Method and computing device using a machine learning algorithm to control a menstrual pain management device

Abstract

Method and control device using a machine learning algorithm to control a menstrual pain management device. Information characterizing a user of the menstrual pain management device is determined. The control device executes the machine learning algorithm, using a predictive model of the machine learning algorithm to determine operating parameter(s) of the menstrual pain management device based on inputs, the inputs comprising the information characterizing the user of the menstrual pain management device. Operation of the menstrual pain management device is controlled based on the determined operating parameter(s). In one implementation, the control device is a computing device sending control command(s) to the menstrual pain management device based on the determined operating parameter(s). In another implementation, the control device is the menstrual pain management device, using a control module of the menstrual pain management device for controlling components of the menstrual pain management device based on the determined operating parameter(s).

Description

METHOD AND COMPUTING DEVICE USING A MACHINE LEARNING ALGORITHM TO CONTROL A MENSTRUAL PAIN MANAGEMENT DEVICETECHNICAL FIELD

[0001] The present disclosure relates to the field of menstrual pain management. More particularly, the present disclosure presents a method and computing device using a machine learning algorithm to control a menstrual pain management device.BACKGROUND

[0002] Menstrual pain is a condition that accompanies or precedes menses. The intensity of menstrual pain varies from one person to another, and for different reasons: period of the month, flow of menses, hormone levels, stress level, time of day, etc.

[0003] Menstrual pain may adversely affect daily activities and render the person affected incapable of pursuing regular daily activities. This condition may be experienced occasionally or systematically every month for up to several days.

[0004] Menstrual pain management options typically include the following types: pharmaceutical or non-pharmaceutical. Not all menstrual pain sufferers react favorably to the pharmaceutical options. Non-pharmaceutical options often include applying warmth on the abdomen or using a tens machine. Transcutaneous Electrical Nerve Simulation (TENS) machine do bring some level of relief when the tens are properly positioned on the abdomen. However, TENS machines include wires for interconnecting the electrode to a control module, and the control module must be carried in a pocket. This renders the use of TENS machines suitable for home use, but not as convenient for use during regular activities outside of the home. There is therefore a need for a menstrual pain management device which is simple to use, portable and operated discreetly.

[0005] Furthermore, an optimal mode of operation of the menstrual pain management device varies significantly from one person to another, based on a plurality of information characterizing the person. This large amount of information makes it difficult to determine an optimal mode of operation of the menstrual pain management device for a given person, using the information characterizing this given person.

[0006] There is therefore a need for a new method and computing device using a machine learning algorithm to control menstrual pain management.SUMMARY

[0007] According to a first aspect, the present disclosure relates to a method using a machine learning algorithm to control a menstrual pain management device. The method comprises storing a predictive model of the machine learning algorithm in a memory of a control device. The method comprises determining information characterizing a user of the menstrual pain management device. The method comprises executing, by a processor of the control device, the machine learning algorithm. The machine learning algorithm uses the predictive model for determining at least one operating parameter of the menstrual pain management device based on inputs. The inputs comprise the information characterizing the user of the menstrual pain management device. The method comprises controlling operation of the menstrual pain management device based on the at least one determined operating parameter.

[0008] In a particular aspect, the machine learning algorithm implements a neural network, the predictive model comprising weights of the neural network.

[0009] In another particular aspect, controlling operation of the menstrual pain management device based on the at least one determined operating parameter comprises sending a control command generated based on the at least one determined operating parameter to the menstrual pain management device via acommunication interface of the control device. A control module of the menstrual pain management device applies the control command to control operation the menstrual pain management device. In a particular embodiment, the control device is a mobile computing device or a cloud server.

[0010] In still another particular aspect, the control device is the menstrual pain management device. The menstrual pain management device comprises a control module in electric connection with an electrode layer of the menstrual pain management device and a heating layer of the menstrual pain management device. The control module also comprises the memory and the processor. The control module controls operation of at least one of the electrode layer and the heating layer based on the at least one determined operating parameter.

[0011] In yet another particular aspect, the information characterizing the user of the menstrual pain management device comprises information belonging to at least one of the following categories: demographic characteristics of the user, medical and gynecological conditions of the user, physical characteristics of the user and lifestyle characteristics of the user.

[0012] In another particular aspect, the information characterizing the user of the menstrual pain management device comprises a combination of at least some of the following information: sex, age, race, ethnicity, occupation, place of residence, presence or absence of endometriosis, primary or secondary dysmenorrhea, pregnancy history, body fat percentage, gynecological disease, menstrual regularity, use of intrauterine device, use of contraceptive pill, size, weight, body mass index, pain level, American Society of Anesthesiologists (ASA) score, use of medications and quality of life.

[0013] In still another particular aspect, the menstrual pain management device comprises an electrode layer and a heating layer. The at least one operating parameter of the menstrual pain management device comprises at least one of the following: one or more electric stimulation modes of electrodes of the electrodelayer, a temperature of a heating element of the heating layer, a duration of use of the menstrual pain management device, a number of treatment cycles by the menstrual pain management device, and a frequency of treatment by the menstrual pain management device.

[0014] In yet another particular aspect, the menstrual pain management device comprises an electrode layer. The at least one operating parameter of the menstrual pain management device comprises one or more electric stimulation modes of electrodes of the electrode layer. The one or more electric stimulation modes comprise at least one of the following: a waveform of electric impulses generated by the electrodes, a nature of the electric impulses generated by the electrodes, a pulse length of the electric impulses generated by the electrodes, a global frequency of the electric impulses generated by the electrodes, an internal frequency of the electric impulses generated by the electrodes, and an intensity of the electric impulses generated by the electrodes.

[0015] According to a second aspect, the present disclosure relates to a computing device. The computing device comprises a communication interface, memory for storing a predictive model of a machine learning algorithm and a processor. The processor determines information characterizing a user of a menstrual pain management device. The processor executes the machine learning algorithm. The machine learning algorithm uses the predictive model for determining at least one operating parameter of the menstrual pain management device based on input. The inputs comprise the information characterizing the user of the menstrual pain management device. The processor sends a control command generated based on the at least one determined operating parameter to the menstrual pain management device via the communication interface, a control module of the menstrual pain management device applying the control command to control operation of the menstrual pain management device.

[0016] According to a third aspect, the present disclosure relates to a non-transitory computer readable medium comprising instructions executable by a processor of acomputing device. The execution of the instructions by the processor of the computing device provides for using a machine learning algorithm to control a menstrual pain management device, by implementing the following steps. The processor executes the instructions to store a predictive model of the machine learning algorithm in a memory of the computing device. The processor executes the instructions to determine information characterizing a user of the menstrual pain management device. The processor executes the instructions to execute the machine learning algorithm. The machine learning algorithm uses the predictive model for determining at least one operating parameter of the menstrual pain management device based on inputs. The inputs comprise the information characterizing the user of the menstrual pain management device. The processor executes the instructions to send a control command generated based on the at least one determined operating parameter to the menstrual pain management device via a communication interface of the computing device. A control module of the menstrual pain management device applies the control command to control operation of the menstrual pain management device.

[0017] In a particular aspect, the machine learning algorithm implements a neural network, the predictive model comprising weights of the neural network.

[0018] In another particular aspect, the computing device consists of a mobile computing device or a cloud server.

[0019] In still another particular aspect, the information characterizing the user of the menstrual pain management device comprises information belonging to at least one of the following categories: demographic characteristics of the user, medical and gynecological conditions of the user, physical characteristics of the user and lifestyle characteristics of the user.

[0020] In yet another particular aspect, the information characterizing the user of the menstrual pain management device comprises a combination of at least some of the following information: sex, age, race, ethnicity, occupation, place ofresidence, presence or absence of endometriosis, primary or secondary dysmenorrhea, pregnancy history, body fat percentage, gynecological disease, menstrual regularity, use of intrauterine device, use of contraceptive pill, size, weight, body mass index, pain level, American Society of Anesthesiologists (ASA) score, use of medications and quality of life.

[0021] In another particular aspect, the menstrual pain management device comprises an electrode layer and a heating layer. The at least one operating parameter of the menstrual pain management device comprises at least one of the following: one or more electric stimulation modes of electrodes of the electrode layer, a temperature of a heating element of the heating layer, a duration of use of the menstrual pain management device, a number of treatment cycles by the menstrual pain management device, and a frequency of treatment by the menstrual pain management device.

[0022] In still another particular aspect, the menstrual pain management device comprises an electrode layer. The at least one operating parameter of the menstrual pain management device comprises one or more electric stimulation modes of electrodes of the electrode layer. The one or more electric stimulation modes comprise at least one of the following: a waveform of electric impulses generated by the electrodes, a nature of the electric impulses generated by the electrodes, a pulse length of the electric impulses generated by the electrodes, a global frequency of the electric impulses generated by the electrodes, an internal frequency of the electric impulses generated by the electrodes, and an intensity of the electric impulses generated by the electrodes.BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Embodiments of the disclosure will be described by way of example only with reference to the accompanying drawings, in which:

[0024] Figure 1 is a perspective top side view of an example of the presentmenstrual pain management device;

[0025] Figure 2 is a side view of the menstrual pain management device of Figure 1 ;

[0026] Figure 3 is a front view of the menstrual pain management device of Figure 1 ;

[0027] Figure 4 is a top view of an intermediate layer;

[0028] Figure 5 is an exploded view of the menstrual pain management device of Figure 1 ;

[0029] Figure 6 is a schematic functional diagram representing the components of the control module 14;

[0030] Figure 7 is a schematic functional diagram representing a computing device executing a machine learning algorithm to control the menstrual pain management device;

[0031] Figure 8 is a schematic functional diagram representing the menstrual pain management device executing the machine learning algorithm;

[0032] Figure 9 represents a method using the machine learning algorithm to control the menstrual pain management device;

[0033] Figure 10 is a schematic representation of the machine learning algorithm; and

[0034] Figure 11 illustrates an exemplary implementation of the machine learning algorithm by a neural network.DETAILED DESCRIPTION

[0035] The foregoing and other features will become more apparent upon reading of the following non-restrictive description of illustrative embodiments thereof, given by way of example only with reference to the accompanying drawings. Likenumerals represent like features on the various drawings.

[0036] Various aspects of the present disclosure generally address problems with current menstrual pain management devices, including the difficulty of determining an optimal mode of operation of the menstrual pain management device for a given person, using information characterizing this given person.

[0037] Referring concurrently to Figures 1-5 there are shown various views of an example of the present menstrual pain management device. The example shown on Figures 1-5 is for explaining the concepts, construction and operation of the present menstrual pain management device and should not be interpreted to limit the present menstrual pain management device to the design and implementation illustrated therein.

[0038] The menstrual pain management device 100 is composed of four layers: a base layer 110, an electrode layer 120, a heating layer 130 and an electronic layer 140. The four-layer configuration provides a menstrual pain management device 100 that is simple to operate, which can be discreetly used and very portable.

[0039] The base layer 110 receives the other three layers: the electrode layer 120, the heating layer 130 and the electronic layer 140. The base layer 110 is made of flexible material. The base layer 110 is adapted to removably adhere to skin. For example, the base layer 110 may be adapted to receive a removable skin adhesive which can be changed after use. Alternatively, the base layer 110 may receive a washable adhesive liquid for removably affix the base layer onto skin. In yet another alternative, the material of the base layer 110 may become adhesive when humid. For example, the base layer 110 may be made of medical-grade silicone.

[0040] The base layer 110 is shaped and sized to cover the area of the abdomen of a user impacted by menstrual pain. For example, the base layer 110 may be shaped as a rectangle, a trapezoid, an oval, a butterfly, or any other shape which is adapted for receiving the other three layers of the present menstrual pain management device 100 while properly positioning the electrode layer 120 and theheating layer 130 over the area of the abdomen of the user impacted by menstrual pain.

[0041] The electrode layer 120 sits over the base layer 110. The electrode layer 120 defines two electrodes 122. The two electrodes 122 are positioned between a center of the base layer 110 and opposite sides thereof. The electrode layer 120 covers at least two areas longitudinally located along the base layer 110. The electrode layer 120 comprises two electrodes 122 created by electric wire shaped in similar design on each side of the electrode layer 120. The design of the electric wire of each electrode 122 may comprise multiple interconnected electric wires electrically connected in serial, in parallel, or in a combination of serial and parallel electrical connections. The electric wire of each electrode 122 is selected to be capable of providing Transcutaneous Electrical Nerve Simulation (TENS) when the menstrual pain management device 100 is installed on a user’s abdomen and the electronic layer 140 provides an electric current to the electrodes 122. The design of electric wire shown on Figure 5 is for example only.

[0042] To minimize electrical consumption while optimizing the area over which TENS is provided, the electrodes 122 may be shaped to be positioned towards the sides of the menstrual pain management device 100 while not extending over a center portion thereof, so as to define a central aperture between the electrodes 122. Furthermore, the gauge of the electric wire and the design of the electrodes 122 are selected to keep electric consumption to a minimum while optimizing the TENS when used. The electrode layer 120 is flexible to adapt to the shape of the abdomen of the user. The terminology electrode layer refers to both implementations where the electrodes 122 are designed as a layer inserted between the base layer 110 and the electronic layer 140, or as two separate electrodes 122 affixed therebetween.

[0043] The heating layer 130 is also designed to minimize electrical consumption while optimizing the area over which the heating layer provides warmth. For doing so, the heating layer 130 may be shaped as a belt adjacent aperiphery of the base layer 110. Alternatively, the heating layer 130 may be shaped as another 2D geometric shape over a central area of the base layer 110. The heating layer 130 is adapted to be inserted between the base layer 110 and the electronic layer 140, below or above the electrode layer 120. The heating layer 130 is also flexible to follow a shape of the abdomen of a user. The heating layer 130 is composed of a heating element 135 and a structure for affixing the heating element between the base layer 110 and the electronic layer 140.

[0044] The electronic layer 140 comprises a control module 145 in electric connection with the electrode layer 120 and the heating layer 130. The control module 145 controls operation of the electrodes 122 of the electrode layer 120 and the heating element 135 of the heating layer 130. The electronic layer 140 further comprises a cover 147 which covers the control module 145, the electrode layer 120 and the heating layer 130. The control module 145 electrically connects with the electrodes 122 of the electrode layer 120 and with the heating element 135 of the heating layer 130 with any means known in the art to electrically interconnect components, such as through wires, traces, connectors, etc.

[0045] Referring now further to Figure 6, the control module 145 comprises a memory 210 for storing instructions to be executed by a processor 220, and which when executed control the electrodes 122 of the electrode layer 120 and the heating element 135 of the heating layer 130. The control module 145 further comprises user controls 230 for controlling operation of the electrodes 122 of the electrode layer 120 and the heating element 135 of the heating layer 130. The user controls 230 may consist of manual controls, such as for example on / off button, + and - buttons, a slider, a rotary dial or any other type of manual controls for controlling intensity, or any other type of manual controls known in the art to control electrodes 122 and the heating element 135. The user controls 230 may include separate manual controls for the electrodes 122 of the electrode layer 120 and for the heating element 135 of the heating layer 130. In addition to user controls 230 for controlling the actuation and / or intensity of, the electrodes 122 andthe heating element 135, the user controls 230 may further include a mode selection control for selecting an electric wave pattern for the electrodes 122 of the electrode layer 120 and / or a frequency of the wave used to control the electrodes 122. Wave patterns and wave frequency for TENS are well known in the art. The cover 147 of the electronic layer 140 provides access to the user controls of the control module 145. Alternately, the mode button may select a mode of operation for the menstrual pain management device from a plurality of preprogrammed programs for controlling the electrodes 122 of the electrode layer 120 and the heating element 135 of the heating layer 130 either concurrently or separately. The preprogrammed programs include for example: a low control for concurrently controlling the electrodes 122 of the electrode layer 120 on low intensity and the heating element 135 of the heating layer 130 on low intensity; a high control for concurrently controlling the electrodes 122 of the electrode layer 120 on high intensity and the heating element 135 of the heating layer 130 on high intensity, a sequence of controls of the electrodes 122 of the electrode layer 120 and / or of the heating element 135 of the heating layer 130, a sequence of increases of the intensity of the electrodes 122 of the electrode layer 120 and / or of the heating element 135 of the heating layer 130, a sequence of decreases of the intensity of the electrodes of the electrode layer 120 and / or of the heating element of the heating layer 130, etc.

[0046] The user controls 230 of the control module 145 may further comprise a feedback button for receiving feedback from a user of the menstrual pain management device. Upon receipt of a feedback through the feedback button, the control module 145 adapts the control of the electrodes 122 of the electrode layer 120 and the heating element 135 of the and heating layer 130.

[0047] The cover 147 of the electronic layer 140 may cover the user controls 230 while allowing the user controls 230 to be accessed over the cover 147. Each user control 230 may be represented by an embossed button, for example a molded button including a tactile icon, such as for example a “+” button for increasing intensity of the electrodes 122 or the heating element 135, a fordecreasing intensity of the electrodes 122 or the heating element 135, a “” icon to represent actuating of the heating element 135, and any type of embossed tactile icon which allows operation of the present menstrual pain management device 100 while being worn by a user so as to permitting discrete operation. Any of the discussed user controls 230 could thus be operated through a corresponding button or actuator provided directly with a tactile icon or a regular button or actuator with a nearby tactile icon on the top cover 147.

[0048] The control module 145 further comprises a rechargeable battery240 for powering the control module 145, the electrodes 122 of the electrode layer 120 and the heating element 135 of the heating layer 130. Of course, those skilled in the art will understand that the expression rechargeable battery 240 is meant to encompass both a single battery or a plurality of batteries.

[0049] The control module 145 further comprises a charger 250 for recharging the battery. The charger 250 may consist of a well-known in the art battery charger with an electrical connector for hooking up to electric power through an electric cord. In another embodiment, the charger 250 is a wireless charging module for wirelessly recharging the rechargeable battery. The wireless charging module may be one of an inductive wireless charging module and a capacitive wireless charging module.

[0050] . The control module 145 further comprises a wireless communication module 260 for communicating with a wireless electronic device such as for example a smart phone and more particularly with an application on the wireless electronic device adapted for communication with the menstrual pain management device 100. The wireless communication module 260 may communicate using any known standards such as for example Bluetooth, Bluetooth Low-Energy, Wi-Fi, cell phone standard, or any similar or proprietary communication standard.

[0051] The wireless communication module 260 is adapted to wirelessly communicate a recharge level of the battery to the wireless electronic device.Alternatively, the control module 145 is adapted for controlling the electrodes 122 of the electrode layer 120 to generate a haptic signal indicative of a low recharge level of the rechargeable battery.

[0052] To render the menstrual pain management device 100 water resistant, a periphery of the electronic layer 140 is fused to the base layer 110 to define a protective shell around the electrode layer 120, the heating layer 130 and the control module 145.

[0053] Figure 4 illustrates an electronic base layer. For manufacturing, maintenance and repair purposes, the electronic layer 140 may be provided as an enclosed unit which is then affixed to the base layer 110, the electrode layer 120 and the heating layer 130. In this approach, only the electronic base layer shown on Figure 4 and Figure 5 is fused to the base layer 110, the electrode layer 120 and the heating layer 130 so as to be detachable from the electronic layer 140. Detaching the electronic layer 140 from the electrode layer 120 and the heating layer 130 facilitates cleaning of the base layer 110 without risking damage to the electronic layer 140. Furthermore, this approach allows access to the control module 145 in case a repair or maintenance is required, for example if the rechargeable battery needs replacement. Additionally, by allowing disconnection of the fused base layer 110, electrode layer 120 and heating layer 130 from the electronic layer 140, it is possible to change the fused base layer 110, electrode layer 120 and heating layer 130 to change for another type of electrodes 122 or heating element 135 without having to replace the whole menstrual pain management device 100.USE OF MACHINE LEARNING TO CONTROL THE MENSTRUAL PAIN MANAGEMENT DEVICE

[0054] Referring now to Figure 9, a method 500 using a machine learning algorithm to control the menstrual pain management device 100 is represented. The method 500 is executed by a control device schematically represented in Figure 9. Figure 7illustrates a first implementation where the control device is a computing device 300 different from the menstrual pain management device 100. Figure 8 illustrates a second implementation where the control device is the menstrual pain management device 100.

[0055] Reference is now made concurrently to Figures 5, 6, 7 and 9. Figure 7 comprises a simplified representation of the menstrual pain management device 100 previously described in relation to Figures 1-6. Only the components of the menstrual pain management device 100 used for the following description of the method 500 are illustrated in Figure 7.

[0056] As mentioned previously, the computing device 300 illustrated in Figure 7 corresponds to the control device of Figure 9 executing the method 500. The computing device 300 comprises a processor 310, memory 320, and a communication module 330. The computing device 300 may comprise additional components, such as another communication module 330, a user interface 340, a display 350, etc.

[0057] Examples of computing devices 300 adapted to execute steps 515-535 of the method 500 include a mobile computing device (e.g. smartphone, tablet, etc.), a computer (e.g. laptop, desktop), a cloud server, etc.

[0058] The processor 310 is capable of executing instructions of computer program(s). The processor 310 executes a machine learning algorithm 312 and a remote control software 314, as will be detailed later in the description. The computing device 300 may include additional processor(s), the remote control software 314 being executed by the processor 310 and the machine learning algorithm 312 being executed by another processor (not represented in Figure 7).

[0059] The memory 320 stores instructions of computer program(s) executed by the processor 310 (for implementing the machine learning algorithm 312, the remote control software 314, etc.), data generated by the execution of the computer program(s), data received via the communication module 330 (or anothercommunication module), etc. Only a single memory 320 is represented in Figure 7, but the computing device 300 may comprise several types of memories, including volatile memory (such as a volatile Random Access Memory (RAM), etc.) and non-volatile memory (such as a hard drive, electrically-erasable programmable read-only memory (EEPROM), flash, etc.).

[0060] The communication module 330 allows the computing device 300 to exchange data with remote devices (e.g. with the menstrual pain management device 100, with a training server 400 illustrated in Figure 9, etc.) over one or more communication networks (not represented in Figure 7 for simplification purposes). For example, the communication module 330 is a wired communication module, adapted to support wired communication protocols such as Ethernet, etc. In another example, the communication module 330 is a wireless communication module, adapted to support wireless communication protocols such as Wi-Fi, Bluetooth®, Bluetooth® Low Energy (BLE), etc. The communication module 330 usually comprises a combination of hardware and software executed by the hardware, for implementing the communication functionalities of the communication module 330.

[0061] Details of the training server 400 are not represented in the Figures, since the components of the training server 400 are similar to the components of the computing device 300 (processor(s), memory, communication module(s), etc.).

[0062] Following is a description of the steps of the method 500.

[0063] The training server 400 executes step 505 by generating a predictive model used by the machine learning algorithm 312. This step will be further detailed later in the description.

[0064] The training server 400 executes step 510 by transmitting the predictive model generated at step 505 to the computing device 300.

[0065] The processor 310 of the computing device 300 executes step 515 by receiving the predictive model via the communication module 330 of the computingdevice 300.

[0066] The processor 310 of the computing device 300 executes step 520 by storing the predictive model in the memory 320 of the computing device 300.

[0067] The processor 310 of the computing device 300 executes step 525 by determining information characterizing a user of the menstrual pain management device 100. The determination is performed by at least one of the following means: reading the information from the memory 320, receiving the information via the communication module 330 (from a remote computing device not represented in the Figures), receiving the information via the user interface 340 (e.g. through interactions of the user of the menstrual pain management device 100 with the user interface 340), a combination thereof, etc. For example, a profile of the user of the menstrual pain management device 100 is generated and stored in the memory 320, the profile comprising the information characterizing the user. The profile of the user stored in the memory 320 is read each time step 525 is executed, to retrieve the information characterizing the user. The profile can be updated if some data related to the user have changed. Examples of information characterizing the user of the menstrual pain management device 100 will be provided later in the description.

[0068] The processor 310 of the computing device 300 executes step 530 by executing the machine learning algorithm 312. The machine learning algorithm 312 uses the predictive model (stored in the memory 320 at step 520) for determining at least one operating parameter of the menstrual pain management device 100 based on inputs. The inputs comprise the information characterizing the user of the menstrual pain management device 100 (determined at step 520). Examples of operating parameters of the menstrual pain management device 100 will be provided later in the description.

[0069] The processor 310 of the computing device 300 executes step 535 by controlling operation of the menstrual pain management device 100 based on theat least one operating parameter (determined at step 530).

[0070] The following implementation of step 535 is specifically adapted to the control device illustrated in Figure 9 being the computing device 300 (a different implementation will be described later in relation to Figure 8).

[0071] The processor 310 of the computing device 300 generates a control command based on the at least one operating parameter (determined at step 530). The control command is sent to the menstrual pain management device 100 via the communication module 330 of the computing device 300. For example, the control command consists of a message sent to the menstrual pain management device 100, the message comprising the operating parameter(s) determined at step 530.

[0072] The control module 145 of the menstrual pain management device 100 applies the control command received from the computing device 300, to control operation of the menstrual pain management device 100.

[0073] For example, the processor 210 of the control module 145 executes a control software 214. The control software 214 receives the control command from the computing device 300 via the wireless communication module 260 of the control module 145. The control software 214 processes the control command, to control operation of the menstrual pain management device 100 (e.g. to control operation of the electrode layer 120 and I or the heating layer 130 illustrated in Figure 7). Additional details will be provided later in the description in relation to Figure 8.

[0074] Steps 515, 520, 525 and 535 are performed by the remote control software 314, while step 530 is performed by the machine learning algorithm 312. Figure 7 illustrates the remote control software 314 and the machine learning algorithm 312 being executed by the same processor 310. Alternatively, the machine learning algorithm 312 is executed by a dedicated processor (not represented in Figure 7) optimized for executing machine learning algorithms, and the remote controlsoftware 314 is executed by a standard processor (e.g. 310).

[0075] Reference is now made concurrently to Figures 5, 6, 8 and 9. Figure 8 also comprises a simplified representation of the menstrual pain management device 100 previously described in relation to Figures 1-6. Only the components of the menstrual pain management device 100 used for the following description of the method 500 are illustrated in Figure 8. The menstrual pain management device 100 corresponds to the control device of Figure 9 executing the method 500.

[0076] The processor 210 of the menstrual pain management device 100 executes step 515 by receiving the predictive model via the wireless communication module 260 of the menstrual pain management device 100.

[0077] The processor 210 of the menstrual pain management device 100 executes step 520 by storing the predictive model in the memory 210 of the menstrual pain management device 100.

[0078] The processor 210 of the menstrual pain management device 100 executes step 525 by determining information characterizing a user of the menstrual pain management device 100. The determination is performed by at least one of the following means: reading the information from the memory 210, receiving the information via the wireless communication module 260 (from a remote computing device such as the computing device 300 illustrated in Figure 7), a combination thereof, etc. In an exemplary implementation mentioned previously in relation to Figure 7, a profile of the user of the menstrual pain management device 100 is generated and stored in the memory 210, the profile comprising the information characterizing the user. The profile of the user stored in the memory 210 is read each time step 525 is executed, to retrieve the information characterizing the user. Alternatively, the profile is stored by the computing device 300 illustrated in Figure 7 and transmitted to the menstrual pain management device 100 each time step 525 is executed. Examples of information characterizing the user of the menstrual pain management device 100 will be provided later in the description.

[0079] The processor 210 of the menstrual pain management device 100 executes step 530 by executing the machine learning algorithm 312. As mentioned previously in relation to Figure 7, the machine learning algorithm 312 uses the predictive model (stored in the memory 210 at step 520) for determining at least one operating parameter of the menstrual pain management device 100 based on inputs. The inputs comprise the information characterizing the user of the menstrual pain management device 100 (determined at step 525). Examples of operating parameters of the menstrual pain management device 100 will be provided later in the description.

[0080] The processing capabilities of the processor 210 being potentially limited, the machine learning algorithm 312 executed by the processor 210 may be a lightweight version of the machine learning algorithm 312 executed by the processor 310 illustrated in Figure 7.

[0081] The processor 210 of the menstrual pain management device 100 executes step 535 by controlling operation of at least one of the electrode layer 120 and the heating layer 130 of the menstrual pain management device 100 based on the at least one operating parameter (determined at step 530).

[0082] As mentioned previously in relation to Figures 1-6, the control module 145 is in electric connection with the electrode layer 120 and the heating layer 130. Furthermore, the control module 145 comprises the battery 240.

[0083] Figure 8 illustrates an exemplary implementation where the battery 240 is in electric connection with the electrode layer 120 and the heating layer 130. In this case, controlling operation of the electrode layer 120 and I or the heating layer 130 is implemented by controlling operation of the battery 240 (based on the operating parameter(s) determined at step 530).

[0084] In an alternative implementation (not represented in the Figures), the battery 240 is in electric connection with an electric interface module of the control module 145, the electric interface module being in electric connection with the electrodelayer 120 and the heating layer 130. In this case, controlling operation of the electrode layer 120 and I or the heating layer 130 is implemented by controlling operation of the electric interface module (based on the operating parameter(s) determined at step 530).

[0085] The control software 214 executed by the processor 210 executes steps 515, 520, 525 and 535.

[0086] Following are examples of the information (determined at step 525) characterizing the user of the menstrual pain management device 100. The exemplary information belongs to one of the following categories.

[0087] A first category comprises demographic characteristics of the user, such as sex, age, race, ethnicity, occupation (e.g. student, worker, etc.), place of residence, etc.

[0088] A second category comprises medical and gynecological conditions of the user, such as presence or absence of endometriosis, primary or secondary dysmenorrhea, pregnancy history, body fat percentage, gynecological disease, menstrual regularity, use of intrauterine device, use of contraceptive pill, etc.

[0089] A third category comprises physical characteristics of the user, such as size, weight, body mass index, pain level (e.g. a numeric rating scale (NRS) of the pain), American Society of Anesthesiologists (ASA) score, etc.

[0090] A fourth category comprises lifestyle characteristics of the user, such as use of medications, quality of life, etc.

[0091] Any combination of the aforementioned examples of information can be used as inputs of the machine learning algorithm 312 at step 530. For example, one or more types of information of each of the four categories is used as inputs. In another example, one or more types of information of only a subset of the four categories is used as inputs. Furthermore, a person skilled in the art will readily understand that the aforementioned examples of information characterizing theuser are for illustrations purposes only, and that additional information may be used.

[0092] Following are examples of the operating parameter(s) of the menstrual pain management device 100 (determined at step 530).

[0093] A first type of operating parameters comprises electric stimulation mode(s) of the electrodes 122 (illustrated in Figure 5) of the electrode layer 120. The electrodes 122 generate electric impulses, which are delivered through a surface of the skin of the user. Examples of such electric simulation modes include a waveform of the electric impulses, a nature (monophasic or biphasic) of the electric impulses, a pulse length of the electric impulses, a global frequency of the electric impulses, an internal frequency of the electric impulses, an intensity of the electric impulses, etc.

[0094] In an exemplary implementation, the control software 214 controls operation of the battery 240 by configuring the battery 240 to deliver an electrical current to the electrodes 122 of the electrode layer 120 with characteristics allowing to achieve the electric simulation mode(s) determined at step 530.

[0095] Another type of operating parameter is a temperature of the heating element 135 (illustrated in Figure 5) of the heating layer 130.

[0096] In an exemplary implementation, the control software 214 controls operation of the battery 240 by configuring the battery 240 to deliver an electrical current to the heating element 135 of the heating layer 130 with an intensity (and I or voltage) allowing to achieve the temperature determined at step 530.

[0097] Still another type of operating parameter is a duration of use of the menstrual pain management device 100.

[0098] In an exemplary implementation, the control software 214 controls operation of the battery 240 by configuring the battery 240 to deliver an electrical current to the electrodes 122 of the electrode layer 120 and to the heating element 135 of theheating layer 130 for the duration determined at step 530.

[0099] Yet another type of operating parameter is a number of treatment cycles by the menstrual pain management device 100.

[0100] Another type of operating parameter is a frequency of treatment by the menstrual pain management device 100.[00i0i]Any combination of the aforementioned examples of operating parameters can be generated (determined) as outputs of the machine learning algorithm 312 at step 530. Furthermore, a person skilled in the art will readily understand that the aforementioned examples of operating parameters are for illustrations purposes only, and that additional types of operating parameters may be generated.

[0102] Reference is now made concurrently to Figures 7, 8, 9 and 10, where Figure 10 is a schematic representation of the machine learning algorithm 312 is illustrated in Figures 7 and 8, representing the inputs and the outputs used by the machine learning algorithm 312 when performing step 530 of the method 500.

[0103] For illustration purposes only, the inputs are represented as comprising at least one instance of each one of the previously described categories of information characterizing the user.

[0104] Reference is now made concurrently to Figures 7, 8, 9, 10 and 11 , where Figure 11 illustrates an exemplary implementation of the machine learning algorithm 312 of Figure 10 by a neural network 600.

[0105] The neural network 600 illustrated in Figure 11 is for illustration purposes only. A person skilled in the art will readily understand that other implementations of the neural network 600 may be used for performing step 530 of the method 500.

[0106] The neural network 600 includes an input layer for receiving the information characterizing the user, followed by a plurality of fully connected layers. The last layer among the plurality of fully connected layers is an output layer for outputting the inferred operating parameter(s) of the menstrual pain management device 100.

[0107] For illustration purposes only, the input layer represented in Figure 11 comprises neurons for receiving information belonging to each one of the previously described categories of information characterizing the user. For example, the input layer comprises respective neurons for receiving the sex of the person, the age of the person, the pregnancy history of the person, the size of the person, the weight of the person, etc.

[0108] The operations of the fully connected layers are well known in the art. The number of fully connected layers is an integer greater than 2, including the output layer (Figure 11 represents four fully connected layers, including the output layer, for illustration purposes only). The number of neurons in each fully connected layer may vary. During the training phase of the neural network, the number of fully connected layers and the number of neurons for each fully connected layer are selected; and may be adapted experimentally.

[0109] The output layer comprises one neuron for outputting each one of the operating parameter(s) of the menstrual pain management device 100 inferred by the neural network 600.

[0110] Following is a description of a procedure for training the neural network 600 to generate the inferred operating parameter(s) of the menstrual pain management device 100. The training procedure is implemented by the training server 400 represented in Figure 9. The training procedure is adapted to an implementation of the neural network 600 supporting step 530 of the method 500. The training procedure can be adapted by a person skilled in the art to other types of machine learning algorithms 312.[00iii]A processing unit of the training server 400 executes a neural network training engine (not represented in the Figures). The neural network training engine implements functionalities of a neural network, allowing to generate a predictive model ready to be used by the neural network 600 (executed by the processor 310 in Figure 7 and the processor 210 in Figure 8) at the end of the training, as is wellknown in the art.[00ii2]The training procedure comprises a step of initializing a predictive model used by the neural network implemented by the neural network training engine. The predictive model comprises various parameters which depend on the characteristics of the neural network implemented by the neural network training engine. The predictive model is stored in a memory of the training server 400.[00ii3]The initialization of the predictive model comprises defining a number of layers of the neural network, a functionality for each layer (e.g. input layer, fully connected layer, etc.), initial values of parameters used for implementing the functionality of each layer, etc. For example, the initialization of the parameters of a fully connected layer includes determining the number of neurons of the fully connected layer and determining an initial value for the weights of each neuron. Different algorithms (well documented in the art) can be used for allocating an initial value to the weights of each neuron. A comprehensive description of the initialization of the predictive model is out of the scope of the present disclosure, since it is well known in the art.

[0114] The training procedure comprises an initial step of generating training data. The training data comprise a plurality of instances of inputs and a corresponding plurality of instances of expected output(s). Each instance of inputs comprises a set of values for the information characterizing the user. Each corresponding set of output(s) comprises expected value(s) for the operating parameter(s) of the menstrual pain management device 100. The set of training data need to be large enough to properly train the neural network.[00ii5]The training data can be determined experimentally, using feedbacks from various users with different values for the information characterizing each user. For each user, several configurations of operating parameter(s) are tested and the feedback consists in determining which configuration is better adapted to the user.[00ii6]The training procedure comprises a step (I) of executing the neural networkimplemented by the neural network training engine, using the predictive model to generate respective instances of calculated output(s) based on the instances of inputs of the training data.[00ii7]The neural network implemented by the neural network training engine corresponds to the neural network 600 executed at step 530 of the method 500 (same number of layers, same functionality for each layer, same parameters used for implementing the functionality of each layer, etc.).[00ii8]The training procedure comprises a step (II) of adjusting the predictive model of the neural network implemented by the neural network training engine, to minimize a difference between the instances of expected output(s) and the corresponding instances of calculated output(s). For example, for a fully connected layer of the neural network, the adjustment comprises adjusting the weights associated to the neurons of the fully connected layer.

[0119] Various algorithms may be used for minimizing the difference between the expected output(s) and the calculated output(s). For example, the predictive model is adjusted so that a difference between the expected output(s) and the calculated output(s) is lower than a threshold (e.g. a difference of only 1 % is tolerated).

[0120] The aforementioned steps of the training procedure correspond to step 505 of the method 500. At the end of the training procedure, the neural network is considered to be properly trained (the predictive model of the neural network has been adjusted so that a difference between the expected output(s) and the calculated output(s) has been sufficiently minimized). The predictive model, comprising the adjusted parameters of the neural network, is transmitted to the control device (e.g. computing device 300 of Figure 7 or menstrual pain management device 100 of Figure 8), as illustrated by step 510 of the method 500. Test data are optionally used to validate the accuracy of the predictive model. The test data are different from the training data used for the training procedure.

[0121] Various techniques well known in the art of neural networks can be used forperforming step (II). For example, the adjustment of the predictive model of the neural network at step (II) uses back propagation. Other techniques, such as the usage of bias in addition to the weights (bias and weights are generally collectively referred to as weights in the neural network terminology), reinforcement learning, etc., may also be used.[00i22]Although the present disclosure has been described hereinabove by way of non-restrictive, illustrative embodiments thereof, these embodiments may be modified at will within the scope of the appended claims without departing from the spirit and nature of the present disclosure.

Claims

WHAT IS CLAIMED IS:

1. A method using a machine learning algorithm to control a menstrual pain management device, the method comprising: storing a predictive model of the machine learning algorithm in a memory of a control device; determining information characterizing a user of the menstrual pain management device; executing by a processor of the control device the machine learning algorithm, the machine learning algorithm using the predictive model for determining at least one operating parameter of the menstrual pain management device based on inputs, the inputs comprising the information characterizing the user of the menstrual pain management device; and controlling operation of the menstrual pain management device based on the at least one determined operating parameter.

2. The method of claim 1 , wherein the machine learning algorithm implements a neural network, and the predictive model comprises weights of the neural network.

3. The method of claim 1 , wherein controlling operation of the menstrual pain management device based on the at least one determined operating parameter comprises sending a control command generated based on the at least one determined operating parameter to the menstrual pain management device via a communication interface of the control device, a control module of the menstrual pain management device applying the control command to control operation the menstrual pain management device.

4. The method of claim 3, wherein the control device is a mobile computingdevice or a cloud server.

5. The method of claim 1 , wherein the control device is the menstrual pain management device, the menstrual pain management device comprises a control module in electric connection with an electrode layer of the menstrual pain management device and a heating layer of the menstrual pain management device, the control module comprising the memory and the processor, the control module controlling operation of at least one of the electrode layer and the heating layer based on the at least one determined operating parameter.

6. The method of claim 1 , wherein the information characterizing the user of the menstrual pain management device comprises information belonging to at least one of the following categories: demographic characteristics of the user, medical and gynecological conditions of the user, physical characteristics of the user and lifestyle characteristics of the user.

7. The method of claim 1 , wherein the information characterizing the user of the menstrual pain management device comprises a combination of at least some of the following information: sex, age, race, ethnicity, occupation, place of residence, presence or absence of endometriosis, primary or secondary dysmenorrhea, pregnancy history, body fat percentage, gynecological disease, menstrual regularity, use of intrauterine device, use of contraceptive pill, size, weight, body mass index, pain level, American Society of Anesthesiologists (ASA) score, use of medications and quality of life.

8. The method of claim 1 , wherein the menstrual pain management device comprises an electrode layer and a heating layer, and the at least one operating parameter of the menstrual pain management device comprises at least one of the following: one or more electric stimulation modes ofelectrodes of the electrode layer, a temperature of a heating element of the heating layer, a duration of use of the menstrual pain management device, a number of treatment cycles by the menstrual pain management device, and a frequency of treatment by the menstrual pain management device.

9. The method of claim 1 , wherein the menstrual pain management device comprises an electrode layer, and the at least one operating parameter of the menstrual pain management device comprises one or more electric stimulation modes of electrodes of the electrode layer, the one or more electric stimulation modes comprising at least one of the following: a waveform of electric impulses generated by the electrodes, a nature of the electric impulses generated by the electrodes, a pulse length of the electric impulses generated by the electrodes, a global frequency of the electric impulses generated by the electrodes, an internal frequency of the electric impulses generated by the electrodes, and an intensity of the electric impulses generated by the electrodes.

10. A computing device comprising: a communication interface; memory for storing a predictive model of a machine learning algorithm; and a processor for: determining information characterizing a user of a menstrual pain management device; executing the machine learning algorithm, the machine learning algorithm using the predictive model for determining at least one operating parameter of the menstrual pain management device based on inputs, the inputs comprising the information characterizing the user of the menstrual pain management device; andsending a control command generated based on the at least one determined operating parameter to the menstrual pain management device via the communication interface, a control module of the menstrual pain management device applying the control command to control operation of the menstrual pain management device.

11. The computing device of claim 10, wherein the machine learning algorithm implements a neural network, and the predictive model comprises weights of the neural network.

12. The computing device of claim 10, consisting of a mobile computing device or a cloud server.

13. The computing device of claim 10, wherein the information characterizing the user of the menstrual pain management device comprises information belonging to at least one of the following categories: demographic characteristics of the user, medical and gynecological conditions of the user, physical characteristics of the user and lifestyle characteristics of the user.

14. The computing device of claim 10, wherein the information characterizing the user of the menstrual pain management device comprises a combination of at least some of the following information: sex, age, race, ethnicity, occupation, place of residence, presence or absence of endometriosis, primary or secondary dysmenorrhea, pregnancy history, body fat percentage, gynecological disease, menstrual regularity, use of intrauterine device, use of contraceptive pill, size, weight, body mass index, pain level, American Society of Anesthesiologists (ASA) score, use of medications and quality of life.

15. The computing device of claim 10, wherein the menstrual pain management device comprises an electrode layer and a heating layer, and the at leastone operating parameter of the menstrual pain management device comprises at least one of the following: one or more electric stimulation modes of electrodes of the electrode layer, a temperature of a heating element of the heating layer, a duration of use of the menstrual pain management device, a number of treatment cycles by the menstrual pain management device, and a frequency of treatment by the menstrual pain management device.

16. The computing device of claim 10, wherein the menstrual pain management device comprises an electrode layer, and the at least one operating parameter of the menstrual pain management device comprises one or more electric stimulation modes of electrodes of the electrode layer, the one or more electric stimulation modes comprising at least one of the following: a waveform of electric impulses generated by the electrodes, a nature of the electric impulses generated by the electrodes, a pulse length of the electric impulses generated by the electrodes, a global frequency of the electric impulses generated by the electrodes, an internal frequency of the electric impulses generated by the electrodes, and an intensity of the electric impulses generated by the electrodes.

17. A non-transitory computer readable medium comprising instructions executable by a processor of a computing device, the execution of the instructions by the processor of the computing device providing for using a machine learning algorithm to control a menstrual pain management device by: storing by the processor a predictive model of the machine learning algorithm in a memory of the computing device; determining by the processor information characterizing a user of the menstrual pain management device; executing by the processor the machine learning algorithm, the machine learning algorithm using the predictive model for determining atleast one operating parameter of the menstrual pain management device based on inputs, the inputs comprising the information characterizing the user of the menstrual pain management device; and sending by the processor a control command generated based on the at least one determined operating parameter to the menstrual pain management device via a communication interface of the computing device, a control module of the menstrual pain management device applying the control command to control operation of the menstrual pain management device.