UCLA researchers in the Department of Radiological Sciences and Biostatistics developed a domain-knowledge assisted automatic diagnosis of idiopathic pulmonary fibrosis using a high-resolution computed tomography.
BACKGROUND:
Idiopathic pulmonary fibrosis (IPF) is a specific form of chronic, fibrosing interstitial pneumonia of unknown cause. The chronic scarring of lung leads to progressive and irreversible decline in lung function. IPF affects around 14 to 27.9 cases per 100,000 people in the United States. The median survival time ranges from 3 to 5 years. The diagnosis of IPF is a complex and lengthy process that involves collaborations of multi-disciplinary specialists: clinicians, radiologists, and pathologists. A fast and reliable diagnostic tool is needed for IPF diagnosis.
INNOVATION:
UCLA researchers in the Department of Radiological Sciences and Biostatistics developed a domain-knowledge assisted automatic diagnosis of IPF using a high-resolution computed tomography (HRCT). Axial lung CT images were acquired from five multi-center studies, which sum up to 330 IPF patients and 650 non-IPF ILD patients. The scientists leveraged the population-level domain knowledge based on the D-optimal design criterion to judiciously select CT slices that could be used for the disease diagnostics. This technology achieved satisfactory results with overall sensitivity, specificity, and accuracy greater than 90%. The clinical significance of this technology includes (1) facilitating automatic classification of IPF, (2) enabling early diagnosis and treatment and potentially a prolonged survival time, and (3) ideally reducing the need for lung biopsy in the diagnosis process.
POTENTIAL APPLICATIONS:
ADVANTAGES:
DEVELOPMENT-TO-DATE:
The study has been validated in mice.