Gregory DiGirolamo
University of Massachusetts Chan Medical School, MA·Psychology
Radiologists can miss up to half of hard-to-see lung nodules, the small spots on chest scans that may be early lung cancer. Yet trained experts often process more than they consciously notice.
- Fields
Psychology
Medicine & Clinical Research
Computer Science
- Topics
lung cancer
lung nodules
nodule detection
- Grants
- 1 known grant, through Jul 2027
What they work on
- Test how well eye-tracking biomarkers of non-conscious detection, recorded while radiologists search chest CTs, can train and refine machine learning models to find missed lung nodules.
- Measure whether feeding back the locations that the ML models flag as missed nodules increases the number of nodules radiologists detect.
- Determine whether the trained ML models generalize to new radiologists reading a new set of chest CT images.
More about the research
More than 80% of lung cancer patients are diagnosed at an advanced stage, and the 5-year survival rate is 21%. This project asks whether a radiologist's non-conscious detection of a nodule can be captured and used to catch nodules that would otherwise be missed.
Why it matters
Catching small lung nodules that radiologists currently miss could lead to earlier lung cancer diagnosis, when treatment is more likely to succeed. Machine learning tools that use radiologists' own non-conscious detection could reduce diagnostic errors in chest CT reading and may help raise lung cancer's 21% five-year survival rate.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Iakovos Toumazis
University of Texas MD Anderson Cancer Center, TX·Medicine & Clinical Research
Expand the ENGAGE lung cancer screening framework beyond age, sex and smoking history to include non-smoking risk factors such as race, family history and history of pulmonary disease.
Kang Wang
University of California-San Francisco, CA·Medicine & Clinical Research
Build a large standardized database of abdominal CT images and radiology reports, using large language models to automatically extract findings from reports and select data for AI training.
Daniela Raicu
DePaul University, IL·Computer Science
Develop a Semantic Deep-learning Neural Network (SDNN), a deep-learning model enhanced with knowledge of what visual features mean, that generates visual patterns and example images resembling the concepts radiologists use to read lung nodules.
The school
University of Massachusetts Chan Medical School: 29 labs on PI Pages, $350M a year on research, 55 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.