What happened
Shanghai United Imaging Intelligence built uAI MI Onco, software that uses deep learning on whole-body 18F-FDG PET/CT images of adults. It gives trained doctors triage prompts for focal FDG uptake in the chest. The Center for Medical Device Evaluation (CMDE) of China's National Medical Products Administration reviewed the company's application for Class III registration. The trial was the maker's own, a retrospective study of 596 cases at two sites in China. An expert panel's reading served as the reference. The software's prompts reached 90.97% sensitivity (95% CI 87.60% to 94.35%) and 94.98% specificity (92.59% to 97.38%). Both lower bounds exceeded the target of 80%. The software's processing time was 46.78±2.98 seconds. On 8 May 2026, CMDE concluded that “该产品上市带来的受益大于风险” (the benefits of putting the product on the market outweigh the risks) and recommended approval. The software read stored images, so it guided no care. The report does not name the sites.
Sources
-
Regulatory document, by authors with no tie to the system.
FAS 集和 PPS 集软件提示的灵敏度为 90.97%, 95%置信区间为 (87.60%, 94.35%) ,特异度为 94.98%,95%置信区间为(92.59%,97.38%)
经综合评价,在目前认知水平上,认为该产品上市带来的受益大于风险,综合剩余风险可接受。
Details
AI system
- Reported configuration
-
Release 1.0 (full version 1.00-2-24)
- Workflow
-
The software segmented anatomical structures on CT and analyzed the PET images with deep learning. It gave triage prompts for focal FDG uptake in the chest.
Follow-up
- Conditions for replication
-
Patients are adults with whole-body 18F-FDG PET/CT images. Trained doctors use the prompts and do not base a clinical decision on the software alone.
Entry details
- Sector
-
Healthcare.
- Purpose
-
Shanghai United Imaging Intelligence built uAI MI Onco to give trained doctors triage prompts for focal FDG uptake in the chest on whole-body PET/CT images of adults.
30 more questions not yet checked
model, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.
Cite this record
Peter H. Vartanian. (2026). uAI MI Onco flagged chest uptake on PET/CT scans in China's review (AIBD-1092). Artificial Intelligence Benefits Database, edition of 1 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1092.html (archived at https://doi.org/10.7910/DVN/AADSII, version 7.0)