Engineering Journal: Volume-11, Issue-11, November, 2025

Abstract— Retained surgical items (RSIs), particularly gauze pieces and mops, remain a preventable yet persistent threat to patient safety in orthopedic operations. Manual counting protocols, though standardized worldwide, are vulnerable to human fatigue, distraction, and workflow complexity conditions frequently intensified in lengthy orthopedic procedures that involve multiple instruments, draping layers, and substantial blood loss. To address these challenges, diverse automation systems have been developed using bar-coding, radio-frequency identification (RFID), radiofrequency detection (RFD) wands, computer vision (CV), and sensor-fusion architectures. This paper presents a comprehensive review of gauze and mop counting automation systems with a specific focus on orthopedic surgical safety. A systematic literature search was conducted across Scopus, PubMed, IEEE Xplore, and ScienceDirect databases for the period January 2010 to May 2025 using PRISMA based screening criteria. Each study was analyzed for detection accuracy, workflow integration, sterility, human factors, and compliance with international standards (ISO 13485, 14971, IEC 60601).
The review identifies that while RFID and RFD technologies achieve high detection sensitivity, they face interference and sterilization constraints; computer vision approaches offer real-time potential but remain limited by dataset variability and occlusion. Few studies report on orthopedic specific validation or multimodal fusion strategies. The synthesis highlights critical research gaps in interoperability, calibration, regulatory validation, and human-automation collaboration.
Building upon these findings, the paper proposes a robotics oriented framework integrating RFID and CV within a closed-loop counting ecosystem capable of edge-level decision making and standardized audit trails. Such an approach could substantially enhance count accuracy, reduce intra-operative delays, and strengthen traceability. The review thus provides a consolidated evidence base and future roadmap toward intelligent, standards-compliant counting automation for safer orthopedic surgery. This review identifies critical pathways for future research in robotics-assisted surgical safety systems.
Keywords— Surgical Safety; RFID; Computer Vision; Retained Surgical Items; Orthopedic Surgery; Automation Framework.

Abstract— Accurate short-horizon stock-movement forecasting remains a central problem in computational finance, where even small directional errors can accumulate into significant trading risk. The most challenging regime is the neutral state—intervals with minor price changes that are easily masked by noise. To address this challenge, we compare three complementary learning paradigms and their combinations across multiple lookback horizons for three representative equities (AAPL, GOOG, TSLA). We evaluate Long Short-Term Memory (LSTM) networks for temporal dynamics, Convolutional Neural Networks (CNNs) on polar-transformed price images for spatial pattern extraction, and XGBoost on tabular technical indicators for structured feature learning.
Empirical results (Appendices A–C) reveal distinct horizon-dependent behaviors: CNNs excel at ultra-short windows (W = 1–3) with perfect accuracy and neutral-F1 ≈ 1.00 but deteriorate rapidly as horizons lengthen; LSTMs gain overall accuracy with longer windows (W = 30–60o) but lose sensitivity to neutral segments; and XGBoost remains the most stable single model, maintaining accuracy ≈ 0.89–0.93, low loss ≈ 0.4–0.6, and neutral-F1 ≈ 0.89–0.96 across assets.
Building on these complementary patterns, we propose fusion frameworks that integrate CNN and XGBoost outputs through weighted voting, cascaded thresholds, and probability-smoothed blending. The best configuration—probability-smoothed fusion—achieves roughly a 3–4 percentage-point improvement in neutral-F1 over the strongest standalone model while preserving comparable accuracy and calibration loss. The LSTM is retained solely as a benchmark to illustrate sequence-model trade-offs and is not included in the fusion.
Together, the results demonstrate that combining spatial and tabular perspectives yields more balanced recognition of neutral states without sacrificing directional accuracy. Accuracy measures overall correctness, loss captures probabilistic calibration, and F1 quantifies class-wise precision–recall balance. Viewed jointly, these metrics show that CNN–XGBoost fusion produces smoother and more interpretable predictions across assets and horizons. Such stability can reduce overtrading during ambiguous market phases, improving risk-adjusted decision-making in algorithmic trading strategies.
Keywords— Short-horizon stock prediction, CNN–XGBoost fusion, polar-coordinate transformation, financial time-series classification, probabilistic calibration, neutral-class F1.