Large language models (LLMs) show great potential for clinical decision-making, yet most applications remain narrow, task-specific chat tools rather than systems integrated into clinical workflows ^1,2 . However, building physician copilots will require models that operate within the electronic health record (EHR), with governed access to patient data and the ability to initiate permitted EHR actions within defined safety constraints. Yet it remains unproven whether such a system can manage patient cases with physician-level performance. Here we show that MIRA (Medical Intelligence for Reasoning and Action), an autonomous artificial intelligence agent operating in a sandboxed EHR environment, can navigate a large clinical action space to obtain patient histories; order and interpret laboratory, imaging and microbiology tests; generate differential diagnoses; and formulate treatment plans such as prescribing medications, scheduling surgical procedures and planning admissions. In simulations on real patient cases spanning multiple diagnoses, MIRA outperformed physicians in diagnostic accuracy and made guideline-concordant, medication-safe and appropriate admission decisions. Compared with previous LLM applications that addressed isolated subtasks or provided free-text advice, these results suggest that an EHR-integrated artificial intelligence agent can turn clinical intent into structured, actionable EHR operations, possibly making it a more effective decision-support partner for physicians. Further work is needed to establish generalization, safety and governance through prospective, real-world studies.

Background: Peripheral nerve pathology is frequently encountered in clinical practice among peripheral nerve and extremity surgeons. One major factor limiting nerve regeneration and possibly leading to revision surgeries is the development of traumatic or postoperative adhesions and scarring around nerves. In experimental models, different materials have been studied to limit scar tissue formation when wrapped around nerves. Methods: A systematic review of studies describing nerve-wrapping materials in a non-transectional rat sciatic nerve model was performed following the PRISMA guidelines. Literature describing nerve-wrapping methods for the prevention of peripheral nerve scarring in rat sciatic nerve models was identified using PubMed and Web of Science, scanned for relevance and analyzed. Results: A total of 15 original articles describing 23 different materials or material combinations for nerve wrapping were included. The heterogeneity of the methods used did not allow a meta-analysis, thus, a systematic review was performed. Out of 28 intervention groups, 21 demonstrated a preventive effect on scar tissue formation in at least one qualitative or quantitative assessment method. Conclusions: The analyzed literature describes a variety of materials from different origins to limit peripheral nerve scarring and adhesions. Thus, a scar-preventive effect by wrapping peripheral nerves as adhesion prophylaxis seems likely. However, a quantitative comparison of the studies to identify the optimal material or technique is not possible with the diversity of used models and study designs. Therefore, further research needs to be performed to identify the optimal nerve wraps to be used routinely in clinical practice.

Scar tissue formation presents a significant barrier to peripheral nerve recovery in clinical practice. While different experimental methods have been described, there is no clinically available gold standard for its prevention. This study aims to determine the potential of fibrin glue (FG) to limit scarring around peripheral nerves. Thirty rats were divided into three groups: glutaraldehyde-induced sciatic nerve injury treated with FG (GA + FG), sciatic nerve injury with no treatment (GA), and no sciatic nerve injury (Sham). Neural regeneration was assessed with weekly measurements of the visual static sciatic index as a parameter for sciatic nerve function across a 12-week period. After 12 weeks, qualitative and quantitative histological analysis of scar tissue formation was performed. Furthermore, histomorphometric analysis and wet muscle weight analysis were performed after the postoperative observation period. The GA + FG group showed a faster functional recovery (6 versus 9 weeks) compared to the GA group. The FG-treated group showed significantly lower perineural scar tissue formation and significantly higher fiber density, myelin thickness, axon thickness, and myelinated fiber thickness than the GA group. A significantly higher wet muscle weight ratio of the tibialis anterior muscle was found in the GA + FG group compared to the GA group. Our results suggest that applying FG to injured nerves is a promising scar tissue prevention strategy associated with improved regeneration both at the microscopic and at the functional level. Our results can serve as a platform for innovation in the field of perineural regeneration with immense clinical potential.

Introduction: Extensive scar tissue formation after peripheral nerve injury or surgery is a common problem. To avoid perineural scarring, implanting a mechanical barrier protecting the nerve from inflammation processes in the perineural environment has shown promising results for functional recovery. This study investigates the potential of an acellular collagen-elastin matrix wrapped around a peripheral nerve after induction of scar tissue formation.Materials and methods: In the present study, 30 Lewis rats were separated into three groups and sciatic nerve scarring was induced with 2.5% glutaraldehyde (GA-CM) or 2.5% glutaraldehyde with a supplemental FDA-approved acellular collagen-elastin matrix application (GA+CM). Additionally, a sham group was included for control. Nerve regeneration was assessed by functional analysis using the Visual Statisc Sciatic Index (SSI) and MR neurography during the 12-week regeneration period. Histological and histomorphometry analysis were performed to evaluate the degree of postoperative scar tissue formation.Results: Histological analysis showed an extensive scar tissue formation for GA-CM. Connective tissue ratio was significantly (p < 0.009) reduced for GA+CM (1.347 ± 0.017) compared to GA-CM (1.518 ± 0.057). Similarly, compared to GA+CM, MR-Neurography revealed extensive scar tissue formation for GA-CM with a direct connection between nerve and paraneural environment. Distal to the injury site, quantitative analysis presented significantly higher axon density (p = 0.0145), thicker axon diameter (p = 0.0002) and thicker myelinated fiber thickness (p = 0.0008) for GA+CM compared to GA-CM. Evaluation of functional recovery revealed a significantly faster regeneration for GA+CM.Conclusion: The supplemental application of an acellular collagen-elastin matrix showed beneficial effects in histological, radiological, and functional analysis. Therefore, applying a collagen-elastin matrix around the nerve after peripheral nerve injury or surgery may have beneficial effects on preventing scar tissue formation in the long run. This represents a feasible approach to avoid scar tissue formation in peripheral nerve surgery.