Differentiation Between Functional Properties And Microscopic Structures Of Peripheral Nerve Fibers And Fascicles With Raman Spectroscopy And Hyperspectral Imaging Technology

Research backgroundPeripheral nerve injuries are frequently encountered during clinical practice. The repair and regeneration of peripheral nerves have been hot research topics both domestic and abroad. Rapid intraoperative identification of motor and sensory fascicles is an urgent issue that needs to be tackled as correctly matching the nerve stumps is essential for the restoration of nerve function. It is also of much referential value to surgeons for it helps to make optimal decisions during peripheral nerve surgeries.Traditionally, methods adopted for clinical use were (1) intraoperative exploration and anatomical approach,(2) intraoperative arousal and electric current stimulation,(3) histochemical staining of cholinesterase or carbohydrase,(4) measurement of choline acetyltransferase activity.Recently, some researchers suggested that spectral analysis was another possible method to differentiate between the functional properties of peripheral nerve fascicles. Raman spectroscopy is capable of describing the amount of nucleic acids, proteins and lipids that compose tiny biological samples, as well as their properties and changes in the chemical structure. It is fast, non-invasive, high in sensitivity and resolution, and capable of quantitative analysis. Hyperspectral imaging microscopy can acquire both spatial and spectral information from tissue samples. It is an exceptional method for analyzing qualitatively, quantitatively, and also spatially. Up till now, many literatures have been published on successfully distinguishing between cells and tissues using Raman spectral and hyperspectral imaging analysis across multiple platforms such as histology, pathology, pharmacology and even traditional Chinese medicine.Objectives1. To study the Raman spectra of motor and sensory nerve fascicles of peripheral nerves.2. To study the modifications in the Raman spectra of motor and sensory nerve fascicles after Karnovsky-Roots staining.3. To evaluate the reliability and feasibility of achieving rapid differentiation between functions of motor and sensory nerve fibers using Raman analysis on nerve fascicles after being Karnovsky-Roots stained for30minutes. 4. To study the hyperspectral imaging properties of peripheral nerve sections.5. Discuss the feasibility of recognition and differentiation between the microscopic structures of peripheral nerves with computer software and histochemically assisted hyperspectal imaging analysis.Methods1.20New Zealand Rabbits were sacrificed for the harvest of their S1anterior and posterior spinal nerve roots.1human common peroneal nerve sample was collected from a case of amputation due to some disease other than ischemic necrosis. Samples were stored under-80℃and frozen sectioned under-20℃.30μm thick sections of rabbit samples were used in Raman spectroscopy experiments while10μm thick sections of both rabbit and human samples were used in hyperspectral imaging expertiments.2. Sections were made from ten randomly chosen samples from both anterior and posterior spinal nerve root groups. Raman spectra data of nerve fascicles on each slide was collected. Raman intensities and their ratios were statistically analyzed.3. Four continuous sections were made from of each of the three randomly chosen samples from both anterior and posterior spinal nerve root groups. Raman spectra data of nerve fascicles on each slide was collected directly after sectioning and after one of four types of processing. Such processing included Karnovsky-Roots staining for8hours or30minutes and incubating in0.1M Acetylthiocholine iodide solutions or0.05%Neostigmine injections, under four distinct groups. This was to study the modifications to the original Raman spectra of the fascicles.4. Sections were made from twenty anterior and twenty posterior root samples. Raman spectra data of nerve fascicles on each slide was collected after Karnovsky-Roots staining for30minutes. This was to statistically analyze the modification of the Raman spectra under quick staining.5. Sections were chosen from newly acquired samples from twenty anterior and twenty posterior spinal nerve roots. Raman spectra data of nerve fascicles on each slide was collected after Karnovsky-Roots staining for30minutes. Diagnostic test was performed to describe the feasibility of such method of differentiation involving Raman spectroscopy assisted with short Karnovksy-Roots staining.6. Sections were made from three randomly chosen samples from both anterior and posterior spinal nerve root groups and from both rabbit and human origins for direct molecular hyperspectral imaging microscopy. Both x20and x40images were captured/Fifteen regions of interests of axons or myelin sheaths were manually selected from each image and calculated for their characteristic spectral plots.7. Differentiated myelin sheaths from other microscopic structures via Spectral Angle Mapper in ENVI software and synthesized false-color pictures for direct viewing.8. Sections were made from three randomly chosen samples from both anterior and posterior spinal nerve root groups and stained with Karnovsky-Roots method for molecular hyperspectral imaging microscopy. Fifteen regions of interests of axons were manually selected from each slide and calculated for their characteristic spectral plots after staining. Modifications of spectral plots were discussed.Results1. Peripheral nerve fascicles have conspicuous Raman scatterings near550cm-1,1080cm-1,1110cm-1,1280cm-1,1440cm-1, and1660cm-1, with intensities at1440cm-1being strongest. The difference between motor and sensory spectra is non-significant.2. Additional Raman scatterings could be detected near480cm-1,510cm-1,595cm-1,2110cm-1, and2155cm-1in motor fascicles after8hours of Karnvosky-Roots staining. These additional scatterings were less significant after30minutes of staining yet remained visible at2100cm-1～2160cm-1. No such change would apply for sensory fascicles. The value of I2100/I1440in motor fascicles are significantly different (P<0.001) than in sensory fascicles after staining for30minutes. Staining can be categorized as positive if I2100/I1440>0.2.3. Differentiation between motor and sensory nerve fascicles could be achieved via the detection of2100cm-1～2160cm-1Raman scatterings after Karnovsky-Roots staining for30minutes.20anterior roots and19posterior roots were correctly classified while1posterior root resulted as false positive.4. Axons and myelin sheaths of peripheral nerves had distinct spectral plots under hyperspectral imaging microscopy. The degree of magnification and the species that were sampled from had no major impact on their manifestations. Differences between motor and sensory spectra were minimal.5. Myelin sheaths were differentiated from other microscopic structures in a peripheral nerve section via Spectral Angle Mapper when a=0.10. False-colored images were synthesized for direct viewing. Such method is unable to differentiate between motor and nerve fascicles directly and also lacks repeatability even after staining.Conclusions1. Both motor and sensory nerve fascicles manifest Raman spectrum of similar characteristics and are unable to differentiate directly.2. The two types of nerve fascicles can be differentiated from their counterparts in less time utilizing Raman micro-spectroscopy assisted with30minutes Karnovsky-Roots staining than conventional staining and light microscopy. The cut-off point in this paper was I2100/I1440=0.2. Value above0.2was considered positively stained. Such method had100%sensitivity,95%specificity and97.5%accuracy when identifying motor fascicles in this paper.3. Myelin sheaths have distincted spectral plots under hyperspectral imaging microscopy which can be differentiated from others via Spectral Angle Mapper. False-color images can be synthesized.4. Molecular hyperspectral imaging microscopy is unable to differentiate between motor and nerve fascicles. Its ability to differentiate after staining lacks repeatability and would require further research and experimentation.