Chapter 11 points, i.e. (DR2‐DR1). DR = diffuse reflectance; = wavelength. The features used for classification are potentially significant due to differences in chemical composition (e.g. haemoglobin, water and lipid content) of the investigated tissues and structures. This study shows that Si‐ and InGaAs‐sensors (i.e. the spectral detection ranges which were used in the spectral data analysis, as outlined above) are equally suited for automated classification of ureter versus surrounding adipose tissue. Si‐sensors seem better suited for classifying artery versus mesenteric adipose tissue. Also, intraoperative nerve localization is of great importance in surgery. In certain procedures, where nerves show visual resemblance to surrounding adipose tissue, this can be particularly challenging for the human eye. A camera system, enabling nerve‐specific 170 image enhancement, would be useful. Chapter 9 identifies InGaAs‐sensors as better suited for automated discrimination between nerves and surrounding adipose tissue than Si‐sensors. This is probably based on the different ratio in which the endogenous chromophores water and lipid are present in respectively nerve and adipose tissue. Selective enhancement of nerves versus surrounding adipose tissue can be expected to be beneficial within all surgical disciplines. In thyroid and parathyroid surgery iatrogenic parathyroid injury should be prevented. Detecting the small sized parathyroid glands can be challenging and time consuming. In Chapter 10, automated parathyroid differentiation is investigated. Si‐ and InGaAs‐sensors seem fairly accurate for automated tissue classification of parathyroid versus adipose tissue or adjacent thyroid tissue. However, clinically relevant accuracy levels were not established for the single sensor types. The combination of the two sensor technologies significantly improves accuracy, especially regarding parathyroid‐adipose tissue classification. Given the results obtained by the several explorative studies using fibre‐optic DRS in this thesis, the potential for automated tissue classification is underlined. Therefore, further investigation of hyperspectral enhanced surgical imaging certainly looks worthwhile. The intraoperative visual judgment of a surgeon is not solely based on color (spectral) information, but also relies on the recognition of spatial anatomical position of a specific tissue. Therefore, the spot‐wise probe measurements (i.e. the “ground‐truth” reference spectra of specific tissue types) performed in the explorative studies reported in this thesis, should be considered just a first step towards the clinically more relevant technique of hyperspectral imaging of the whole surgical field (i.e. offering tissue‐specific contrast‐enhancement).
proefschrift_Schols_SLV
To see the actual publication please follow the link above