Imperial College London
Dr James McKenzie

DrJamesMcKenzie

Faculty of Medicine, Department of Metabolism, Digestion and Reproduction

Research Associate
 
 
 
//

Contact

 

j.mckenzie

 
 
//

Location

 

E311Burlington DanesHammersmith Campus

//

Summary

 

Publications

Citation

BibTex format

@article{St:2016:10.1016/j.ejso.2016.02.102,
author = {St, John E and Rossi, M and Balog, J and McKenzie, J and Muirhead, L and Speller, A and Gildea, L and Veselkov, K and Shousha, S and Ramakrishnan, R and Takats, Z and Darzi, A and Leff, D},
doi = {10.1016/j.ejso.2016.02.102},
journal = {European Journal of Surgical Oncology (EJSO)},
pages = {S25--S25},
title = {Real time intraoperative classification of breast tissue with the intelligent knife},
url = {http://dx.doi.org/10.1016/j.ejso.2016.02.102},
volume = {42},
year = {2016}
}
Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - Introduction: Re-operation for close/positive margins following breast-conserving surgery occurs frequently (≈20–25%), is cost-inefficient, and leads to physical and psychological morbidity. Rapid Evaporative Ionisation Mass Spectrometry (REIMS) determines the structural lipid profile of tissues by the on-line analysis of electrosurgical smoke and uses this information for the rapid identification of dissected tissues. We evaluated its performance regarding real-time classification of heterogeneous breast tissue.Method: 155 patients enrolled in this study comprising method optimization (n=40), construction of a tissue specific ex-vivo database (n=87), and intraoperative analysis (n=28). Electrosurgical aerosol produced from ex-vivo and in-vivo breast samples was aspirated into a mass spectrometer via a modified surgical hand-piece. Tissue identification results obtained by the multivariate statistical analysis of MS data were validated by histopathology. Intraoperative REIMS data was acquired from resection margins and time-synchronized to operative videos. Ex-vivo classification models were used to predict intraoperative margin status.Results: An ex-vivo classification model using spectral data from healthy breast (n=561) and breast tumours (n=139) provided 92.1% sensitivity, 96.4% specificity and 95.6% overall accuracy. 17,974 spectra from 28 patients were obtained intra-operatively in real-time. The method demonstrated 100% sensitivity and 77.3% specificity regarding intraoperative positive margin detection. For histologically negative margins, misclassification (REIMS false positive) was observed for only 0.5% of total tissue dissection time (69/14,023 spectra).Conclusions: The REIMS method has been optimized for real-time analysis of intraoperative heterogeneous breast tissue, and the results suggest spectral analysis is accurate and rapid for determination of oncological margin status intra-operatively as an “Intelligent Knife”.
AU  - St,John E
AU  - Rossi,M
AU  - Balog,J
AU  - McKenzie,J
AU  - Muirhead,L
AU  - Speller,A
AU  - Gildea,L
AU  - Veselkov,K
AU  - Shousha,S
AU  - Ramakrishnan,R
AU  - Takats,Z
AU  - Darzi,A
AU  - Leff,D
DO  - 10.1016/j.ejso.2016.02.102
EP  - 25
PY  - 2016///
SN  - 0748-7983
SP  - 25
TI  - Real time intraoperative classification of breast tissue with the intelligent knife
T2  - European Journal of Surgical Oncology (EJSO)
UR  - http://dx.doi.org/10.1016/j.ejso.2016.02.102
VL  - 42
ER  -
Download