Datar Cancer Genetics Unveils Exacta AI: A Revolutionary AI-Driven Cancer Treatment Platform

Metropolis Healthcare Partners with Roche to Launch Self-Sampling HPV DNA Test for Cervical Cancer Screening in India

Union Budget 2025: Full Exemption from Basic Customs Duty (BCD) for 36 Lifesaving Drugs

Dr. Jitendra Singh Inaugurates Centre for Advanced Genomics & Precision Medicine at AIIMS Jammu

Union Budget 2025-26: Key Healthcare Measures for Cancer Care and Access to Life-Saving Drugs

FDA Approves Roche’s PATHWAY HER2 Test for Identifying HER2-Ultralow Breast Cancer Patients Eligible for ENHERTU

Bristol Myers Squibb’s Breyanzi® Receives CHMP Recommendation for Relapsed or Refractory Follicular Lymphoma in the EU

NHS to Offer Groundbreaking Gene Therapy for Sickle Cell Disease

Quibim Secures $50M Series A to Advance AI-Powered Imaging for Precision Medicine

Apollo Hospitals Unveils Advanced Oncology Facility in Chennai

Beckman Coulter’s Alzheimer’s Blood Test Receives FDA Breakthrough Device Designation

ANGLE Announces Breakthrough Dual CTC-DNA and ctDNA Analysis with Illumina’s NGS Solution

Diatech Pharmacogenetics and Merck Expand Collaboration to Enhance RAS Biomarker Testing in MEA Region

DCG Unveils AI-Powered Video-Reporting for Personalized Chemotherapy Selection

FDA Approves Monthly Maintenance Dosing for LEQEMBI® in Alzheimer’s Treatment

Home
/
update
/
details

Deep Learning Tool Alerts to Elevated Lung Cancer Risk in Non-Smoking Population

NextEdge Admin

06 Nov 2023

2 min 24 sec

Researchers from Massachusetts General Hospital have developed an AI-deep learning tool, known as the Chest-X-ray (CXR) Lung-Risk tool, aimed at identifying 10% to 20% of lung cancers in non-smokers. This tool analyzes routine chest X-rays, which are often available in a patient’s electronic medical record. In a study involving 17,407 never-smoker patients, the CXR tool demonstrated predictive ability for identifying high-risk individuals for lung cancer.

Senior author Michael Lu, MD, MPH, highlighted the challenge of identifying high-risk individuals who don’t smoke, stating that current tools rely on pack-based smoking history. The CXR tool, however, provides a means to identify high-risk patients from routine chest X-rays, potentially allowing for early detection in non-smokers.

The study focused on a group of never-smokers who underwent outpatient chest X-rays, with the primary outcome being a six-year incident lung cancer. The deep learning model categorized patients into low, moderate, and high-risk groups. Of the total patients, 28% were deemed high-risk, with 2.9% later receiving a lung cancer diagnosis, exceeding the 1.3% six-year risk threshold for recommended lung cancer screening CT.

While not considered a screening test, the CXR-Lung-Risk tool could identify high-risk non-smokers based potentially improving early detection and intervention in the absence of smoking history on routine chest X-rays, prompting further evaluation with lung cancer screening CT. The AI tool, developed using 147,497 chest X-rays, offers a reproducible method for flagging high-risk individuals, potentially improving early detection and intervention in the absence of smoking history.