1 PLANTEAMIENTO DEL PROBLEMA DE INVESTIGACION
1.3 OBJETIVOS DE LA INVESTIGACION
2.2.5 SISTEMAS DE PROTECCIÓN ACTIVA
2.2.5.3 Sistema Central de Alarma
2.2.5.3.3 Partes Básicas de una Central de Alarma
The study was approved by the local ethics commitee, and all patients and control subjects gave their informed consent for the magnetic resonance investigation. 3 groups of patients with breast cancer were studied using 3 ip MRS.
1) Patients presenting with untreated, operable breast cancer (UlCC Stage T1-4, NO-1,MO) were studied the day before surgery.
2) Patients not considered fit for surgery were treated with tamoxifen 20mg twice a day. They were studied before starting treatment and then at intervals of 1 week, 1 month, 3 months, 6 months and 1 year after starting treatment.
3) Patients not responding to tamoxifen were treated with surgery. They were studied the day before their operation.
3 ip MR spectra of normal breast were obtained from volunteers with no known breast pathology.
Localised 3 ip MR spectra of the breast were obtained on a Biospec spectrom eter interfaced to a I.QTesla, 60cm clear bore
superconducting magnet (Oxford Research Systems), operating at 32.7MHz for phosphorus and 80.8MHz for proton. The probe was a double surface coil with the 4 cm diameter receiver coil
electrically isolated from and positioned 1.5cm forward from the 10cm diameter transm itter coil. A glass vial of 75mM
diphenylphosphate (DPP) in chloroform was placed in the centre of the receiver coil as a reference for the position of the coil in the
phosphorus image.
Subjects were positioned supine and slightly rotated to one side so that the tumour was within the homogenous region of the magnet. The back was supported by foam to maintain the position and for patient comfort. In all patients the tum our was located very easily by palpation. The coil centre was placed over the tumour ensuring that the plane of the coil was parallel to the chest wall, and that the coil axis and therefore the B i field were perpendicular to the
Bo fie ld .
Both coils were tuned to the phosphorus frequency and the transm itter coil (receiver coil for later studies) for the proton frequency. The magnetic field homogeneity was optimised by
observing the proton signal from the region of breast to be studied. A pulse and collect sequence (48 pulses, 150|is, 3s interpulse
interval) was first recorded to select the appropriate centre frequency for obtaining depth resolved spectra.
Depth resolved spectra were obtained from the tumour separate from the underlying chest wall muscle using Phase Modulated
Rotating Frame Imaging (PMRFI) as originally proposed by Hoult (1979) and implemented as described by Blackledge et al., (1987). A total of 8 data accumulations were acquired, the pulse sequence comprising an incremented spatial encoding pulse (0) followed by a
phase encoding pulse (A.). After a short delay, the free induction decay was collected
0 ±x - Xy - delay - acquire Figure 2.1
Each accumulation comprised 48 transients, collected with an interpulse delay of 3s and covering a spectral width of 2000Hz. Using 100 watts of transmitter power, a pulse length of IBOps for both incremental and phase encoding pulse produced an
approxim ately 90° tip angle in the centre of the tumour. To remove the phase twist inherent in the 2-D fourier transform ation, a
second data set with the phase of the 0 pulse increased by IBOO was acquired in exactly the same way.
As there is an inverse relationship between spatial resolution and signal to noise ratio, the number of increments was the minimum required for adequate spatial discrimination. The phosphorus data was collected in 40 minutes. With positioning the patient and optimising the magnetic field, the total time of the study was about 1 hour, which was acceptable to all subjects.
The free induction decays were multiplied by an exponential line broadening of 15Hz in the chemical shift dimension (F2). 2-D Fourier transform ation was performed after zero filling and applying apodisation with a Gaussian function in the spatial dimension (F I). The 2 images obtained with 0x and 0_x were then added after reversing the second data set 0_x about the origin to produce a phophorus image without a phase twist. From the resulting 2-D data set, tumour spectra were selected by the
presence of PME (undetectable In muscle) and summed to maximise signal to noise. The poor signal to noise in individual rows from the contour plot did not allow accurate assessment of intratumour variations in the 3 ip m r spectrum.
A frequency domain, Lorentzian line fitting routine (developed by Counsell 0., MRC Radiobiology Unit, Chilton) was used to measure the area of the peaks in the 3 ip m r spectrum after optimising the baseline fit using polynomial corrections. The spectrum has an underlying hump that is due to the relatively immobile phospholipid bilayer with short T2 relaxation ( Murphy et al., 1989), and field inhomogeneity. The baseline was fitted to the hump rather than removed by convolution difference so as not to degrade the signal to noise. A mean of three fits was used to calculate PME relative to ATP using the ATP y-P signal as off resonance effects cause the ATP p-P peak to be displaced forward in the ^ ip m r spectroscopic
image (Styles, 1991). Consequently, the tumour spectrum contains a mixture of muscle and tumour ATP. The PDE peak was often
difficult to quantitate due to poor signal to noise, and a broad, non- Lorentzian line shape. Corrections for partial saturation effects were not made.
Measurement of absolute concentration of tissue metabolite levels can be made by relating the area of the peak of interest to the area of the phosphorus reference of known concentration, with
corrections for loss of signal intensity with depth. This was not possible as the tumour volume was not known and was often sm aller than the sensitive volume detected by the receiver coil.
2.2 HIGH RESOLUTION SPECTROSCOPY IN VITRO
Perchloric acid (PGA) extraction of tissue samples isolates the intermediates in phospholipid and energy metabolism and is therefore important for identification of the peaks in the 3 ip m r spectrum in vivo. Composition of the membrane phospholipids can be examined by chloroform/methanol (C/M) extraction of the pellet produced after PC A extraction.
2.2.1 EXTRACTION PROCEDURES
Samples of breast and brain tumours were frozen in liquid nitrogen within 15 minutes of removal at surgery until the time of
extraction. Samples of normal breast tissue were obtained from women undergoing reduction mammoplasty.