Canonical text (the exact bytes fingerprinted)

TRUVACE RECORD VERSION
record: TRV-2026-1805
version: 1
kind: certified
reason: Certified into the record
timestamp: 2026-10-18T10:41:35.373534Z
status: published
lens: g_space
sector: health
headline: A Hybrid Human-AI Workflow for Ultrasound Report Quality Control: A Multicenter Performance and Efficiency Evaluation
dek: Rationale and objectives: This study evaluated a hybrid human-AI workflow for ultrasound report quality assurance using multicenter retrospective data.
Materials and methods: 786 anonymized ultrasound reports covering 22 examination categories were collected from four hospitals. Reports containing naturally occurring or implanted errors were matched across arms. GPT-4o and DeepSeek-v3 performed initial screening before review by six human reviewers (HRs). HR with blinded expert over-reading (HR-ex) served as con…
gain_title: Pairing GPT-4o or DeepSeek-v3 screening with human review matched single-reviewer accuracy on ultrasound reports while cutting median audit time from 51.0 s to about 11 to 12 s.
problem_title: (none)
trace_subject: (none)
gain_reading: Pairing GPT-4o or DeepSeek-v3 screening with human review matched single-reviewer accuracy on ultrasound reports while cutting median audit time from 51.0 s to about 11 to 12 s.
gain_evidence: Median review time decreased substantially from 51.0 s (HR-ex) to 11.0 s (4o-HYB) and 12.0 s (v3-HYB) | accuracy comparable to single-physician review while substantially reducing audit time
problem_reading: (none)
problem_evidence: (none)
quick_read: Researchers tested a workflow in which large language models screened ultrasound reports for errors before human reviewers checked them. The test used 786 anonymized reports from four hospitals across 22 examination categories, with some errors implanted. A second validation used 1680 consecutive reports with no implanted errors.

The result suggests AI-assisted screening could speed up report auditing without lowering accuracy below that of a single physician. However, it did not match expert over-reading and let more errors through. The authors say prospective real-world validation is needed before routine clinical use.
limitation: The study was retrospective and the authors state that prospective real-world validation is required before routine clinical implementation; the hybrid also trailed expert over-reading on accuracy and false negatives.
tag: Evidence-backed gain
key_points: Accuracy was 80.2% for GPT-4o hybrid review, 82.5% for DeepSeek-v3 hybrid review and 82.3% for human review, versus 87.3% for expert over-reading. | Human verification reduced both false negatives and false positives produced by the language models. | Severity assessment did not indicate preferential omission of clinically significant errors. | In the supplementary set with 9.6% error prevalence, the DeepSeek hybrid had a 22.6% false-negative rate versus 24.3% for human review.
rundown: The study used retrospective data in which reports with naturally occurring or implanted errors were matched across arms. GPT-4o and DeepSeek-v3 did the first screening, and six human reviewers then examined the results. Human review with blinded expert over-reading served as the control.

On review time, the hybrid arms took a median of 11.0 and 12.0 seconds per report, compared with 51.0 seconds for expert over-reading. In the supplementary validation, median times were 15.4 seconds for the DeepSeek-assisted hybrid and 24.3 seconds for human review.

In that validation, false-positive rates were 7.7% for the hybrid and 10.0% for human review. Both approaches still missed more than a fifth of errors, at 22.6% and 24.3% respectively.
sources:
- peer_reviewed | Academic Radiology | https://doi.org/10.1016/j.acra.2026.09.069 | 2026-10-17
prev: 0000000000000000000000000000000000000000000000000000000000000000

sha256

01300edaa7cea936935f8ac5f23e7d158758c5822cd6976ed34fe7f906740779

previous

0000000000000000000000000000000000000000000000000000000000000000