Optical Geometry;

The proximal ends of the fibres are connected to the lamp and spectrometer by SMA connectors. A number of short pulses of light (typically 5-10) are sent through the delivery fibre. Following elastic scattering as described above, only a small fraction of the scattered light from the tissue reaches the receive fibre. The collected light is then guided to the spectrometer where an optical spectrum is generated for further processing.

The core diameters of the ’delivery’ and ‘receive’ fibres are 400pm and 200pm respectively. Both fibres lie adjacent to each other and at the distal end are encased in a metal sheath. The centre-to-centre fibre separation within this sheath is set at 340 pm. The external diameter of the metal casing in different probes varies between 0.8mm and 4mm depending on the situation that optical readings are being taken. For this probe geometry, the volume of tissue visited by the collected photons has been postulated to occupy a zone approximately 500 pm long, 300 pm wide and 300 pm deep. This has been determined from computational simulations using a Monte-Carlo code, which incorporates Mie theory for the details of the scattering events {Mourant, Boyer, et al. 1996 861 /id}. The separation of the fibres also has a profound bearing on the character of the received signal. For narrow separation, as with this design, the reflectance spectrum depends mainly on the scattering properties of tissue although some absorptive features also influence the signal. For a probe design with wider separation, the reverse is true.

The geometry described above has been developed and used for several years by our co-workers in Los Alamos. The sizes of the send and receive fibres are partly determined by the constraints of the connections to the xenon lamp and spectrometer and also have been selected for their broadband light transmission over the spectral range used in this study. The precise diameters have however been optimised to allow an adequate amount of light to enter the tissue but not to overpower the spectrometer with the received light. The separation of the fibres is also standardised by these fibre sizes when they are abutted together. This separation is conveniently a suitable distance apart to optimise signal due to the scattering properties of tissue

{Mourant, Fuselier, et al. 1997 426 /id}. Furthermore, by using such slender fibres, the overall diameter of the entire probe can be kept small enough to pass through narrow gauge cannula. It would be feasible to produce a system that has a single send and receive fibre, which could be made even smaller. This however would have limitations, as a significant proportion of the signal would be determined by reflection from the tissue surface. Also absorption would play a greater role and the technical considerations of rapid switching between send and receive modes would be difficult. For all these reasons, the current probe geometry offers the best compromise between size, simplicity of construction, and predominance of scattering within the returning signal.

8.3 Cl in i c a l p r o b e d e s ig n

Several methods of collecting data were investigated (see section 9.2). Preliminary ex-vivo work used hand held probes approximately 4mm in diameter but it was obvious that it would be necessary to produce probes of much smaller calibre in order to access tissue within the breast. Although it would be feasible to take in-vivo optical measurements from the cut surface of benign breast tumours during an operation, this is not possible for breast cancers. It is a strict surgical edict, to where possible, not incise through breast cancer tissue as this can cause seeding of cancer cells to the surgical wound or deep within the tumour bed.

As it has been speculated that the probe was only obtaining information from tissue to a depth of 1mm from its tip (see above), the main difficulty was to obtain a tissue sample for histological analysis from precisely the same point. It was therefore evident that a probe capable of passing through a standard biopsy needle would need to be developed so an initial optical reading could be obtained before taking a standard tissue biopsy for comparison. It was therefore decided to develop a probe small enough to pass through a standard 14-gauge core-cut biopsy needle. In conjunction with the collaborators in Los Alamos, an initial flexible probe was built with a diameter of 0.8mm. This was the smallest possible diameter due to the size of the optical fibres within the probe. It was found however that this allowed too much blood to gather within the bore of the core-cut needle which interfered with the optical reading due to excessive absorption by haemoglobin (Fig 8.8). A metal clad

optical probe was subsequently m anufactured with a diam eter o f 1.85mm. This created a snug fit within the bore o f the outer part o f the core-cut needle, which m inim ised the am ount a blood present at the tip giving better optical readings (Fig 8.4).

(a)

(b)

Figure 8.4 R igid probe (b) f o r using w ith core cut needle (a), a nd fle x ib le probes (c)

A second new dem and on the design o f the probe w as the ability to w ithstand adequate sterilisation. Initial in-vivo m easurem ents used probes sterilised with gluteraldehyde in a m anner sim ilar to that used for endoscopic instrum ents. The main draw back how ever was that the adhesive used at the SM A connector end w ould eventually degrade under these conditions so therefore only the distal end could be sterilised. It was felt how ever that a fully autolclavable probe w ould be beneficial as the w hole probe could be packaged and sterilised sim ilar to other operating instrum ents. T his w ould allow the surgeon to handle the w hole probe and then ‘hand o f f the SM A connectors to be attached to the rest o f the optical biopsy system. Several probe designs were tested. Initial problem s with partial m elting of the outer plastic sheath were overcom e by using a polyam ide coating. T his allow ed sterilisation at I26°C for Ilm in s , w ithout dam age. T his is a standard regim e for sim ilar instrum ents used during laparoscopy and w as im plem ented follow ing

direction from the sterile services department at the Middlesex Hospital. This inevitably has lead to a more reproducible and verifiable sterilisation process. Another sterilisation method that has been employed more recently is the use of ethylene oxide. This avoids high temperatures that may have a long-term effect on the probes. Both techniques are however adequate to fulfil ‘operating theatre standards’.