High precision in the manual control of needles and biopsy probes in medical treatment requires high skill and dexterity levels. In anaesthesia, force sensation is an important feedback mechanism, and the practitioner needs to refresh or develop skills to improve on the interpretation of needle progress towards the target site. This paper describes an experimental tactile force simulator for uniaxial needle action for which the force resisting progress of the needle is derived from measured data. As an example, the approach taken to develop the simulation of the insertion of epidural needles is described. Adaptation to other procedures would be possible by adopting new reference models based on appropriate measured force data.

We have studied 150 women undergoing elective Caesarean section under spinal anaesthesia. They were allocated randomly to have a 22-gauge Whitacre, a 25-gauge Whitacre or a 26-gauge Quincke needle inserted into the lumbar subarachnoid space. The groups were compared for ease of insertion, number of attempted needle insertions before identification of cerebrospinal fluid, quality of subsequent analgesia and incidence of postoperative complications. There were differences between groups, but they did not reach statistical significance. Postdural puncture headache (PDPH) was experienced by one mother in the 22-gauge Whitacre group, none in the 25-gauge Whitacre group and five in the 26-gauge Quincke group. Five of the six PDPH occurred after a single successful needle insertion. Seven of the 15 mothers in whom more than two needle insertions were made experienced backache, compared with 12 of the 129 receiving two or less (P < 0.001). We conclude that the use of 22- and 25-gauge Whitacre needles in elective Caesarean section patients is associated with a low incidence of PDPH and that postoperative backache is more likely when more than two attempts are made to insert a spinal needle.

Precise control of automated invasive surgical tools requires real-time identification of tissue types and their deformation. At the focus of this paper is the epidural puncture, for which it is shown that the tissue type and deformation can respectively be determined from laser-based spectroscopy and the change in force required to push the needle through the various tissues. Studies have shown that physiological variations from one patient to another are too great to allow absolute values to be reliably used to indicate the position of the needle tip. However, the pattern of force variation during penetration is shown to be similar between specimens. Interpretation of this information in conjunction with spectroscopic techniques can be used to discriminate between tissues and tissue structure at the needle tip. This paper describes results from an investigation on automatic techniques for interpreting the type and deformation of tissues under tool action.

There have been several reports of cardiac output measurements during laparoscopy for gynaecological purposes, using dye dilution techniques. Techniques for measuring cardiac output which involve arterial and central venous cannulation did not seem appropriate to us in healthy patients undergoing a relatively minor surgical procedure. It was therefore decided to measure cardiac output during laparoscopy using a Minnesota Impedance C a r d i~g r a p h .~-~ This is a non-invasive technique which enables repeated measurements to be made at short time intervals. The values of stroke volume produced correlate satisfactorily with other methods currently and, in particular, with electromagnetic flowmeters' which also measure stroke volume on a beat by beat basis. Impedance cardiographyThis technique is based on the observation that if an alternating current is passed across the thorax it is possible to demonstrate that there is a change of electrical impedance with each cardiac cycle. A constant sinusoidal alternating current of 4 mA RMS at 100 kHz is passed between electrodes 1 and 4 and the transthoracic impedance is measured between the inner two electrodes 2 and 3 (Fig. I). These impedance changes with each cardiac cycle (Az) are differentiated to produce the trace - (Fig. 2).The peak of this trace represents the maximum rate of change of electrical impedance

We have compared two non-invasive methods of arterial pressure (AP) measurement used in labour wards: an automatic oscillometric measurement obtained by Dinamap 1846, and a conventional auscultatory measurement obtained by midwives. A total of 369 AP measurements were recorded, involving 28 normotensive and hypertensive pregnant women during labour, with or without extradural analgesia. Compared with the midwife group, the Dinamap group had a greater systolic AP, by 2.7 mm Hg (P < 0.01) and smaller diastolic AP, by 9.8 mm Hg (P < 0.01). The correlations between the two methods were highly significant, but the limits of agreement were relatively wide for both systolic and diastolic AP measurements. We conclude that a clinically important difference exists in diastolic AP measurements. Dinamap diastolic AP must be corrected using a regression equation, or simply by adding 10 mm Hg, before being compared with the available normal and hypertensive AP values.