Feature driven classification of Raman spectra for real-time spectral brain tumour diagnosis using sound.

Spectroscopic diagnostics have been shown to be an effective tool for the analysis and discrimination of disease states from human tissue. Furthermore, Raman spectroscopic probes are of particular interest as they allow for in vivo spectroscopic diagnostics, for tasks such as the identification of tumour margins during surgery. In this study, we investigate a feature-driven approach to the classification of metastatic brain cancer, glioblastoma (GB) and non-cancer from tissue samples, and we provide a real-time feedback method for endoscopic diagnostics using sound. To do this, we first evaluate the sensitivity and specificity of three classifiers (SVM, KNN and LDA), when trained with both sub-band spectral features and principal components taken directly from Raman spectra. We demonstrate that the feature extraction approach provides an increase in classification accuracy of 26.25% for SVM and 25% for KNN. We then discuss the molecular assignment of the most salient sub-bands in the dataset. The most salient sub-band features are mapped to parameters of a frequency modulation (FM) synthesizer in order to generate audio clips from each tissue sample. Based on the properties of the sub-band features, the synthesizer was able to maintain similar sound timbres within the disease classes and provide different timbres between disease classes. This was reinforced via listening tests, in which participants were able to discriminate between classes with mean classification accuracy of 71.1%. Providing intuitive feedback via sound frees the surgeons' visual attention to remain on the patient, allowing for greater control over diagnostic and surgical tools during surgery, and thus promoting clinical translation of spectroscopic diagnostics.

Characterisation of a fibre optic Raman probe within a hypodermic needle.

We demonstrate the first use of a multifibre Raman probe that fits inside the bore of a hypodermic needle. A Raman probe containing multiple collection fibres provides improved signal collection efficiency in biological samples compared with a previous two-fibre design. Furthermore, probe performance (signal-to-noise ratios) compared favourably with the performance achieved in previous Raman microscope experiments able to distinguish between benign lymph nodes, primary malignancies in lymph nodes and secondary malignancies in lymph nodes. The experimental measurements presented here give an indication of the sampling volume of the Raman needle probe in lymphoid tissues. Liquid tissue phantoms were used that contained scattering medium encompassing a range of scattering properties similar to those of a variety of tissue types, including lymph node tissues. To validate the appropriateness of the phantoms, the sampling depth of the probe was also measured in excised lymph node tissue. More than 50 % of Raman photons collected were found to originate from between the tip of the needle and a depth of 500 µm into the tissue. The needle probe presented here achieves spectral quality comparable to that in numerous studies previously demonstrating Raman disease discrimination. It is expected that this approach could achieve targeted subcutaneous tissue measurements and be viable for use for the in vivo Raman diagnostics of solid organs located within a few centimetres below the skin's surface. Graphical Abstract Schematic of multi-fibre Raman needle probe with disposible tips and proximal optical filtration.

Rapid discrimination of maggots utilising ATR-FTIR spectroscopy.

Entomological evidence is used in forensic investigations to indicate time since death. The species and age of maggots or flies that are present at the scene can be used when estimating how much time has passed since death. Current methods that are used to identify species and developmental stage of larvae and fly samples are highly subjective, costly and often time consuming processes and require the expertise of an entomologist or species identification via DNA analysis. The use of vibrational spectroscopy, as an alternative identification method, would allow for a quicker, cheaper and less subjective technique and would allow entomological evidence to be used more commonly in the forensic process. This proof of principle study shows the potential for using attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR) as a rapid tool for differentiating between various species of larvae, such as those commonly found at crime scenes. The proposed regime would provide a rapid and valuable tool resulting in reduced time for both species identification and life cycle determination, particularly in forensic situations.

Effect of substrate choice and tissue type on tissue preparation for spectral histopathology by Raman microspectroscopy.

Raman spectroscopy is a non-destructive, non-invasive, rapid and economical technique which has the potential to be an excellent method for the diagnosis of cancer and understanding disease progression through retrospective studies of archived tissue samples. Historically, biobanks are generally comprised of formalin fixed paraffin preserved tissue and as a result these specimens are often used in spectroscopic research. Tissue in this state has to be dewaxed prior to Raman analysis to reduce paraffin contributions in the spectra. However, although the procedures are derived from histopathological clinical practice, the efficacy of the dewaxing procedures that are currently employed is questionable. Ineffective removal of paraffin results in corruption of the spectra and previous experiments have shown that the efficacy can depend on the dewaxing medium and processing time. The aim of this study was to investigate the influence of commonly used spectroscopic substrates (CaF2, Spectrosil quartz and low-E slides) and the influence of different histological tissue types (normal, cancerous and metastatic) on tissue preparation and to assess their use for spectral histopathology. Results show that CaF2 followed by Spectrosil contribute the least to the spectral background. However, both substrates retain paraffin after dewaxing. Low-E substrates, which exhibit the most intense spectral background, do not retain wax and resulting spectra are not affected by paraffin peaks. We also show a disparity in paraffin retention depending upon the histological identity of the tissue with abnormal tissue retaining more paraffin than normal.