Tissue classification in surgical workflows between healthy and tumoral regions remains challenging both during and post-surgery. The current standard practice consists of taking small biopsies directly after tumor resection and sending them to pathologists for an intraoperative margin assessment, which is time-consuming and error prone due to the necessarily limited size and number of samples. Then, after the surgery is completed, the resected tumor is sent to the pathology lab, where its type and grading are further confirmed. The present workflow is prone to inaccuracies and particularly difficult when the sample is resected in several pieces. Therefore, an intraoperative tissue classification technology is highly sought-after for a simplified surgical workflow and better patient outcome. Our work aims at using hyperspectral images (HSI) for contact- and tracer-free tissue differentiation. We introduce a deep learning-based algorithm for the classification of tissue type that is based on spectral information and can be applied simultaneously to the whole sample. We illustrate the performance of our method on ex vivo head and neck squamous cell cancer samples. The proposed algorithm can differentiate between three main classes: background, tumor, and healthy tissues. Our experiments first assess the generalization of the neural network on data from unseen cases. We then determine the minimal number of training examples needed to cover the variety of tissue spectral appearances seen in the clinical dataset. We evaluate the influence of the delay between resection and start of image acquisition on the quality of the recorded HSI and the prediction. Qualitative and quantitative evaluations support the applicability of hyperspectral imaging for tissue classification and demonstrate an agreement between surgeon annotations and neural network predictions in most test cases.
KEYWORDS: Deep learning, Tumors, Surgery, Neural networks, Hyperspectral imaging, RGB color model, Tissues, Cameras, Brain, Real time optical diagnostics
Surgery for gliomas (intrinsic brain tumors), especially when low-grade, is challenging due to the infiltrative nature of the lesion. Currently, no real-time, intra-operative, label-free and wide-field tool is available to assist and guide the surgeon to find the relevant demarcations for these tumors. While marker-based methods exist for the high-grade glioma case, there is no convenient solution available for the low-grade case; thus, marker-free optical techniques represent an attractive option. Although RGB imaging is a standard tool in surgical microscopes, it does not contain sufficient information for tissue differentiation. We leverage the richer information from hyperspectral imaging (HSI), acquired with a snapscan camera in the 468 − 787 nm range, coupled to a surgical microscope, to build a deep-learning-based diagnostic tool for cancer resection with potential for intra-operative guidance. However, the main limitation of the HSI snapscan camera is the image acquisition time, limiting its widespread deployment in the operation theater. Here, we investigate the effect of HSI channel reduction and pre-selection to scope the design space for the development of cheaper and faster sensors. Neural networks are used to identify the most important spectral channels for tumor tissue differentiation, optimizing the trade-off between the number of channels and precision to enable real-time intra-surgical application. We evaluate the performance of our method on a clinical dataset that was acquired during surgery on five patients. By demonstrating the possibility to efficiently detect low-grade glioma, these results can lead to better cancer resection demarcations, potentially improving treatment effectiveness and patient outcome.
Protoporphyrin IX (PpIX) fluorescence-guided surgery has established as a standard for resecting malign glioma. However, low-grade glioma or sparsely infiltrated brain often emit weak PpIX fluorescence and are hard to distinguish from non-pathological tissue. Furthermore, spectrally overlapping autofluorescence inherently limits the sensitivity of fluorescence-intensity based PpIX detection. We therefore integrated frequency-domain fluorescence lifetime imaging together with a spectrometer in a surgical microscope. When analyzing human glioma samples ex vivo, weak PpIX fluorescence could be differentiated from the autofluorescence background through increased lifetimes. Characteristic peaks in the spectral measurements (635, 705nm) confirmed low concentrations of PpIX in the tissue.
Significance: 5-Aminolevulinic acid (5-ALA)-based fluorescence guidance in conventional neurosurgical microscopes is limited to strongly fluorescent tumor tissue. Therefore, more sensitive, intrasurgical 5-ALA fluorescence visualization is needed.
Aim: Macroscopic fluorescence lifetime imaging (FLIM) was performed ex vivo on 5-ALA-labeled human glioma tissue through a surgical microscope to evaluate its feasibility and to compare it to fluorescence intensity imaging.
Approach: Frequency-domain FLIM was integrated into a surgical microscope, which enabled parallel wide-field white-light and fluorescence imaging. We first characterized our system and performed imaging of two samples of suspected low-grade glioma, which were compared to histopathology.
Results: Our imaging system enabled macroscopic FLIM of a 6.5 × 6.5 mm2 field of view at spatial resolutions <20 μm. A frame of 512 × 512 pixels with a lifetime accuracy <1 ns was obtained in 65 s. Compared to conventional fluorescence imaging, FLIM considerably highlighted areas with weak 5-ALA fluorescence, which was in good agreement with histopathology.
Conclusions: Integration of macroscopic FLIM into a surgical microscope is feasible and a promising method for improved tumor delineation.
During open brain surgery we acquire perfusion images non-invasively using laser Doppler imaging. The regions of
brain activity show a distinct signal in response to stimulation providing intraoperative functional brain maps of
remarkably strong contrast.
A stationary low coherence interferometer for optical coherence tomography (linear OCT, LOCT) based on Young's two-pinhole experiment is characterized theoretically. All OCT sensors either work in the time (TDOCT) or Fourier domain (FDOCT). In contrast to these setups, the interferometer described in this paper employs no moving parts in the reference arm and no spectrometers for depth profiling. Depth profiling is achieved by detecting the interference signal on a linear CCD-array. Different positions of the interference signal on
the CCD-array correspond to different depths inside the sample. The
interference signal of the setup and the sensitivity in the case of shot noise limited detection are derived theoretically and compared to sensors in the time domain. In-vitro images of porcine cornea demonstrate the clinical potential of the setup.
In this paper we demonstrate real-time in vivo and in vitro OCT images of human dental tissue obtained in a clinical setting. For the first time we have used a compact, commercial prototype OCT system with a surgical microscope as a beam delivery system for investigations of dental tissue. We have imaged demineralised tissue, caries lesions and restored teeth and demonstrate the detection of changes in tissue microstructure. We discuss the details of this system and its potential and limitations with respect to dental applications.
Optical coherence tomography (OCT) is a noninvasive imaging technology, which provides subsurface imaging of biological tissue with a resolution in the micrometer range. OCT sensors either work in the time or Fourier domain. We present a new interferometer setup based on a fiber double pinhole arrangement. Two fibers are placed in parallel similar to Young’s two-pinhole interference experiment with spatial coherent and temporal incoherent light. The interference pattern is observed on a linear CCD-array. A complete A-scan can be derived from a single readout of the CCD-array. The experimental setup is described in detail. The main parameters of the setup are derived theoretically and compared with experiments. First images of technical and biological samples are presented.
Time-resolved backscattering from randomly scattering media is studied experimentally with the aim of human skin diagnosis. The experimental setup consists of a self-modelocked Ti:Sapphire laser and a light gating technique based on sum-frequency generation. Aqueous solutions of latex microspheres were used as scattering medium. The experimentally determined temporal profiles of the recorded backscattered photons significantly depend on the specific conditions of the whole optical system. A systematic variation of the optical parameters was performed and an optimum arrangement was determined. In a first set of experiments, relatively weak concentrations of the scatterers were investigated and scattering lengths were determined. In a second series of experiments two-layered samples of latex suspensions with scattering properties similar to human skin were studied. Under these scattering conditions penetration depths of more than 1 mm could be obtained.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.