Therapy System for Acute Patient Care

This invention relates to a system for preventing physical deterioration of the muscles of severely injured patients. A serious injury may cause permanent spinal injury resulting in paraplegia or quadriplegia. The resulting prolonged disuse of the limbs produces atrophy of the muscles, deterioration of the tendons and osteoporosis in the bones. As the bones and muscles begin to decondition, there is a reduction in circulation of the blood, thereby predisposing the patient to pressure sores and blood clots.

It has been found that such physical deterioration can be reversed, even after years of muscular disuse, through a physical conditioning program based upon functional electrical stimulation of the paralyzed muscle. Typical apparatus for producing appropriate muscular stimulation is disclosed in Petrofsky et al U.S. Pat. No. 4,421,336 and includes pairs of alternately pulsed electrodes which may be applied to the surface of the skin above those muscles which are to be stimulated. Such stimulation devices have been used to stimulate operation of wheelchairs, tricycles, exercise bicycles and exercise chairs equipped with leg weights. Such systems have produced remarkable rehabilitation of long term paraplegic and quadriplegic patients.

It is known that the effects of muscle disuse begin to manifest themselves almost immediately upon cessation of muscular activity. The above-described prior art devices may be used only by persons who have recovered from the trauma of their injury and have regained a relatively stable physical condition. In the interim, however, such persons may have experienced a long period of physical inactivity during which muscular deterioration has commenced. If the injured person is in a state of coma, the inactivity may continue for many months. It is therefore seen that there is a need for a system to provide physical therapy for such acutely injured persons prior to the time when use of the above-mentioned prior art devices becomes appropriate.

This invention provides apparatus and method for preventing muscular deterioration in a spinal cord patient during the period of time following the spinal cord injury and prior to the period when stimulated dynamic exercise may be commenced. Apparatus in accordance with the invention includes means for stimulating a muscle to be exercised and restraining means for restraining movement of a joint activated by that muscle. The system further includes load selection means for selecting a target load for the muscle and a controller for controlling the operation of the stimulation means.

The controller causes the stimulation means to stimulate the muscle at the required level for causing an effort near the selected target load. Since the system restrains movement of the joint, the net result is a loading of the muscle which is isometric. The loading is maintained for a predetermined period of time, after which the muscle is rested. The entire cycle is repeated until a predetermined routine has been completed or the muscle has become exhausted.

In the preferred embodiment the restraining means comprises a leg brace having load cells which maintain the knee and the ankle at predetermined angles and provide feedback signals indicating the loading which has been achieved. The system alternately stimulates and rests agonist and antagonist muscles for each joint.

FIG. 1 is a pictorial illustration of an injured person undergoing therapy in accordance with this invention;

FIG. 2 illustrates electrode attachment points at the front of the leg;

FIG. 3 illustrates electrode attachment points at the rear of a leg;

FIG. 4 is a schematic block diagram of the electrical components comprising the invention;

FIG. 5 is a schematic illustration of a stimulation waveform;

FIG. 6 is a schematic diagram of computer pin connections;

FIG. 7 is a schematic diagram of feedback elements;

FIG. 8 is a schematic diagram of a stimulator;

FIG. 9 is a side elevation view of a leg brace;

FIG. 10 is a view taken along lines 10--10 of FIG. 9;

FIG. 11 is an enlarged pictorial illustration of the lower end of a leg brace;

FIG. 12 is an enlarged view of that portion of a leg brace at the knee joint, as seen in side elevation;

FIG. 13 is a perspective view of that portion of a leg brace shown in FIG. 12; and

FIGS. 14A and 14B are a flow chart for portions of a computer program for carrying out the present invention.

Physical therapy in accordance with this invention may be carried out as generally illustrated in FIG. 1. An injured person, who may be completely unconcious, rests on a bed 20 with his legs strapped in a pair of articulated braces 21, 21. As hereinafter described, only one brace is functional at any given time, so it is quite feasible to use only a single leg brace 21 and to change it from one leg to the other.

Leg braces 21 and 21 each consist of an upper leg portion 22, a lower leg portion 23 and a foot portion 24. A series of straps 25 and a pair of shoes 26, 26 are provided for securing leg braces 21, 21 to the legs of the patient. The leg braces maintain the legs of the patient with his knees slightly bent. Therefore, a rest 27 of triangular cross section is provided.

During the practice of this invention 24 electrodes 28 may be applied to the legs of the patient as illustrated generally in FIG. 1 and more particularly in FIGS. 2 and 3. These electrodes are applied in groups of three over agonist and antagonist muscles for the knee and ankle joints of each leg. Electrodes 28 are transcutaneous electrodes of the type set forth in TABLE II hereof. These electrodes are attached to the surface of the skin as described in Petrofsky et al U.S. Pat. No. 4,421,336. They are placed at locations for providing stimulation access to the muscles connected for causing movement of a joint to be loaded. The muscles stimulated by the illustrated electrode placement are the hamstring, quadriceps, tibialis anterior and gastrocnemius muscles of each leg. The hamstring and quadriceps muscles function as agonist and antagonist muscles respectively for the knee joint while the tibilias anterior and gastrocnemius muscles function as agonist and antagonist muscles, respectively, for the ankle joint.

In the embodiment as illustrated and as hereinafter described, the quadriceps muscles are stimulated for approximately four seconds to cause extending action of the knee against the restraining action of the leg brace. Thereafter, the quadriceps muscles are rested and the hamstring muscles are stimulated for a four second period to cause flexing of the knee against the restraining action of the leg brace. Thereafter, the hamstring muscles are rested for a four second period while continuing the resting action of the quadriceps muscles. The sequence is continually repeated, so that each muscle group exercises for four seconds, rests for eight seconds, and then exercises again. Since the leg braces permit only imperceptably small movement, an isometric exercise routine is achieved.

After the hamstring and quadriceps muscles have been exercised, the control system is connected for stimulating the tibialis anterior and gastrocnemius muscles of one leg. This causes controlled loading of those muscles against the restraining action of the leg brace at the ankle joint. The same stimulation sequence is performed as described above in connection with the hamstring and quadriceps muscles. Thereafter, the control system may be connected to the electrodes which stimulate muscular activity in the other leg.

As hereinafter described in detail the system measures the effort exerted by the stimulated muscles and compares this effort against a target load. The difference between the actual effort and the target load is utilized for adjusting the stimulation signals. In this manner the muscles may be stimulated to work isometrically at a precisely determined level of effort.

FIG. 4 presents a block diagram of the electrical controls for the system. The heart of the system is a programmed computer 40, which may be an APPLE II PLUS computer manufactured by Apple Computer Inc. of Cupertino, Calif. Computer 40 includes a keyboard 41 which permits entry of control commands. In particular, keyboard 41 is utilized for entry of target loads, as hereinafter described.

The electrical control system also includes a pair of load cells 43a, 43b for alternate selective usage. As shown in FIG. 4 load cell 43a is connected for computer control of muscular activity at the ankle joint. Output signals from load cell 43a are amplified by amplifier 44 and digitized by A/D converter 45. These output signals function as feedback signals, which computer 40 compares against input data from keyboard 41 to generate stimulation control signals. The stimulation control signals, so generated, are applied to a pair of D/A converters 46a, 46b for conversion to analog form. The analog signals then are applied to stimulation generators 47a, 47b which generate stimulation voltages for application to electrodes 28.

The waveform for the signals generated by stimulation generators 47 is illustrated in FIG. 5. That figure illustrates two signals 51 and 52. As described above, the stimulation electrodes are connected for operation in sets of three. The center electrode of each set is grounded. Signal 51 of FIG. 5 represents the voltage between the center electrode and one of the other two electrodes. Signal 52 represents the voltage between the center electrode and the third electrode of the set. Each signal comprises a series of pulses having a pulse width of about 300 microseconds generated at a frequency of 40 pulses per second.

FIG. 6 illustrates the details of the connections between the Apple Computer, A/D converter 45 and D/A converter 46a. The connections to D/A converter 46b are similar and are not illustrated in detail. Connection to the Apple Computer is made via a peripheral connector which plugs into peripheral slot 3 of the Apple Computer. Pin numbers for the connector are set forth on FIG. 6 and are encircled. TABLE I identifies the functions for each of the illustrated pins. The integrated circuits illustrated in FIG. 6, as well as all components of FIGS. 6-8 are identified in TABLE II.

Decoder 48 decodes signals from computer 40 calling for output stimulation commands to be directed to one or the other of D/A converters 46a or 46b. The selection is made by the generated address of one or the other of the instructions POKE 49920, I1 or POKE 49922, I2. (See the complete program listing set forth in TABLE III). When these instructions are executed the decoder 48 transmits a write enabling command on one or the other of lines 118 or 119 for enabling D/A converter 46a or 46b. The selected D/A converter then causes an analog equivalent of I1 or I2 to appear on one or the other of lines 175a or 175b. Potentiometer 189 provides means for stimulation threshold adjustment as hereinafter described.

Reading of analog feedback signals from load cells 43a or 43b is initiated by execution of the instruction POKE 49328,0. This enables A/D converter 45 to read and digitize any analog signal which may be present on line 143. That line may be connected alternatively to one or the other of load cells 43a or 43b, depending upon the joint which has been selected for exercising. Reading of the digitized feedback signal is accomplished by execution of the instruction D=PEEK (49328).

Load cell 43a is a commercially available device, as listed in TABLE I, and the output thereof is applied to amplifier 44 which is also commercially available. The internal details of amplifier 44 are illustrated in FIG. 7 as including four amplifier stages comprising operational amplifiers 53 through 56 and associated resistors. Amplifier 53 provides a gain of 10, while amplifier 54 is connected for use as a balancing amplifier. Thus one input terminal of amplifier 54 is connected to a ten turn potentiometer 72 which is adjusted to produce a 2.5 volt output for a zero load output from load cell 43a. Amplifiers 55 and 56 each provide factor-of-ten gains. A potentiometer 74 provides gain adjustment.

Analog control signals on line 175a are processed by stimulator 47a through use of circuitry which is illustrated in detail in FIG. 8. Stimulator 47b is of identical construction and is not illustrated in detail. Stimulator 47a includes two integrated timing circuits 141 and 142. IC 141 is connected to operate as a 40 Hz free-running multivibrator. The output of IC 141 is applied to input pins 6 and 8 of IC 142. IC 142 produces alternating 300 microsecond pulses each at a frequency of 40 Hz for application to the collector terminals of transistors 145 and 146. The pulse width is set by an appropriate selection of the resistance for resistor 185 and the capacitance of capacitor 187, as shown in the manufacturer's data sheets for integrated circuit 142.

The analog control voltage generated on line 175a by D/A converter 46a is applied to the base terminals of transistors 145 and 146. Concomitantly, output pulses from pins 5 and 9 of IC 142 are applied to the collectors of transistors 146 and 145, respectively. As a result thereof, transistors 146 and 145 generate emitter currents across resistors 130 and 127 providing voltage profiles of the general shape illustrated in FIG. 5. These voltages are applied to the base terminals of transistors 148 and 147 for generating voltage pulses ranging between 0 to 12 volts across the primary windings of transformers 150 and 149.

The voltage pulses across the primary windings of transformers 150 and 149 produce low current, high voltage pulses ranging from 0 to 60 volts across the secondary windings of transformers 150 and 149. The secondary windings of transformers 150 and 149 have one side grounded to a high voltage ground which is different from the ground utilized for the primary windings thereof. The output pulses from the secondary windings are thereby RF isolated to maintain the safety of the paralyzed person.

Output voltages from transformers 150 and 149 are used for driving transistor pairs 162, 164 and 161, 163, respectively. Transistor pairs 162, 164 and 161, 163 provide a current gain so as to have high current, high voltage and low duty cycle pulses available for application across electrode pairs 28b, 28c and 28b, 28a. The signals across the two terminal pairs are represented by signal lines 51 and 52, respectively, of FIG. 5.

FIG. 9 illustrates the construction details of a typical brace 21. As described above, brace 21 includes an upper leg portion 22, a lower leg portion 23 and a foot portion 24. Upper leg portion 22 and lower leg portion 23 are joined along a pivot line 81. Pivotal motion therebetween is substantially restrained, however, by load cell 43b as more fully illustrated in FIGS. 12 and 13. Lower leg portion 23 and foot portion 24 are joined along another pivot line 80. Pivotal motion about pivot line 80 is substantially restrained by load cell 43a as illustrated in FIG. 11.

Upper leg section 22 of a leg brace 21 comprises four longitudinally extending framing rods 82 through 85, and lower leg section 23 comprises four longitudinally extending framing rods 86 through 89. Framing rods 82, 84, 86 and 88 are on that side of leg brace 21 facing load cells 43a and 43b and therefore are visible in FIG. 9. Framing rods 83, 85, 87 and 89 are on the opposite side of leg brace 21 and therefore are not visible in FIG. 9. The framing rods are assembled in pairs in overlapping relationship as illustrated generally in FIG. 9 and more particularly in FIG. 10. The framing rods are provided with a series of mating apertures 90 which are threaded for reception of thumb screws 91. This provides means for adjusting the length of leg sections 22 and 23.

As best illustrated in FIG. 11, framing rods 88 and 89 are pivotally joined to a pair of steel straps 92 and 93, which in turn are secured to the sole 94 of shoe 26. Load cell 43a is supported by a pair of support members 95 and 96, which are mounted upon framing rod 88 and strap 92, respectively.

In operation, shoe 26 is laced securely to the foot of the patient, so that when the ankle muscles are stimulated the foot produces a torque about hinge line 80. Sole 94 of shoe 26 is of rigid construction, so that the torque so produced is transmitted through strap 92 and support member 96 to load cell 43a. Load cell 43a is of relatively rigid construction so that there is no perceptable rotation of shoe 26 about hinge line 80. There is, however, a very slight rotation which is sensed by a resistance bridge arrangement within load cell 43a (see FIG. 6) and converted into an electrical indication of load.

The mounting of load cell 43b is similar to the mounting of load cell 43a, as may be seen by reference to FIGS. 12 and 13. Load cell 43b is supported by mounting blocks 97 and 98 which are secured to rod members 84 and 86, respectively. There is a support member 99 secured to mounting block 97, and load cell 43b is mounted between support member 99 and mounting block 98. This prevents any significant rotary motion between framing rods 84 and 86. Thus angular movement of the knee is restrained when the knee muscles are stimulated.

The operation of computer 40 will now be described with reference to the flow chart of FIGS. 14A and 14B and the program listing of TABLE III. The program as set forth in TABLE III is written in APPLESOFT, a form of the BASIC language adapted for operation on an APPLE II PLUS computer. This is a relatively high level source language, and the program listing is therefore self-explanatory. However, some comments are appropriate.

The program begins with a series of instructions related to load cell calibration. During this part of the program the display monitor provides instructions directing the operator to disconnect the operative load cell from leg brace 21 and to suspend a reference weight (5 lbs.) therefrom. Prior to attachment of the weight, however, balance potentiometer 72 is adjusted under zero load conditions until a value of 125 units is read on the computer monitor. The program loops between steps 700 and 800 while this adjustment is made. After the adjustment is made, the operator may press any key of keyboard 41 to cause an exit from the looping routine. Thereafter, the operator is instructed to place a 5 lb. load on the sensor and input a scale factor through keyboard 41. The scale factor corresponds to the sensitivity of load cell 43a in terms of volts per pound. This factor is established prior to the therapy routine and is marked on the equipment. The factor may be established rather easily by running the program in the laboratory with an added loop to permit trial and error variation of both the input scale factor and the adjustment of potentiometer 74. In such a procedure potentiometer 74 is set to a mid-range value, and the scale factor is varied between 1 and 10 until a value near 5 lbs. is obtained. Then potentiometer 74 is adjusted until a reading of exactly 5 lbs. is obtained. The computer obtains the weight reading as shown at program steps 1020 to 1040 by examining the voltage value stored in location 49328, adjusting for the zero load value, and then applying the scale factor.

After calibration has been completed, then the stimulation threshold is set for both the ventral and dorsal muscles. This setting is accomplished through use of potentiometer 189 in each stimulation channel, as documented at program steps 1300 through 1860. For each muscle the potentiometer is adjusted from a zero position up to another position where the muscle begins to twitch. Threshold settings are established while the computer is transmitting the hexadecimal equivalent of the decimal number 64 to D/A converters 46a and 46b. Thereafter, the computer uses the number 64 as the value for threshold stimulation voltage.

After threshold setting has been completed the attending physician checks the patient and prescribes target loads for the ventral and dorsal muscles. These target loads may have any value up to 30 lbs. and are input to the computer at program step 3150 as variables VENT and DS. Keyboard 41 is used for this input.

After the target loads have been read into the computer the stimulation sequence commences. The steps for that sequence are outlined in the flow chart of FIGS. 14A and 14B. As noted in the flow chart the computer begins by setting the ventral stimulation at the threshold value and setting the dorsal stimulation at 0. The ventral stimulation then is increased stepwise until the target value has been reached. The target value is held for four seconds, following which the ventral stimulation is set to zero and the dorsal stimulation is set at a value equal to its target load. Thereafter, the dorsal stimulation is increased stepwise to its target value and held for four seconds. Following this, both muscles are rested for four seconds and the entire sequence is repeated. The routine produces periodic isometric loading of the stimulated muscles.

As the routine progresses the muscles will begin to tire, and the feedback signal from the load cell will begin to decrease. The computer compensates for this decrease by increasing the stimulation level until a maximum value of 200 units is reached (corresponds to 50 volts). No further voltage increases are made beyond that point. The routine is terminated at any time by depressing any key of keyboard 41. The computer senses the depression of the key at one or the other of program steps 3350 or 3450. The program then discontinues all stimulation and awaits an instruction to recommence, if desired. Stimulation may be resumed by typing the letter R, as noted at program step 3240.

While the method herein described, and the form of apparatus for carrying this method into effect, constitute preferred embodiments of this invention, it is to be understood that the invention is not limited to this precise method and form of apparatus, and that changes may be made in either without departing from the scope of the invention, which is defined in the appended claims. ##SPC1##