Bronchiectasis is a heterogeneous condition with diverse etiologies and clinical manifestations. Its progression involves a vicious cycle of airway inflammation, recurrent infection, and structural damage, leading to persistent symptoms and declining lung function. Current management focuses on airway clearance and antibiotics, with no disease-modifying therapies available. Recognizing this heterogeneity is crucial for advancing targeted treatments and precision medicine.

Radiomics converts medical images into mineable data to reveal underlying pathophysiology. While applied in other respiratory diseases, its potential in bronchiectasis remains underexplored. Both radiomics and the lung microbiome are independently linked to disease severity in conditions like COPD, but their interplay is unclear. Integrating these modalities with clinical data could unlock novel insights, identify new therapeutic targets, and improve diagnostic and prognostic models.

However, few studies have investigated multimodal models combining radiomics, microbiome, and clinical features to predict outcomes in bronchiectasis. To address this gap, we designed a multicenter, retrospective study. It will analyze data from patients diagnosed between January 2020 and July 2025 to evaluate the combined value of radiomics, microbial features, and clinical parameters in diagnosing and predicting the progression of bronchiectasis.

Inclusion Criteria:

• ① Diagnosis of bronchiectasis;

  • Availability of both raw high-resolution computed tomography (HRCT) chest images and the corresponding radiology report; ③ Age ≥ 18 years.

Exclusion Criteria:

• ① Incomplete clinical data;

  • Absence of chest CT imaging studies and reports; ③ Other patients deemed ineligible for enrollment at the investigator's discretion.

Patients with a diagnosis of bronchiectasis established by high-resolution CT (HRCT) of the chest and its corresponding radiology report.