical adoption, the technology will need further prospective validation, standardized testing procedures, and evidence that it changes clinical decision-making or improves outcomes. Cost-effectiveness, ease of implementation, and long-term reproducibility will also matter.
Bottom Line
This international validation study shows that electronic nose technology can accurately distinguish breath profiles from patients with different fibrotic interstitial lung diseases. The findings suggest that eNose testing could become a useful, easy-to-use supportive tool in the diagnosis of fILDs. While it is not ready to replace established diagnostic methods, it represents a promising step toward faster, less invasive, and more personalized lung disease care.
