Aidoc reports faster PE response and broader therapy use with AI workflows
AI-enabled pulmonary embolism workflows were linked to faster PERT activation, higher advanced therapy utilization and shorter hospital stays in research presented at the 2026 PERT Consortium Annual Meeting in Boston. The findings, highlighted by Aidoc, examine how clinical AI can connect imaging results with specialist teams and subsequent care decisions.
- Mean PERT activation time fell from 211 to 111 minutes in one study.
- Advanced therapy utilization rose from 25.7% to 86.5% at Yale New Haven Health.
- A multi-site study reported shorter inpatient and ICU stays after implementation.
- PERT alerts occurred less often in incidental PE cases than in non-incidental cases.
Faster activation and more interventions
Researchers at University Hospitals Cleveland Medical Center compared traditional PERT activation with a workflow implemented in partnership with Aidoc. The system analyzed CTPA scans for suspected PE cases and automatically notified the PERT team through a mobile platform.
In the retrospective study presented by Satawart Upadhyay, MD, the mean time to PERT activation fell 47%, from 211 to 111 minutes. Intervention rates increased from 7.3% to 20.8%, which was nearly three times higher in the AI-enabled group. Both outcomes had a p-value of less than 0.001.
The findings suggest that sending imaging insights directly to care teams may reduce communication delays and allow clinicians to evaluate treatment options sooner, according to the research highlighted by Aidoc.
Advanced therapy consultation increased at Yale New Haven Health
A separate study at Yale New Haven Health examined changes after implementing an AI-enabled PE triage and risk stratification workflow incorporating Aidoc. Matteo Conte, MD, and colleagues compared 208 patients before implementation with 222 patients afterward.
Among eligible patients, PERT consultation rose from 77.1% to 100%. Advanced therapy utilization increased from 25.7% to 86.5%, with both increases reported as statistically significant. Large-bore mechanical thrombectomy accounted for 13.1% of patients in the post-AI cohort, compared with none before implementation.
The study points to a possible role for AI-assisted triage, automated PERT notification and structured clinical review in creating more consistent pathways for patients who may benefit from advanced therapies.
The reported effects extended beyond the initial alert
At Jefferson Health, Mark Edelstein, MD, and colleagues analyzed more than 17,000 CTPA cases before and after implementation of an AI-assisted PE workflow with automated PERT notifications. The multi-site study reported an 18.4% reduction in median inpatient length of stay, a 20.7% reduction in median ICU length of stay and a 12.9% reduction in median time to thrombectomy treatment.
Among patients undergoing thrombectomy, 30-day readmissions fell by 49.4%, from 21.95% to 11.11%. Aidoc says the study was observational and did not establish that AI alone caused the reported changes. That limitation makes the findings evidence of an association rather than proof of a standalone AI effect.
Incidental PE cases followed different pathways
Research from Temple University Hospital examined incidental pulmonary embolism, or IPE, identified during imaging performed for unrelated reasons. Parth Rali, MD, analyzed 1,475 positive PE and IPE cases, including 190 documented PERT activation alerts.
PERT alerts occurred in 5.0% of incidental PE cases and 18.6% of non-incidental cases. Incidental cases were more frequently associated with inpatient settings, while non-incidental cases were more commonly identified in emergency departments.
For hospitals, the findings indicate that patients may enter PE care through different settings and may require different evaluation pathways. Aidoc presents this distinction as relevant to efforts to refine PE risk stratification and PERT activation protocols.
Why the findings matter for PE care
Together, the studies shift attention from detecting a suspected PE to coordinating what happens afterward. The research covered specialist activation, consultation, advanced therapy use, treatment timing, hospital stays and readmissions.
Aidoc says the findings are relevant as hospitals evaluate how to operationalize the 2026 AHA/ACC multisociety acute PE guideline, which introduces updated clinical categories and emphasizes multidisciplinary care. The company’s stated opportunity is to bring imaging insights, patient-specific clinical information and communication between care teams into a connected workflow.
The evidence presented at PERT Symposium 2026 supports further evaluation of how hospitals can use AI to coordinate PE care. It also reinforces that the appropriate response may differ according to how and where a patient’s PE is identified.


