FIELD OF THE INVENTION

The invention relates to the field of magnetic resonance imaging (MRI), and more particularly to RF pluse amplifiers for RF pulse excitation in MRI systems. BACKGROUND OF THE INVENTION

RF power amplifiers are used in a wide variety of applications including wireless communication, TV transmissions, radar, RF heating, and medical devices. These RF power amplifiers must meet strict performance specifications, efficiency, output power and linearity dependent on different applications. The output signal of an ideal power amplifier is an exact replica of the input signal, only the amplitude is greater by some fixed multiple. Typically, the power amplifier increases a power level of a continuous wave input signal and consumes most of the power of the overall system. A power amplifier with a low efficiency results in increased power usage and heat removal requirements, which may increase the equipment and operating costs of the overall system. For this reason, much effort has been expended on increasing the efficiency of RF power amplifiers. Power amplifier architectures are developed with the aim to realize power stages with almost constant and higher efficiency behavior for a limited output back-off range, which is the amount of backoff from the peak power output, e.g., a typical 6db back-off from the peak power output. US2012/0315162A1 discloses a 3 -way Doherty amplfiier with minimum output network with improved efficiency in a 12dB power back-off demanded by new communication ssytems such as 3G-LTE. WO2012/076924 A 1 discloses another RF amp lifer circuit in

communication system that maintains a good matching in both low powr mode and high power mode.

By MRI systems, the nuclear spins of the body tissue to be examined are aligned by a static main magnetic field Bo and are excited by transverse magnetic fields Bi oscillating in the radio frequency band. The Bi field is produced by driving electrical currents through specialized RF transmit coils. In nearly all clinical MR imaging applications, the Bi field is transmitted in short 1-5 ms bursts called RF pluses, but not in continuous waves. A RF transmit chain including a frequency synthesizer, a pulse modulator and a pulse amplifier is responsible for generating the electric currents required to produce the Bi field. More specifically, the frequency synthesizer produces a continuous sinusoidal carrier wave at the Larmor frequency, which is sent to the pulse modulator for being shaped into amplitude- modulated pulses as dictated by a particular imaging application. Down the RF transmit chain, the RF pulse further passes through the pulse amplifier which raises the power level of the small input RF pulse up from a milli-Watt range to a level high enough to drive the RF coils. The RF pulse amplifiers in modern MRI systems typically produce peak power in a wide dynamic range of 0.5KW-35KW, e.g., 0.5KW-2KW (legs and arms), 4KW-8KW head, and up to 35KW whole body, at the frequency of 64MHZ or 128MHZ. To meet the high power output levels, many power amplifiers composed of the same type of amplifiers with the same power amplification ratings are typically deployed at the amplification stage, so divider and combiner networks are also used to divide the small input pulse evenly among the power amplifiers before the amplification and sum the amplifier output power together after the amplification.

However, the theoretically possible efficiency of the pulse amplifiers cannot cover the full dynamic range of the output power and this efficiency drops sharply when the output power is in a low amplitude range far away from the maximum output power. Fig. 1 indicates a maximum average RF power available for various peak output power of a 1.5T MRI system, e.g., Philips Achieva 1.5T MRI. The axial coordinate shows the peak output power of the pulse amplifier in the 1.5T MRI system, and the vertical coordinate shows the maximum average RF power in accordance with different peak output powers. The dotted line 101 indicates the short term maximum average power, e.g., the average power calculated in 2 seconds. The solid line 103 indicates the long term maximum average power, e.g., the average power calculated in 30 seconds. As shown by Fig. 1 , due to the poor performance in the low amplitude range below a knee point 105, e.g., 1KW, the pulse amplifier cannot deliver enough average RF power necessary for some clinical MRI applications. In order to maintain the same average RF power, higher power supply and more complicated cooling system need to be introduced, and the power consumption in this case may be doubled or even more, which will increase the total cost and failure rates of the RF pulse amplifiers. Therefore, pulse amplifiers in medical devices call for improved efficiency in lower amplitude range of RF pulses.

SUMMARY OF THE INVENTION It is an object of the invention to provide a RF pulse amplifier for magnetic resonance imaging that effectively mitigate or even eliminate the sharp efficiency drop at the low amplitude range of the input/output power.

Embodiments of the invention provide a RF pulse amplifier, a method for driving a transmit coil using the RF pulse amplifier, and a MRI system embedded with the RF pulse amplifier in the independent claims. Embodiments are given in the dependent claims.

Embodiment of the present invention provides a RF pulse amplifier for driving a transmit coil in a magnetic resonance imaging system. The RF pulse amplifier comprises a signal divider network, multiple auxiliary power chains ad a main power chain. The signal divider network is configured to divide an input signal into a main input signal and multiple auxiliary input signals. Each auxiliary power chain includes an auxiliary amplifier configured to amplify one of the auxiliary input signals to a respective auxiliary amplified signal. The main power chain includes a main amplifier and multiple combiners coupled in series between the signal divider network and the transmit coil. The combiners are further configured to receive the auxiliary amplified signals from the auxiliary amplifiers, respectively, and a control is coupled to the signal divider network and the main and auxiliary amplifiers and configured to adjust phases of the auxiliary amplified signal and operating points of the auxiliary amplifiers based on the input pulse signal to ensure at least part of the auxiliary amplifiers is switched off at a low amplitude range of the input pulse signal far from predetermined peak output of the RF pulse amplifier . Advantageously, compared with conventional pulse amplifiers used in magnetic resonant imaging systems, fewer amplifiers in the RF amplifier of the present invention are activated to operate in the low amplitude range of the output power, thereby the operating amplifiers are set closer to the optimum state that delivers the maximum output power with the maximum efficiency. In other words, the fewer operating amplifiers are more fully used on the optimal points, owing to which, the sharp efficiency drop in the low amplitude range of the output power is mitigated significantly or eliminated.

According to one embodiment of the present invention, each combiner further comprises an impedance transforming network and a combination node. The impedance transforming network is configured to transform a voltage of a signal flowing through the main chain and appearing at an input port of the impedance transforming network to an output current at an output port of the impedance inverting network. The combination node is configured to sum the output current from the output port of the impedance transforming network with the corresponding auxiliary amplified signal received from the corresponding auxiliary amplifier. Advantageously, discarding and/or adding the power output from the auxiliary power chains can be realized in each combination stage and controlled

independently according to output power requirements of the RF pulse amplifier.

According to another embodiment of the present invention, the output current from the output port of the impedance transforming network is independent of a load modulation driven by the corresponding auxiliary amplifier. Advantageously, discarding and/or adding the auxiliary amplified signals from the auxiliary amplifiers 408-1 to 408-N to the main power chain brings no impact on each other, thereby realizing isolated input of each combination stage of the (N+ 1 )-way combiner 416.

According to yet another embodiment of the present invention, the impedance transforming network is implemented in a form selected from at least a lumped T-shaped network, a lumped n-shaped network or a distributed lambda quarter transmission line.

According to yet another embodiment of the present invention, the RF pulse amplifier further comprises multiple phase compensators coupled in series with the auxiliary amplifiers in the auxiliary power chains, respectively. The phase compensators are

configured to adjust phases of the auxiliary amplified signals, respectively. Advantageously, phases of the auxiliary amplified signals can be adjusted to be in phase with the

corresponding output current at the combination node of each combination stage.

According to yet another embodiment of the present invention, a DC operating point of the main amplifier is preset to be in an active region of operation, and DC operating points of the auxiliary amplifiers are preset to be in a cutoff region of operation. A DC operating point of a next successively switched-on auxiliary amplifier is preset to further below a DC operating point of a preceding successively switched-on auxiliary amplifier. Advantageously, since only the main amplifier is preset to be in the cutoff region of operation, the power consumption of DC biasing is further reduced.

According to yet another embodiment of the present invention, the main amplifier is biased to operate in class AB mode and each auxiliary amplifier is biased to operate in class C mode. Advantageously, since only the main amplifier is biased to operate in class AB mode, the power consumption of DC biasing is further reduced.

According to yet another embodiment of the present invention, the RF pulse amplifier further comprises a controller coupled to the signal divider network and the main and auxiliary amplifiers. The controller is configured to adjust phases of the auxiliary amplified signal and operating points of the auxiliary amplifiers based on the input signal. Advantageously, the controller can adjust phases of the auxiliary amplified signals in phase with the corresponding the corresponding output current at the combination node of each combination stage. Moreover, the controller can adjust the operating points of the auxiliary amplifiers to ensure that as few as possible amplifiers are biased to switch on at the low amplitude range of the input/output power of the RF pulse amplifier, thereby significantly mitigate or even eliminate the sharp efficiency drop below the knee point.

According to yet another embodiment of the present invention, the auxiliary amplifiers begin to be switched on when the power level of the input signal increases to a predetermined threshold value, preferably 1/3-1/2 of the peak of the input signal.

Embodiment of the present invention provides a method for driving a transmit coil in a magnetic resonance imaging system. The method comprises the steps of dividing an input signal into a main input signal and multiple auxiliary input signals, providing the auxiliary input signals to multiple auxiliary amplifiers in multiple auxiliary power chains, respectively, adjusting operating points of the auxiliary amplifiers based on the input pulse signal by a controller 414) to provide at least as possible switched-on auxiliary amplifiers at at a low amplitude range of the input pusle signal far from predetermined peak output of the RF pulse amplifier, amplify the auxiliary input signals to auxiliary amplified signals by switched-on auxiliary amplifiers, respectively, adjusting phases of the auxiliary amplified signal by the controller based on the input pulse signal; providing the auxiliary amplified signals to multiple serially coupled combiners in a main power chain, respectively, amplifying the main input signal to a main amplified signal by a main amplifier in the main power chain, and combining the main amplified signal sequentially with the auxiliary amplified signals as the main amplified signal flows through the serially coupled combiners sequentially to generate an output signal that drives the transmit coil.

According to one embodiment of the present invention, the step of combining further comprises performing the following steps for each combination in sequence:

transforming a traveling signal appearing at an input port of an impedance transforming network to an output current at an output port of the impedance inverting network, and summing at a combination node the output current with the corresponding auxiliary amplified signal from the corresponding auxiliary amplifier.

According to another embodiment of the present invention, the method further comprises the step of adjusting the output current in phase with the corresponding auxiliary amplified signal at the combination node. According to yet another embodiment of the present invention, the method further comprises the steps of presetting a DC operating point of the main amplifier in an active region of operation, and presetting DC operating points of the auxiliary amplifiers in a cutoff region of operation. A DC operating point of a next successively switched-on auxiliary amplifier is further below a DC operating point of a preceding successively switched-on auxiliary amplifier.

According to yet another embodiment of the present invention, the method further comprises the steps of biasing the main amplifier to operate in class AB mode, and biasing the auxiliary amplifiers to operate in class C mode.

According to yet another embodiment of the present invention, the input signal is a RF pulse chain with a dynamic amplitude range, and the auxiliary amplifiers are switched on successively with increasing of the amplitude of the RF pulse chain.

Various aspects and features of the disclosure are described in further detail below. And other objects and advantages of the present invention will become more apparent and will be easily understood with reference to the description made in combination with the accompanying drawings.

DESCRIPTION OF THE DRAWINGS

The present invention will be described and explained hereinafter in more detail in combination with embodiments and with reference to the drawings, wherein:

Fig. 1 illustrates a maximum average RF power available for various peak output power of a 1.5T MRI system.

Fig. 2 illustrates a schematic diagram of a magnetic resonance imaging system according to one embodiment of the present invention.

- Fig. 3 illustrates a block diagram of a pulse amplifier according to one embodiment of the present invention.

Fig. 4 illustrates a schematic diagram of a pulse amplifier according to one embodiment of the present invention.

Fig. 5 illustrates a schematic diagram of a (N+l)-way combiner according to one embodiment of the present invention.

Fig. 6 illustrates a schematic diagram of a pulse amplifier according to another embodiment of the present invention.

Fig. 7 illustrate a flow chart of a method for driving a transmit coil in a magnetic resonance imaging system according to one embodiment of the present invention. The present invention will be described with respect to particular

embodiments and with reference to certain drawings but the invention is not limited thereto but only by the claims. The drawings described are only schematic and are non-limiting. In the drawings, the size of some of the elements may be exaggerated and not drawn on scale for illustrative purposes.

DETAILED DESCRIPTION OF THE EMBODIMENTS

Like numbered elements in these figures are either equivalent elements or perform the same function. Elements which have been discussed previously will not necessarily be discussed in later figures if the function is equivalent.

Fig. 2 illustrates a magnetic resonance imaging (MRI) system 200 that excites a nuclei (e.g., associated with isotopes such as IH, 19F, 13C, 31p, etc.) within a subject using a RF power amplifier. The system 200 includes a housing 4. A subject 6 (e.g., a human, an object, etc.) is at least partially disposed within a bore 8 of the housing 4 for one or more MRI procedures (e.g., spin echo, gradient echo, stimulated echo, etc.). A magnet 10 resides in the housing 4. The magnet 10 typically is a persistent superconducting magnet surrounded by a cryo shrouding 12. However, other known magnets (e.g., a resistive magnet, a permanent magnet, etc.) can be employed. The magnet 10 produces a stationary and substantially homogeneous main magnetic field B0 in the subject 6. As a result, the nuclei within the subject 6 preferentially align in a parallel and/or an anti-parallel direction with respect to the magnetic flux lines of the magnetic field B0. Typical magnetic field strengths are about 0.5 Tesla (0.5T), LOT, 1.5T, 3T or higher (e.g., about 7T).

Magnetic field gradient coils 14 are arranged in and/or on the housing 4. The coils 14 superimpose various magnetic field gradients G on the magnetic field B0 in order to define an imaging slice or volume and to otherwise spatially encode excited nuclei. Image data signals are produced by switching gradient fields in a controlled sequence by a gradient controller 16. One or more radio frequency (RF) coils or resonators are used for single and/or multi-nuclei excitation pulses within an imaging region. Suitable RF coils include a full body coil 18 located in the bore 8 of the system 2, a local coil (e.g., a head coil 20 surrounding a head of the subject 6), and/or one or more surface coils.

A transmitter 22 including a frequency synthesizer 24, a pulse modulator 26 and a pulse amplifier 28, generates the single and/or multi-nuclei excitation pulses and provides these pulses to the RF coils 18 and/or 20 through a T/R switch 44. A scanner controller 30 controls the transmitter 22 based on operator instructions. For instance, if an operator selects a protocol for acquisition of proton spectra, the scanner controller 30 accordingly instructs the transmitter 22 to generate excitation pulses at a corresponding frequency. The scanner controller 30 also controls the T/R switch 44. During an excitation phase, the scanner controller 30 controls the T/R switch 44 and allows the single or multi-nuclei excitation pulses to pass through the T/R switch 44 to the RF coils 18 or 20, but not to a receive system 32. Upon receiving the single or multi-nuclei excitation pulses, the RF coils 18 or 20 resonate and apply the pulses into the imaging region. The gradient controller 16 suitably operates the gradient coils 14 to spatially encode the resulting MR signals. During the readout phase, the T/R switch 44 connects the receive system 32 to one or more receive coils to acquire the spatially encoded MR signals. The receive system 32 includes one or more receivers 34 dependent on the receive coil configuration. The acquired MR signals are conveyed (serially and/or in parallel) through a data pipeline 36 and processed by a processing component 38 to produce one or more images.

The reconstructed images are stored in a storage component 40 and/or displayed on an interface 42, other display device, printed, communicated over a network (e.g., the Internet, a local area network (LAN), etc.), stored within a storage medium, and/or otherwise used. The interface 42 also allows an operator to control the magnetic resonance imaging scanner 2 through conveying instructions to the scanner controller 30.

Fig. 3 illustrates a block diagram of a pulse amplifier 300 according to one embodiment of the present invention. In the embodiment of Fig. 3, the pulse amplifier 300 includes a signal divider network 302, a main power chain 320 and multiple auxiliary power chains 330-1 to 330-N, where N is an integral greater than 1. The pulse amplifier 300 raises the power level of a small input RF pulse to a power level high enough to drive a transmit coil 310. The main power chain 320 further includes a main amplifier 304 and multiple combiners 306-1 to 306-N which are coupled in series between the signal divider network 302 and the transmit coil 310. The auxiliary power chains 330-1 to 330-N including auxiliary amplifiers 308-1 to 308-N are coupled between the signal divider network 302 and the combiners 306-1 to 306-N, respectively. Each auxiliary power chain is coupled in parallel with the main power chain. In operation, the signal divider network 302 divides the small input pulse into a main input signal provided to the main amplifier 304 and multiple auxiliary input signals provided to the auxiliary amplifiers 308-1 to 308-N, respectively. The main amplifier 304 amplifies the main input signal to a main amplified signal. The auxiliary amplifiers 308-1 to 308-N are suitably biased to be switched on or off based upon a power level of the input RF pulse. Upon switching-on of a certain auxiliary amplifier, e.g., the auxiliary amplifier 308-K, the auxiliary input signal received by the auxiliary amplifier 308- K is amplified to an auxiliary amplified signal by the switched-on auxiliary amplifier, where K is an integral selected from 1 to N. The auxiliary amplified signal from the auxiliary amplifier 308-K is further provided to the combiner 306-K. More specifically, with increasing of the power level of the input RF pulse signal, the auxiliary amplifiers 308-1 to 308-N are biased to be switched on successively to provide the auxiliary amplified signals. The main amplified signal traveling through the serially coupled combiners 306-1 to 306-N is combined sequentially with the auxiliary amplified signals to generate an output signal that drives the transmit coil 310.

As such, the serially coupled combiners 306-1 to 306-N realize a (N+l)-way combination in a chain configuration and each combination stage adds a certain share of the output power to the output. Furthermore, owing to successive switching-on of the auxiliary amplifiers 308-1 to 308-N, the power contribution from the auxiliary amplifiers 308-1 to 308- N, at least in part, is discarded in a low amplitude range of the input RF pulse but

successively added up as the amplitude of the input RF pulse exceeds some predetermined power thresholds successively. Since fewer amplifiers are activated to operate in the low amplitude range of the output power, the operating amplifiers are set closer to the optimum state that delivers the maximum output power with the maximum efficiency. In other words, the fewer operating amplifiers are more fully used. Accordingly, owing to the fully used fewer operating amplifiers, the sharp efficiency drop in the low amplitude range of the output power is mitigated significantly or eliminated.

Fig. 4 illustrates a schematic diagram of a RF pulse amplifier 400 according to one embodiment of the present invention. In the embodiment of Fig. 4, the pulse amplifier 400 includes a signal divider network 402, a main amplifier 404, auxiliary amplifiers 408-1 to 408-N, a controller 414 and a (N+l)-way combiner 416. The (N+l)-way combiner 416 further comprises N serially coupled combiners, each of which includes an impedance inverting network, e.g., impedance transforming network 406-1, 406-2,... 406-N, and a combination node, e.g., combination node 412-1, 412-2,... 412-N. The main amplifier 404 and the (N+l)-way combiner 416 form the main power chain, and the auxiliary amplifiers 408-1 to 408-N form N auxiliary power chains. In a fashion similar to the embodiment discussed with respect to Fig. 3, the pulse amplifier 400 raises the power level of a small input RF pulse to a power level high enough to drive a transmit coil 410. Description of certain similarities between the two embodiments may be omitted herein for the sake of brevity and convenience, but such omission is not limiting.

In the embodiment of Fig. 4, the controller 414 is incorporated to bias the main and auxiliary amplifiers 404 and 408-1 to 408-N to operate properly. The RF pulse amplifier 400 features differential bias controls for the main amplifier 404 and the auxiliary amplifiers 408-1 to 408-N. In one embodiment, the bias controls can be predetermined as static DC bias voltages for a particular implementation. For example, based on the predetermined static DC bias voltages, a DC operating point of the main amplifier 404 is preset to be in an active region of operation, e.g., biased in class AB mode, while DC operating points of the auxiliary amplifiers 408-1 to 408-N are preset to be in a cutoff region of operation and are different from each other, e.g., biased in class C mode. Since the main amplifier 404 is biased in the active region of operation, e.g., class AB mode, the main amplifier 404 can provide the main amplified signal to add output power to the main power chain as soon as the main input signal from the signal divider network 402 is provided to the main amplifier 404. While, cutoff region biasing of the auxiliary amplifiers 408-1 to 408-N, e.g., class C mode, causes a lag before the auxiliary amplifiers 408-1 to 408-N begin to add the output power to the main power chain. This is because in the cutoff biasing, the auxiliary amplifiers 408-1 to 408-N are turned-off until the auxiliary input signals are strong enough to turn on the respective auxiliary amplifiers. Due to the different DC operating points preset for the auxiliary amplifiers 408-1 to 408-N, the auxiliary amplifiers 408-1 to 408-N are turned on successively to provide the output power to the main power chain as the power level of the input signal increases.

Alternatively, the controller 414 can be preset with multiple power thresholds P _TH for the input RF pulse which are associated with the biasing controls that switch on the auxiliary amplifiers 408-1 to 408-N, respectively. As such, by monitoring the input RF pulse signal, the controller 414 successively turns on the auxiliary amplifiers 408-1 to 408-N as the input RF pulse exceeds the respective power thresholds. As discussed previously, different from the RF amplifiers used in constant envelope modulation communication system which emphasizes more on efficiency in a limited back-off range (e.g., 6 or 12dB back-off) from the peak output power, RF pulse amplifiers features dynamic wide range (e.g., 70dB) of operation and the input pulse is changing dynamically according to the MR scan sequence requirements, in which circumstance, the operation of the RF amplifier in the low power range far from the peak power needs to be optimized accordingly based on the dynamic input pulse, e.g., based on the power level, the pulse width and/or the pulse-width-modulaiton (PWM) of the MR sequence.

As observed, in accordance with the pulse amplifier configuration of Fig. 4, the auxiliary amplifiers 408-1 to 408-N, at least in part, are discarded at a low amplitude range of the input power/output power, but successively turned on to add output power to the main power chain as the input signal becomes stronger. More specifically, if the power level or magnitude of the RF input pulse signal is below a predetermined power threshold P _TH , e.g., 1/3-1/2 of the peak input power, only the main amplifier 404 is in operation to provide the output power that is conventionally provided by all partially used power amplifiers. As such, the operating state of the main amplifier 404 is set closer to the optimum state of the peak output power with maximum efficiency. Additionally, load modulation driven by the turned- off auxiliary amplifiers enables the main amplifier 404 to see increased load impedance, which leads to a peak efficiency of the main amplifiers 404 before it delivers maximum output power. Once the power level or magnitude of the RF input pulse signal exceeds the predetermined power threshold P _TH , the auxiliary amplifiers 308-1 and 308-N are

successively turned on to provide the required output power. For each auxiliary amplifier that is turned on to add the output power, there is a corresponding increase in the power range over which the high efficiency is maintained since the amplifiers are operating on the optimal points of high efficiency.

Advantageously, the efficiency of the RF pulse amplifier 400 at the low amplitude range of the input power/output power is significantly enhanced by full usage of fewer operating power amplifiers. Accordingly, the sharp efficiency drop as shown by Fig. 1 can be effectively mitigated or even eliminated. As such, the RF pulse amplifier 400 can meet various pulse sequence requirements of various clinical applications. Moreover, because only the main amplifier 404 is biased in the active region, e.g., biased in class AB mode, the power consumption of DC biasing is further reduced. These factors contribute to a significantly enhanced efficiency of the RF pulse amplifier 400.

In order to discard and/or add the auxiliary amplified signals from the auxiliary amplifiers 408-1 to 408-N without impact on each other, each combination stage of the (N+l)-way combiner 416 needs to have an isolated input. This requires that the resultant output current from each of the impedance transforming network 406-1 to 406-N is independent of the load modulation driven by the auxiliary amplifiers 408-1 to 408-N, respectively, both in amplitude and phase. To the end, each of the impedance inverting networks 406-1 to 406-N can be implemented in a form selected from at least a lumped T- shaped network, a lumped n-shaped network and a distributed lambda quarter transmission line.

Fig. 5 illustrates a schematic diagram of the (N+l)-way combiner 500 according to one embodiment of the present invention. In the embodiment of Fig. 5, the (N+l)-way combiner 500 includes N combiners in a serial configuration. Each of N combiners includes an impedance transforming network and a combination node, e.g., the impedance inverting networks 503-1 to 503-N and combination nodes 505-1 to 505-N. Each of the impedance inverting networks 503-1 to 503-N is a two-port n-shaped network, which includes two capacitors and an inductor coupled in n-shape, e.g., capacitors 507-1 to 507-N and 509-1 to 509-N, inductors 511-1 to 511-N. Assuming each capacitor has a capacitance X and each inductor has an inductance X, the voltage V _a appearing at an input port of each impedance inverting network, e.g., an input port 513 of the impedance transforming network 503-2 as shown in Fig. 5, is transformed into a current I _a at an output of this impedance inverting network, e.g., an output port 515 of the impedance transforming network 503-2 as shown in Fig. 5, according to the following equation I _a = V _a /jX (1). As can be noted, the resulting output current I _a does not depend, both in amplitude and phase, on the terminating impedance nature. As such, although the states of the auxiliary amplifiers 408-1 to 408-N can modulate the terminating impedance of the impedance inverting networks 503-1 to 503- N, such terminating impedance modulation will not change the output current I _a from the output port, thereby realizing an input isolated (N+l)-way combiner 500.

Besides a lumped n-shaped network, a lumped T-shaped network or a distributed lambda quarter transmission line can also realize the input isolation purpose. For the lumped T-shaped network, the equation (1) still applies. For the distributed lambda quarter transmission line, the output current from the output port is related to the voltage at the input port by an characteristic impedance of the transmission line only, not depending on the termination through which the output current is flowing.

In addition to the input isolation, to sum the signal traveling through the main power chain with the auxiliary amplified signals from the auxiliary power chains properly, the output current I _a from each impedance transforming network 503-1 to 503-N needs to be in phase with the respective auxiliary amplified signals at each combination node 505-1 to 505-N. Referring back to Fig. 4, the controller 414 is coupled to the signal divider network 402 to adjust the phases of the main input signal and auxiliary input signals to ensure the signal traveling through the main chain in phase with the amplified signals from the auxiliary power chains. Alternatively, multiple phase compensators 601-1 to 601-N can be coupled in series with auxiliary amplifiers 608-1 to 608-N in the auxiliary power chains, respectively, as shown in a pulse amplifier 600 of Fig. 6. Optionally, a phase compensator 614 is coupled in series with a main amplifier 604 in the main power chain. As acknowledged by the skilled in the art, all these phase compensators are configured to cooperate with each other to produce predetermined phase delays which ensure the current flowing out of the output ports of impedance transforming network 606-1 to 606-N in phase with the auxiliary amplified signals from the auxiliary amplifiers 608-1 to 608-N at the combination nodes 612-1 to 612- N, respectively. Furthermore, the main amplifier 614 and auxiliary amplifier 616-1 to 616-N are preset to proper operating modes by static DC bias voltages. The operation of the pulse amplifier 600 is similar to that discussed with reference to Fig. 4. Specifically, upon reception of the input RF pulse, the main amplifier 604 operates in the class AB mode automatically and the operating state is set closer to the optimum state. The main amplifier 604 features an improved efficiency due to the preset operating state and the load modulation driven by the auxiliary amplifiers 608-1 to 608-N. As the input RF pulse signal exceeds the preset power thresholds, the associated auxiliary amplifiers is turned on successively to add more output power as required by a particular clinical application. Advantageously, no digital control is required for the embodiment of Fig. 6, which can bring at least the benefits of reducing system complexity and cost.

Fig. 7 illustrate a flow chart 700 of a method for driving a transmit coil in a magnetic resonance imaging system according to one embodiment of the present invention. Fig. 7 is described in combination with Figs 4 and 6.

At step 702, an input signal is divided into a main input signal and multiple auxiliary input signals. For example, the signal divider network 402 divides the input signal into a main input signal and multiple auxiliary input signals.

At step 704, the auxiliary input signals are provided to auxiliary amplifiers in multiple auxiliary power chains, respectively. For example, the auxiliary input signals from the signal divider network 402 are provided to the auxiliary amplifiers 408-1 to 408-N, respectively.

At step 706, the auxiliary amplifiers are switched on based upon a power level of the input signal. For example, a DC operating point of the main amplifier 604 is preset in an active region of operation by DC static bias voltage, e.g., class AB mode. DC operating points of the auxiliary amplifiers 608-1 to 608-N are preset in a cutoff region of operation by DC static bias voltages, e.g., class C mode. Upon reception of the RF input signal, the main amplifier 604 begins the amplification immediately. While the auxiliary amplifiers begin to be switched on successively until the RF input signal exceeds associated power thresholds preset for the RF input signal. Alternatively, the controller 414 controls the biasing of the auxiliary amplifiers 408-1 to 408-N based on the power level of the input signal.

At step 708, the auxiliary input signals are amplified to auxiliary amplified signals by switched-on auxiliary amplifiers, respectively. For example, following the switching on of auxiliary amplifiers 408-1 to 408-N, the auxiliary input signals received by these auxiliary amplifiers are amplified to respective auxiliary amplified signals.

At step 710, the auxiliary amplified signals are provided to serially coupled combiners in a main power chain, respectively. For example, the auxiliary amplified signals from auxiliary amplifiers 408-1 to 408-N are provided to (N+l)-way combiner 416 which is composed of N serially coupled combiners.

At step 712, the main input signal is amplified to a main amplified signal by a main amplifier in the main power chain. For example, the main amplifier 404 is biased in class AB mode. Upon reception of the main input signal from the signal divider network 402, the main amplifier 404 functions to amplify the main input signal into a main amplified signal.

At step 714, the main amplified signal and the auxiliary amplified signals are combined sequentially as the main amplified signal flows through the serially coupled combiners sequentially to generate an output signal that drives the transmit coil. For example, the main amplified signal from the main amplifier 404 flows through the (N+l)-way combiners 416 to combine with the auxiliary amplified signals sequentially by each serially coupled combiner sequentially to generate an output signal that drives the transmit coil 410. Each combiner further includes an impedance transforming network 406-1, 406-2,...406-N and a combination node 412-1, 412-2,...412-N. For each combination, the travelling signal appearing at an input port of an impedance transforming network is transformed to an output current at an output port of the impedance inverting network, and the output current is further summed at the combination node with the corresponding auxiliary amplified signal to generate the signal travelling to the next combination stage.

The invention has been described with reference to the preferred embodiments. Modifications and alterations may occur to others upon reading and understanding the preceding detailed description. It is intended that the invention be constructed as including all such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.