Novel handheld magnetometer probe based on magnetic tunnelling junction sensors for intraoperative sentinel lymph node identification

Abstract

Using magnetic tunnelling junction sensors, a novel magnetometer probe for the identification of the sentinel lymph node using magnetic tracers was developed. Probe performance was characterised in vitro and validated in a preclinical swine model. Compared to conventional gamma probes, the magnetometer probe showed excellent spatial resolution of 4.0 mm, and the potential to detect as few as 5 μg of magnetic tracer. Due to the high sensitivity of the magnetometer, all first-tier nodes were identified in the preclinical experiments, and there were no instances of false positive or false negative detection. Furthermore, these preliminary data encourage the application of the magnetometer probe for use in more complex lymphatic environments, such as in gastrointestinal cancers, where the sentinel node is often in close proximity to other non-sentinel nodes, and high spatial resolution detection is required.

Figure 1. MTJ magnetometer probe schematic diagram.

(a) The MTJ sensor bridge is mounted on the end of a small double-sided PCB, and positioned inside a cylindrical electromagnet’s aperture, forming the probe tip. The electromagnet consisted of a 90-turn coil wound onto a ferrous core and was used to excite magnetic nanoparticles in either parallel or antiparallel magnetisation with respect to the sensing axis of the probe. (b) Due to the hollow core of the bobbin-shaped electromagnet, from finite element analysis, the magnetic signal can be seen to peak at approximately 2.5 mm from the probe tip, before decaying in a manner similar to that of solid-core electromagnets.

Figure 2. Probe limit of detection.

By measuring a range of magnetic tracer quantities at a fixed distance of 4.0 mm from the signal source, a 50 μg limit of detection (SNR = 2.0) was calculated. Due to the short-range nature of magnetic fields, this value can be significantly improved if the probe is moved closer to the signal source.  = Measured data;  = Calculated datum.

Figure 3. Range of magnetic signal.

The magnetic signal from a simulated lymph node approximately 6 mm in diameter decays at a lower rate than predicted by theoretical inverse-cubed law for magnetic dipoles due predominantly to the finite volume of the phantom.

Figure 4. Spatial resolution of probe.

Normalised response curve of a point-source phantom (centred at x = 0) scanned at a height of 1.0 mm. A probe spatial resolution of 4.0 mm can be determined from the response curve’s FWHM, which is considerably less than that of a coil magnetometer or gamma probe.

Figure 5. ‘Sentinel’ node identification in vivo.

(a) Preinjection and postinjection coronal MRI scans of a swine’s hind legs showing the negative contrast in one first-tier node (circled) resulting from the uptake of the magnetic tracer. Postinjection scans were used preoperatively to identify the location and number of first-tier nodes in both limbs for each animal. (b) After the anatomical location of the nodes had been determined with MRI, Patent Blue V dye was used to guide the surgery, and the identification of first-tier nodes (circled). (c) Some nodes were only visualised due to the dark brown staining from magnetic tracer accumulation (top). Compared to a control lymph node (bottom left), either blue dye (bottom right) or magnetic tracer uptake was sufficient to distinguish the nodes from surrounding tissue.

Figure 6. Swine lymph node measurements.

(a) MTJ magnetometer probe measurement of first-tier nodes from three animals. N1, N2, and control (Cont.) were measured ex vivo, and N3 – N6b were measured in vivo. Error bars indicate the uncertainty in measurement due to background signal fluctuations (electronic noise). (b) Distinguishing nodes in close proximity in ex vivo array of lymph nodes. The boundary of each node is indicated with the vertical dotted lines and the threshold level is indicated with the horizontal dotted line. (c) Prussian blue staining of two separate node sections. The presence of magnetic tracer in afferent trabecula and subcapsular sinuses is evident in positive first- and second-tier nodes (e.g. top), but absent in negative nodes (e.g. bottom).