Development of an Electrical Impedance Tomography Coupled Surgical Stapler for Tissue Characterization.

OBJECTIVE: This study explores the feasibility of coupling Electrical Impedance Tomography (EIT) to a standard-of-care laparoscopic surgical stapler, stapler+EIT, with the long-term goal of enabling intraoperative tissue differentiation for tumor margin detection. METHODS: Two custom printed-circuit-board-based electrode arrays with 60 and 8 electrodes, respectively, matching the stapler geometry, served as the jaws of an electrode-integrated surrogate stapler+EIT device. The device was evaluated through a series of simulations and bench-top imaging experiments of agar-gel phantoms and bovine tissue samples to evaluate the technique and determine optimal imaging parameters. RESULTS: Electrodes localized to only one jaw (the 60-electrode side) of the stapler outperformed a dual-jaw distribution of electrodes. Using this one-sided electrode array, reconstructions achieved an Area-Under-the-Curve (AUC) ≥ 0.94 for inclusions with conductivity contrasts of ≥30% for any size considered on measured agar-gel tests, and an AUC of 0.80 for bovine tissue samples. CONCLUSION: A stapler+EIT algorithm has been tuned and evaluated on challenging phantom tests and demonstrated to produce accurate reconstructions. SIGNIFICANCE: This work is an important step in the development of a surgical stapler+EIT technique for margin assessment.

Electrical Impedance Characterization of in Vivo Porcine Tissue Using Machine Learning.

The incorporation of sensors onto the stapling platform has been investigated to overcome the disconnect in our understanding of tissue handling by surgical staplers. The goal of this study was to explore the feasibility of in vivo porcine tissue differentiation using bioimpedance data and machine learning methods. In vivo electrical impedance measurements were obtained in 7 young domestic pigs, using a logarithmic sweep of 50 points over a frequency range of 100 Hz to 1 MHz. Tissues studied included lung, liver, small bowel, colon, and stomach, which was further segmented into fundus, body, and antrum. The data was then parsed through MATLAB's classification learner to identify the best algorithm for tissue type differentiation. The most effective classification scheme was found to be cubic support vector machines with 86.96% accuracy. When fundus, body and antrum were aggregated together as stomach, the accuracy improved to 88.03%. The combination of stomach, small bowel, and colon together as GI tract improved accuracy to 99.79% using fine k nearest neighbors. The results suggest that bioimpedance data can be effectively used to differentiate tissue types in vivo. This study is one of the first that combines in vivo bioimpedance tissue data across multiple tissue types with machine learning methods.

Hydrogeology, chemical and microbial activity measurement through deep permafrost.

Little is known about hydrogeochemical conditions beneath thick permafrost, particularly in fractured crystalline rock, due to difficulty in accessing this environment. The purpose of this investigation was to develop methods to obtain physical, chemical, and microbial information about the subpermafrost environment from a surface-drilled borehole. Using a U-tube, gas and water samples were collected, along with temperature, pressure, and hydraulic conductivity measurements, 420 m below ground surface, within a 535 m long, angled borehole at High Lake, Nunavut, Canada, in an area with 460-m-thick permafrost. Piezometric head was well above the base of the permafrost, near land surface. Initial water samples were contaminated with drill fluid, with later samples <40% drill fluid. The salinity of the non-drill fluid component was <20,000 mg/L, had a Ca/Na ratio above 1, with δ(18) O values ∼5‰ lower than the local surface water. The fluid isotopic composition was affected by the permafrost-formation process. Nonbacteriogenic CH(4) was present and the sample location was within methane hydrate stability field. Sampling lines froze before uncontaminated samples from the subpermafrost environment could be obtained, yet the available time to obtain water samples was extended compared to previous studies. Temperature measurements collected from a distributed temperature sensor indicated that this issue can be overcome easily in the future. The lack of methanogenic CH(4) is consistent with the high sulfate concentrations observed in cores. The combined surface-drilled borehole/U-tube approach can provide a large amount of physical, chemical, and microbial data from the subpermafrost environment with few, controllable, sources of contamination.

An exploration of alternative visualisations of the basic helix-loop-helix protein interaction network.

BACKGROUND: Alternative representations of biochemical networks emphasise different aspects of the data and contribute to the understanding of complex biological systems. In this study we present a variety of automated methods for visualisation of a protein-protein interaction network, using the basic helix-loop-helix (bHLH) family of transcription factors as an example. RESULTS: Network representations that arrange nodes (proteins) according to either continuous or discrete information are investigated, revealing the existence of protein sub-families and the retention of interactions following gene duplication events. Methods of network visualisation in conjunction with a phylogenetic tree are presented, highlighting the evolutionary relationships between proteins, and clarifying the context of network hubs and interaction clusters. Finally, an optimisation technique is used to create a three-dimensional layout of the phylogenetic tree upon which the protein-protein interactions may be projected. CONCLUSION: We show that by incorporating secondary genomic, functional or phylogenetic information into network visualisation, it is possible to move beyond simple layout algorithms based on network topology towards more biologically meaningful representations. These new visualisations can give structure to complex networks and will greatly help in interpreting their evolutionary origins and functional implications. Three open source software packages (InterView, TVi and OptiMage) implementing our methods are available.