Design and performance of an illumination-based hyperspectral laparoscopic system.

Abstract

This paper presents the design and performance evaluation of an illumination-based hyperspectral imaging system for laparoscopic surgery, referred to as the HyperScope. Hyperspectral imaging holds significant promise in minimally invasive surgery due to its ability to capture detailed spectral signatures that can enhance intraoperative tissue characterization and improve surgical decision-making without the need for exogenous contrast agents. The presented HyperScope system integrates a supercontinuum laser source, a laser line tunable filter, and a laparoscope-coupled short-wave infrared camera, enabling spectral imaging across a broad wavelength range (450-1500 nm) with high spectral resolution and sufficient spatial resolution quantified through several metrics. The system's hardware and software components are optimized for clinical usability and spectral fidelity. Performance was assessed through a series of technical experiments, including spectral response characterization, a modulation transfer function measurement using the slanted-edge method, and an evaluation of distance compensation via pre-processing. The system achieves a spectral deviation of 0.16 nm (450-975 nm) and -1.23 nm (1000-1675 nm), with mean spatial resolutions of 0.02 mm in the visible range (450-800 nm) and 0.14 mm in the SWIR range (800-1500 nm). The results demonstrate consistent spectral accuracy, acceptable spatial resolution across the spectral range, and effective correction of distance-induced spectral variability using standard normal variate (SNV) processing. These findings support the potential of the HyperScope system for future integration into clinical MIS workflows, with applications in tumor detection, perfusion assessment, and real-time optical biopsies.