Abstract:C omplete tumor resection is of utmost importance and a cornerstone in the treatment of numerous tumors; it is the aim of every surgeon. Surgery might be followed by adjuvant therapy to complete this task, but it still may fail as a result of limited recognition of tumor cells at the resection site. An incomplete tumor resection is at risk of tumor recurrence regardless of adjuvant therapies, and recurrence decreases quality of life and shortens life expectancy. Tumor diagnosis and intraoperative definition of tumor borders are based on various visualization methods and on the histopathological examination of a limited number of biopsy specimens via frozen sections. Today, gold standard surgical visualization tools include surgical “microscopes,” laparoscopes, ultrasonography, intraoperative magnetic resonance imaging, and computed tomography, as well as more sophisticated navigation devices. None of these is able to differentiate tumor from normal tissue at a cellular level, which is necessary to achieve higher accuracy in surgical tumor therapy, especially for malignant tumors. Therefore, frozen sections are typically used to define and differentiate tumor cells ex vivo. Normally, they are removed when the tumor is exposed but not yet resected. Unfortunately, intraoperative histopathology has several shortcomings, and many biopsies are inconclusive. First, the tissue architecture of the tumor can be altered during specimen preparation and can be mechanically destroyed. In addition, sampling errors are possible as a result of tissue heterogeneity. Finally, the lack of realtime interactivity with the pathologist and the waiting time of about 30 minutes for the result disrupt the surgical workflow. According to our own surgical experiences, we can say that optimal surgical therapy includes a combination of maximal resection of diseased tissue with minimal damage to the normal tissue. This can be achieved only with the ability to identify intraoperative cellular structures and to differentiate tumor from normal functional tissue. This opinion assumes the need for new technological equipment integrated within new surgical concepts. Confocal laser endomicroscopy (CLE) is an imaging technique that provides microscopic information of tissue in real time. Instead of taking a tissue sample from a patient and bringing it under the microscope for analysis, these “optical biopsies” bring the microscope inside the patient through miniaturized fiberoptic probes and allow real-time histopathology with a resolution down to 1 mm. A clear visualization of the cytoarchitecture of the cell can be achieved with a 1000fold magnification. CLE was introduced in 2004 in gastroenterology as a supplement to the current standard endoscopy for performing optical biopsies. It is also currently becoming a useful diagnostic tool in urology, pulmonology, and gynecology. We investigated the use of CLE in our neurosurgical treatment to analyze the technical considerations. Furthermore, we investigated the integration of CLE in the neurosurgical daily workflow in the operation room as an ex vivo diagnostic module, and we tried to create an easy integration of the CLE technique into the neurosurgical daily routine using endoscopic or microscopic settings and providing immediate intraoperative histopathological diagnosis of the entire entity in real time.We also investigated the best conditions for an evaluation of CLE for in vivo diagnosis in different types and models of intracranial and intraspinal neoplasia. Our results are reported here.

XRelevanceCAM: towards explainable tissue characterization with improved localisation of pathological structures in probe-based confocal laser endomicroscopy

Context aware decision support in neurosurgical oncology based on an efficient classification of endomicroscopic data

Analysis of Intraoperative Microscopy Imaging Techniques and Their Future Applications

High-Resolution Image Processing of Probe-Based Confocal Laser Endomicroscopy Based on Multistage Neural Networks and Cross-Channel Attention Module

Multi-label Recognition of Cancer-Related Lesions with Clinical Priors on White-Light Endoscopy

Towards an Efficient Computational Framework for Guiding Surgical Resection through Intra-operative Endo-microscopic Pathology.

Transfer Recurrent Feature Learning for Endomicroscopy Image Recognition.

Weakly Supervised Representation Learning for Endomicroscopy Image Analysis.

Imaging Breast Cancer Morphology Using Probe-Based Confocal Laser Endomicroscopy: Towards a Real-Time Intraoperative Imaging Tool for Cavity Scanning

Dynamic Tissue Surface Geometry Reconstruction for Autonomous Microscopic Tissue Scanning During Tumour Resection

Prospects for Theranostics in Neurosurgical Imaging: Empowering Confocal Laser Endomicroscopy Diagnostics via Deep Learning

Automatic Tissue Differentiation Based on Confocal Endomicroscopic Images for Intraoperative Guidance in Neurosurgery

Robust Mosaicing of Endomicroscopic Videos Via Context-Weighted Correlation Ratio.

Confocal Laser Endomicroscopy for Real-time Histomorphological Diagnosis: Our Clinical Experience With 150 Brain and Spinal Tumor Cases.

Weakly-Supervised Learning-Based Feature Localization in Confocal Laser Endomicroscopy Glioma Images

Few Shot Learning for the Classification of Confocal Laser Endomicroscopy Images of Head and Neck Tumors

Toward Robust Histology-Prior Embedding for Endomicroscopy Image Classification

Novel classification to facilitate recognition of breast cancer morphology on confocal endomicroscopy

3D multiplexed tissue imaging reconstruction and optimized region of interest (ROI) selection through deep learning model of channels embedding

Exploring the Explainability of a Machine Learning Model for Prostate Cancer: Do Lesions Localize with the Most Important Feature Maps?

Transferability of Deep Learning Algorithms for Malignancy Detection in Confocal Laser Endomicroscopy Images from Different Anatomical Locations of the Upper Gastrointestinal Tract