Health Condition 1: J988- Other specified respiratory disorders
Conditions
Interventions
Intervention1: Predible Lung:
Deep Learning Based Software Tool for detection and characterization of Lung Nodules on CT Images
: Deep learning algorithm (Predible Lung) developed using 3D Convolution
Sponsors
Predible Health P Ltd
Eligibility
Inclusion criteria
Inclusion criteria: Retrospectively CT Images of males and females above the age of 18 years and having Lung pathology will be read 1. Subject: Men and Women, 18-years and older 2. Studies: Computed Tomography of the Lung
Exclusion criteria
Exclusion criteria: 1. Individuals with CT scans that have motion artefacts and/or poor image quality 2. Scans with consolidation, lobar or greater atelectasis, airways disease, interstitial lung disease, masses larger than 3 cm, and large pleural effusions 3. Scans with nodules with tree-in-bud appearance, nodules in diseased/interstitial background and clustered nodules 4. Apices cannot be cropped
Design outcomes
Primary
| Measure | Time frame |
|---|---|
| Compare the sensitivity of Predible Lung in Lung Nodule Detection from Lung CT as measured by a Panel of 3 RadiologistsTimepoint: Not Applicable as it is a Retrospective Observational Study | — |
Secondary
| Measure | Time frame |
|---|---|
| Average false positives per scan flagged by Predible Lung as compared to outcome reported by panel of 3 radiologistsTimepoint: Not Applicable as it is a Retrospective Observational Study | — |
Countries
India
Contacts
Public ContactMr Suthirth Vaidya
Mahajan Imaging Education and Research Academy
Outcome results
None listed