Differentiation between nerve and adipose tissue using wide‐band spectroscopy The classification performance for respectively LOO cross‐validation and TT cross‐validation are listed in Table 9.3. LOO cross‐validation is solely based on InGaAs‐sensor range data from thyroid and parathyroid surgery for train and test purposes. TT cross‐validation is based on data from thyroid and parathyroid surgery for train purposes and 139 on additional carpal tunnel release surgery data for test purposes. Table 9.3 Classification performance of selected InGaAs‐sensor features TP TN Sensitivity Specificity PPV NPV Accuracy LOO CV 22/22 16/18 100 (82‐100) 89 (64‐98) 92 (72‐99) 100 (76‐100) 95 TT CV 10/10 5/5 100 (66‐100) 100 (46‐100) 100 (66‐100) 100 (46‐100) 100 TP = true positive; TN = true negative numbers indicate identified tissue spots. A positive test is defined as the tissue observed being nerve; a negative test is defined as the tissue observed being adipose tissue. Sensitivity; specificity; PPV = positive predictive value; NPV = negative predictive value; accuracy numbers are percentages; numbers in parentheses indicate 95% confidence interval. LOO CV = leave‐one‐out cross– validation; TT CV = train‐test cross‐validation. Discussion and conclusion This explorative study reveals the in vivo wide‐band (350 – 1830 nm) diffuse reflectance spectra of human recurrent laryngeal nerve, surrounding adipose tissue and sternocleidomastoid muscle (anatomical region: neck). Also in vivo spectra are presented for human median nerve, subcutaneous adipose tissue and transverse thenar/hypothenar muscle fibres overlying the transverse carpal ligament (anatomical region: wrist). The presented spectra covered both silicon (Si) and indium gallium arsenide (InGaAs) sensor ranges (350 – 1830 nm with 1 nm resolution), thereby exceeding beyond the 1600 nm boundary reported by preceding work21‐23,32. Spectroscopy on human skin samples in the wavelength range of 1000 – 2200 nm has been reported33. However, this report did not include any of the tissues covered by our study. Even though the measured reflectance spectra, shown in Figure 9.3, illustrate a great similarity between different tissue types, the nerve classification within an adipose surrounding was fairly accurate within the detection ranges of both Si and InGaAs. Based on the classification accuracies for these detection ranges (Tables 9.2 and 9.3) we can conclude that InGaAs sensors are better suited for automated discrimination between nerves and surrounding adipose tissue than Si sensors. With respect to the statistically different spectral features of nerve versus adipose tissue in the InGaAs range, our results indicate that especially water and lipid absorption differences around 1210 nm play an important role. This is also strongly
proefschrift_Schols_SLV
To see the actual publication please follow the link above