A Lightweight Multi-Scale Attention Framework for Real-Time Spinal Endoscopic Instance Segmentation

TL;DR

The paper tackles real-time, multi-class instance segmentation in spinal endoscopy under hardware constraints. It introduces LMSF-A, a co-designed architecture with a lightweight C2f-Pro backbone, a dual-branch SSFF+TFE neck, and LMSH head, achieving strong accuracy with low latency. A clinically reviewed PELD dataset (610 images, 61 patients) is released to standardize evaluation. Empirical results show LMSF-A achieves 74.4 mAP@50 and 43.0 mAP50-95 with 1.8M parameters and 8.8 GFLOPs, running at 189 FPS, outperforming or matching heavier baselines. The work demonstrates practical applicability for real-time intraoperative guidance in spinal procedures and provides insights into deployment-friendly design trade-offs.

Abstract

Real-time instance segmentation for spinal endoscopy is important for identifying and protecting critical anatomy during surgery, but it is difficult because of the narrow field of view, specular highlights, smoke/bleeding, unclear boundaries, and large scale changes. Deployment is also constrained by limited surgical hardware, so the model must balance accuracy and speed and remain stable under small-batch (even batch-1) training. We propose LMSF-A, a lightweight multi-scale attention framework co-designed across backbone, neck, and head. The backbone uses a C2f-Pro module that combines RepViT-style re-parameterized convolution (RVB) with efficient multi-scale attention (EMA), enabling multi-branch training while collapsing into a single fast path for inference. The neck improves cross-scale consistency and boundary detail using Scale-Sequence Feature Fusion (SSFF) and Triple Feature Encoding (TFE), which strengthens high-resolution features. The head adopts a Lightweight Multi-task Shared Head (LMSH) with shared convolutions and GroupNorm to reduce parameters and support batch-1 stability. We also release the clinically reviewed PELD dataset (61 patients, 610 images) with instance masks for adipose tissue, bone, ligamentum flavum, and nerve. Experiments show that LMSF-A is highly competitive (or even better than) in all evaluation metrics and much lighter than most instance segmentation methods requiring only 1.8M parameters and 8.8 GFLOPs, and it generalizes well to a public teeth benchmark. Code and dataset: https://github.com/hhwmortal/PELD-Instance-segmentation.

Figures (10)

• Figure 1: Overall architecture of LMSF-A. GN denotes Group Normalization.
• Figure 2: Comparison between the traditional C2f and our C2f-Pro module.
• Figure 3: Representative samples from the PELD instance‑segmentation dataset. Top row: raw endoscopic frames captured at 1080$\times$720 resolution. Bottom row: corresponding pixel‑wise, instance‑level annotations for four surgical structures—adipose tissue (purple), bone (red), ligamentum flavum (yellow), and nerve (green). Examples illustrate typical variations in viewpoint, illumination, bleeding, and specular highlights encountered during percutaneous endoscopic lumbar discectomy.
• Figure 4: Percutaneous endoscopic lumbar discectomy (PELD) procedural progression. (A) Establishment of the working channel with ligamentum flavum obscuring deeper structures. (B) Partial flavectomy exposing subcutaneous fat and adjacent neural tissue. (C) Further clearance revealing bony landmarks and an intact nerve. (D) Adequate decompression of the compressed nerve root. Typical endoscopic artifacts include narrow field of view, specular highlights, irrigation/blood contamination, and frequent instrument occlusions.
• Figure 5: Qualitative comparison on the PELD instance-segmentation dataset.