The edition · Radiology
Commercial AI shaved 15% off chest CT reporting time across 39,000 scans
A real-world before-and-after study puts a number on what nodule AI does to throughput, saline tract sealing more than halves pneumothorax after CT-guided lung biopsy, and photon-counting CT matches MRI for breast staging while missing nearly half the microcalcifications.
The edition in brief
AI in radiology is usually sold on accuracy and bought on throughput, and throughput has rarely been measured. A single-centre before-and-after study of 39,323 chest CT examinations, split around the introduction of commercial nodule software, found adjusted median reporting time falling from 21.3 to 18.2 minutes, a 14.6% reduction (adjusted hazard ratio 1.17, 95% CI 1.14 to 1.21). The gain was uneven: 41% faster for ECG-gated thoracic CT and 25% for thoracic radiologists, but 7% slower in the emergency department. It is a retrospective pre/post design and cannot separate the software from everything else that changed over three years. The most directly usable finding is procedural. A meta-analysis of seven studies and 1,455 patients found that instilling saline into the biopsy tract after CT-guided lung biopsy cut pneumothorax (risk ratio 0.46, 95% CI 0.35 to 0.61) and more than halved the need for a chest drain (risk ratio 0.22, 95% CI 0.12 to 0.42). It is cheap and adds little to the procedure. Two staging papers refine rather than change practice: photon-counting CT matched MRI for breast lesion characterisation but missed 44% of microcalcifications, so it does not replace mammography; and for lateral lymph nodes in rectal cancer, FDG PET/CT was no more sensitive than MRI but far more specific (0.97 versus 0.82), useful for confirming rather than excluding disease. The ACR has updated its appropriateness criteria for myelopathy imaging.
Nodule AI cut chest CT reporting time by 15%, but not evenly
Commercial nodule AI cut chest CT reporting time by about 15% overall, concentrated in thoracic work and absent or negative elsewhere — useful for capacity, but not a uniform speed-up.
Saline in the tract more than halves pneumothorax after lung biopsy
Consider routine saline tract sealing during CT-guided lung biopsy: it roughly halves pneumothorax and cuts chest-drain placement by about three-quarters, at negligible cost.
Photon-counting CT matched MRI for breast staging but missed microcalcifications
Photon-counting CT performed like MRI for breast lesion and node assessment but missed nearly half of microcalcifications, so it is a possible MRI alternative, not a mammography replacement.
For rectal lateral nodes, PET/CT is the specific test, MRI the sensitive one
In rectal cancer, add FDG PET/CT to confirm suspected lateral nodal metastasis before dissection — its high specificity (0.97) rules disease in, though neither test reliably rules it out.
No new drug or device action for imaging today; ACR updates myelopathy criteria
No new imaging-relevant regulatory action today; the ACR's 2026 myelopathy appropriateness criteria are the current reference for choosing first-line imaging in suspected cord dysfunction.
A negative node test is not a clear field
When a modality is specific but not sensitive, say so in the report — a negative result is not the same as absence of disease.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this one is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for radiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free