Kamil Jóźwik

Background: ICU care is time critical and data dense, making it a promising but high-risk setting for AI decision support when tools are weakly validated. ICU AI evidence is heterogeneous, with limited external validation, inconsistent clinically actionable reporting, and scarce real-world impact data, yielding fragmented review conclusions. We mapped five prespecified ICU domains and assessed clinical and implementation maturity to identify key translational gaps. Methods: We performed a PRIOR-aligned overview of systematic reviews with prespecified maturity constructs. PubMed, Embase, and Web of Science were searched (title and abstract) on 13 December 2025, supplemented by backward citation searching. Two reviewers screened and extracted data with arbitration, assessed the review-level risk of bias using ROBIS, and synthesized findings without meta-analysis using a SWiM-guided narrative prioritizing AUROC ranges. Results: We included 34 systematic reviews (2017–2025) across five ICU domains, dominated by prognostic and early warning applications, mostly in adult populations and commonly using EHR and multimodal inputs. Reporting focused on discrimination, with AUROC ranges roughly 0.54–0.99 for prognostic tasks and 0.64–0.99 for diagnostic tasks, while calibration and clinical utility were rarely addressed and overlap suggested partial dependence. Maturity signals clustered at low-to-intermediate levels, with no evidence for routine, and regulated CDS deployment at the review level. Conclusions: Review-level evidence indicates a translational gap between retrospective performance and clinically mature, safely deployable ICU AI, supporting priorities for external validation, prospective impact studies, standardized reporting including calibration, and governance-focused implementation.

Study Design.: Systematic review with meta-analysis. Objective.: This meta-analysis aims to assess robotic and conventional surgical techniques for instrumentation of the spine among pediatric, adolescent, and young adult populations. Summary of Background Data.: Spinal deformities among younger patients may be treated with pedicle screw instrumentation; however, freehand guidance often leads to screw misplacements and consequent complications. Although computer-navigation surgery improved the screw accuracy, physiological limitations of the surgeon still remain a serious challenge. Robot-assisted (RA) surgery became a novel technique for enhancing screw accuracy. Methods.: PRISMA and Cochrane Handbook were followed, and five databases were searched from inception. Studies analyzing patients aged up to 25 years old, and comparing RA to freehand or navigation were included. Outcomes included screw accuracy, perioperative outcomes, and complications. Meta-analysis was pooled with random-effects model. Results.: Finally, 10 studies, 550 patients, and 8061 screws were included in the meta-analysis. Accurate screw placement rates (grade 0: OR 2.33, P P P =0.03). Blood loss, Cobb angle correction, hospital stay, or complications showed no significant differences between the two groups. Conclusions.: RA surgery leads to improved accuracy of the pedicle scores; however, this does not directly translate to better clinical or curvature outcomes. Moreover, longer operation times were observed in the robotic cohort. More high-quality studies are needed to validate these findings.

There is a growing interest in the use of machine learning (ML) for the diagnosis of retinoblastoma and leukocoria, and this study aims to systematically evaluate its performance compared with reference standards. A systematic review and meta-analysis were performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement. We included studies using ML to diagnose retinoblastoma or leukocoria and providing enough data for analysis of diagnostic accuracy. We calculated sensitivity, specificity, and other measures of diagnostic performance. Twelve studies were included. Pooled sensitivity for retinoblastoma and leukocoria detection was 0.972 with ML models, indicating high potential for screening. However, high heterogeneity in the analyses was observed. The review also noted biases in some studies, along with small sample sizes that would limit generalizability. ML models appear to be promising for retinoblastoma diagnosis; however, limitations in specificity and potential methodological bias need further investigation. Incorporating research that used photographs taken with smartphone cameras indicates that ML-based diagnosis may become even more widely available through the use of such technology. Future studies need to have better specificity of the model, less bias in the methodology, must be conducted on large-scale datasets and they should address the cost-effective analysis compared with traditional methods. The incorporation of ML into the practice of retinoblastoma diagnosis has the capacity to transform the mode of detecting this condition and ultimately enhance patient management.

Atlantoaxial dislocation (AAD) is a serious condition in which the first two cervical vertebrae lose their anatomical position and stability. This may lead to neurological complications, including death. The treatment of AAD remains controversial, and posterior instrumentation with pedicle screw placement is one of the commonly used methods. The goal of this study is to evaluate the effectiveness of robot-assisted (RA) surgery compared to conventional techniques for posterior AAD instrumentation. Following Cochrane and PRISMA guidelines, five major databases were searched from inception to December 2024. Studies evaluating RA surgery compared to freehand (FH) or computer navigation methods were included. Outcomes included screw accuracy, surgery time, radiation exposure, and complications. A random-effects mean difference (MD) and odds ratio (OR) were used. After a comprehensive search, three studies were included. Two studies compared RA to FH, and one compared RA to navigation. Blood loss (MD -103.11 [-148.73, -57.49], p < 0.01) and radiation dose (MD -8.19 [-10.18, -6.19], p < 0.01) were lower in the RA group compared to FH. Operating time was significantly longer in the RA group compared to conventional methods (MD 36.55 [30.07, 43.04], p < 0.01). The number of correct screws (OR 2.15 [1.11, 4.15], p = 0.02) and satisfactory screws (OR 3.18 [1.13, 8.90], p = 0.03) was higher in the RA group. RA surgery provides significantly higher pedicle screw accuracy compared to FH and navigation methods, although at the cost of increased surgery time. Blood loss and radiation exposure were lower in the RA group compared to FH. While the results are promising, this meta-analysis included only three small trials. Future randomized controlled trials are needed to fully evaluate the effectiveness of RA surgery for AAD instrumentation.

Renal artery aneurysms (RAAs) are a rare vascular condition. Robot-assisted surgery offers a minimally invasive approach for RAA repair, potentially improving surgical outcomes. This review investigates the current evidence on the effectiveness and limitations of this technique. A systematic search following PRISMA guidelines identified relevant studies across five electronic databases. Studies investigating the use of robot-assisted surgery for RAA repair were included. The review identified 11 studies encompassing a total of 23 patients. Procedures included aneurysmectomy, end-to-end anastomosis, prosthetic graft repair, and even coil embolization. All surgeries were successful, with only minor complications reported in four cases. Robot-assisted RAA repair shows promise as a minimally invasive approach with encouraging preliminary outcomes. However, the limited data come from small studies. Future advancements in robotic technology hold the potential to optimize this approach for improved patient care.

Vestibular Schwannoma (VS) is a rare tumor with varied incidence rates, predominantly affecting the 60–69 age group. In the era of artificial intelligence (AI), deep learning (DL) algorithms show promise in automating diagnosis. However, a knowledge gap exists in the automated segmentation of VS using DL. To address this gap, this meta-analysis aims to provide insights into the current state of DL algorithms applied to MR images of VS. Methodology: Following 2020 PRISMA guidelines, a search across four databases was conducted. Inclusion criteria focused on articles using DL for VS MR image segmentation. The primary metric was the Dice score, supplemented by relative volume error (RVE) and average symmetric surface distance (ASSD). Results: The search process identified 752 articles, leading to 11 studies for meta-analysis. A QUADAS- 2 analysis revealed varying biases. The overall Dice score for 56 models was 0.89 (CI: 0.88–0.90), with high heterogeneity (I2 = 95.9%). Subgroup analyses based on DL architecture, MRI inputs, and testing set sizes revealed performance variations. 2.5D DL networks demonstrated comparable efficacy to 3D networks. Imaging input analyses highlighted the superiority of contrast-enhanced T1-weighted imaging and mixed MRI inputs. Discussion: This study fills a gap in systematic review in the automated segmentation of VS using DL techniques. Despite promising results, limitations include publication bias and high heterogeneity. Future research should focus on standardized designs, larger testing sets, and addressing biases for more reliable results. DL have promising efficacy in VS diagnosis, however further validation and standardization is needed. Conclusion: In conclusion, this meta-analysis provides comprehensive review into the current landscape of automated VS segmentation using DL. The high Dice score indicates promising agreement in segmentation, yet challenges like bias and heterogeneity must be addressed in the future research.