Skip to content

Deep Learning-Based Detection of Pneumothorax on Chest Radiographs after Percutaneous Lung Biopsy and/or Ablation

Deep Learning for Detecting Pneumothorax on Chest X-Rays Following Percutaneous Lung Biopsy and/or Ablation in Colorectal Cancer Patients with Pulmonary Metastases: A Prospective Cohort Clinical Study

Status
Active, not recruiting
Phases
Phase 3
Study type
Observational
Source
ChiCTR
Registry ID
ChiCTR2600128965
Enrollment
Unknown
Registered
2026-07-29
Start date
2026-07-31
Completion date
Unknown
Last updated
2026-08-03

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Pneumothorax

Interventions

Gold Standard:The reference standard was established by thoracic radiologists (A, B, and C, with 10, 15, and 20 years of experience in thoracic radiology. The reference standard for quantifying pneumo
Index test:Primary EndpointComparison of the diagnostic performance between PTX-AI v1 and radiology residents for pneumothorax detection, using the following metrics: sensitivity, specificity, positiv

Sponsors

Fudan University Shanghai Cancer Center
Lead Sponsor

Eligibility

Sex/Gender
All
Age
18 Years to 80 Years

Inclusion criteria

Inclusion criteria: Participants eligible for inclusion in this study must meet criteria 1+2+3+4+6 or 1+2+3+5+6: 1) Have signed an informed consent form; 2) Be aged between 18 and 80 years; 3) Have a pathologically confirmed diagnosis of colon or rectal cancer; 4) Patients with a clinical and imaging diagnosis of a lung tumour, where imaging and follow-up cannot confirm metastasis, and who require a percutaneous needle biopsy to confirm the pathology; or patients who require tissue samples to be taken for a clinical trial; 5) Patients with a clinical, imaging or pathological diagnosis of lung metastasis who require percutaneous local ablation; 6) Patients who have undergone at least one post-operative chest X-ray.

Exclusion criteria

Exclusion criteria: Subjects should not participate in the trial if any of the following conditions apply: 1. Patients with severe cardiac or pulmonary insufficiency who, in the investigator’s assessment, are unsuitable for participation in this clinical trial; 2. Patients with marked coagulation abnormalities; 3. Patients with obvious infectious lesions along the intended puncture site; 4. Patients with significant pleural effusion on the affected side; 5. Patients requiring management of post-operative complications such as haemorrhage, air embolism, or cardiovascular or cerebrovascular events; 6. Pregnant or breastfeeding women; 7. Patients with cognitive impairment; 8. Patients assessed by the investigator as unlikely to cooperate adequately with this trial; 9. Other circumstances deemed by the investigator to render the patient unsuitable for participation in this clinical trial.

Design outcomes

Primary

MeasureTime frame
Diagnostic accuracy for pneumothorax detection;

Secondary

MeasureTime frame
Accuracy of pneumothorax segmentation;Dice coefficient;Accuracy of pneumothorax volume measurement;Degree of deviation from the gold standard (volume of pneumothorax on CT scan images);

Countries

China

Contacts

Public ContactHe Xinhong

Fudan University Shanghai Cancer Center

13817694419@163.com+86 21 64175590

Outcome results

None listed

Source: ChiCTR (via WHO ICTRP) · Data processed: Aug 10, 2026