Matching of inhalers to patient vocal tract for optimization of particle dispersion
11545249 · 2023-01-03
Assignee
Inventors
Cpc classification
A61M2205/6009
HUMAN NECESSITIES
G16Z99/00
PHYSICS
A61M2205/3375
HUMAN NECESSITIES
A61M2205/3553
HUMAN NECESSITIES
A61M2205/505
HUMAN NECESSITIES
International classification
Abstract
An optimization methodology is employed to match vibratory inhaler devices having certain characteristics to the particular anatomical and acoustic properties of a patient's vocal tract, in order to achieve the most effective dispersion of a dry powder medicament using inspiratory effort of a user of the inhaler. In embodiments, optimization involves employing one or more measurements of acoustic frequency spectrum properties as well as one or more anatomical/geometric measurements of the structures comprising the particular patient's mouth, pharynx, and upper respiratory tract and matching a vibratory inhalation device that corresponds thereto.
Claims
1. A device selected to optimize drug particle deposition, the device comprising: a vibratory dispersal element extending across at least a portion of an inspiratory flow path for dispersing drug particles into a respirable aerosol plume, wherein the vibratory dispersal element creates acoustic vibrations having a frequency spectrum that corresponds to acoustic properties of a respiratory flow path of a patient, wherein the acoustic vibrations having the frequency spectrum that corresponds to the acoustic properties of the respiratory flow path of the patient reduces coagulation of the drug particles; and a spacer attached to the device that modifies one or more of size or geometry of a cavity that is associated with an interior space of the device, wherein the one or more of the size or the geometry reduces coagulation of the drug particles.
2. The device of claim 1, further comprising a dry powder inhaler or a dry powder nebulizer.
3. The device of claim 1, wherein a configuration of the device is based on patient data that includes acoustic reflectometry data for the patient, wherein the acoustic reflectometry data includes a sensitivity of the patient to the frequency spectrum produced by the device, and wherein the configuration of the device defines size and geometry of the device.
4. The device of claim 3, wherein the configuration of the device is further based on a predetermined correlation between the patient data and a rate of drug deposition in a deep lung portion of the respiratory flow path of the patient.
5. The device of claim 1, wherein a configuration of the device is based on patient data that includes spirometry data, wherein the spirometry data includes a calculated value for one or more of a forced expiratory volume one-second (FEV1), a forced vital capacity (FVC), or the FEV1 as a percent of the FVC (FEV1% FVC), and wherein the configuration of the device defines size and geometry of the device.
6. The device of claim 5, wherein the configuration of the device is further based on a predetermined correlation between the patient data and a rate of drug deposition in a deep lung portion of the respiratory flow path of the patient.
7. The device of claim 1, wherein the inspiratory flow path for dispersing the drug particles into the respirable aerosol plume includes a deagglomerating portion formed in the device, wherein the deagglomerating portion has a cross-sectional area that is equal to or less than three centimeters squared.
8. The device of claim 7, wherein the deagglomerating portion of the inspiratory flow path further has a length that is equal to or less than about three centimeters.
9. A device selected to optimize drug particle deposition, the device comprising: a vibratory dispersal element extending across at least a portion of an inspiratory flow path for dispersing drug particles into a respirable aerosol plume, wherein the vibratory dispersal element creates acoustic vibrations having a frequency spectrum that corresponds to acoustic properties of a respiratory flow path of a patient, wherein the acoustic vibrations having the frequency spectrum that corresponds to the acoustic properties of the respiratory flow path of the patient reduce coagulation of the drug particles within the respirable aerosol plume and increases deposition of the drug particles in a deep lung portion of the respiratory flow path of the patient; and a spacer attached to the device that modifies one or more of size or geometry of an interior cavity that is associated with an interior space of the device, wherein the one or more of the size or the geometry reduces coagulation of the drug particles.
10. The device of claim 9, further comprising a dry powder inhaler or a dry powder nebulizer.
11. The device of claim 10, wherein the spacer extends from a bounding surface into the interior space of the device, wherein the spacer extends into the inspiratory flow path, wherein the spacer extends orthogonal to a direction of the inspiratory flow path, and wherein the spacer is selected based on the acoustic properties of the respiratory flow path of the patient.
12. The device of claim 10, further comprising a turbulence promoter located within the inspiratory flow path, wherein the turbulence promoter creates at least one vertex having an axis of rotation that is substantially orthogonal to a direction of the inspiratory flow path.
13. The device of claim 10, wherein the vibratory dispersal element is configured to generate one or more frequencies in a range of 40 Hertz (Hz) to 2100 Hz, or one or more frequencies in a range of 8 Hz to 250 Hz.
14. A device selected to optimize drug particle deposition, the device comprising: a vibratory dispersal element extending across at least a portion of an inspiratory flow path for dispersing drug particles into a respirable aerosol plume, wherein the vibratory dispersal element is configured to generate non-linear acoustic vibrations having a range of 10 Hz to 2000 Hz, wherein the non-linear acoustic vibrations reduce coagulation of the drug particles within the respirable aerosol plume and increases deposition of the drug particles in a deep lung portion of a respiratory flow path of a patient; and a spacer attached to the device that modifies one or more of size or geometry of an interior cavity that is associated with an interior space of the device, wherein the one or more of the size or the geometry reduces coagulation of the drug particles.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1) The present invention is described in detail below with reference to the attached drawing figures, wherein:
(2)
(3)
(4)
(5)
(6)
(7)
DETAILED DESCRIPTION
(8) The subject matter of the present invention is described with specificity herein to meet statutory requirements. However, the description itself is not intended to limit the scope of this patent. Rather, the inventors have contemplated that the claimed subject matter might also be embodied in other ways, to include different steps or combinations of steps similar to the ones described in this document, in conjunction with other present or future technologies. Moreover, although the terms “step” and/or “block” may be used herein to connote different components of methods employed, the terms should not be interpreted as implying any particular order among or between various steps herein disclosed unless and except when the order of individual steps is explicitly described.
(9) Referring to the drawings in general, and initially to
(10) The present invention may be operational with numerous other general purpose or special purpose computing system environments or configurations. Examples of well-known computing systems, environments, and/or configurations that may be suitable for use with the present invention include, by way of example only, personal computers, server computers, hand-held or laptop devices, multiprocessor systems, microprocessor-based systems, set top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, distributed computing environments that include any of the above-mentioned systems or devices, and the like.
(11) The present invention may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include, but are not limited to, routines, programs, objects, components, and data structures that perform particular tasks or implement particular abstract data types. The present invention may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in local and/or remote computer storage media including, by way of example only, memory storage devices.
(12) With continued reference to
(13) The server 22 typically includes, or has access to, a variety of computer readable media, for instance, database cluster 24. Computer readable media can be any available media that may be accessed by server 22, and includes volatile and nonvolatile media, as well as removable and non-removable media. By way of example, and not limitation, computer readable media may include computer storage media and communication media. Computer storage media may include, without limitation, volatile and nonvolatile media, as well as removable and nonremovable media implemented in any method or technology for storage of information, such as computer readable instructions, data structures, program modules, or other data. In this regard, computer storage media may include, but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVDs) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage, or other magnetic storage device, or any other medium which can be used to store the desired information and which may be accessed by the server 22. Communication media typically embodies computer readable instructions, data structures, program modules, or other data in a modulated data signal, such as a carrier wave or other transport mechanism, and may include any information delivery media. As used herein, the term “modulated data signal” refers to a signal that has one or more of its attributes set or changed in such a manner as to encode information in the signal. By way of example, and not limitation, communication media includes wired media such as a wired network or direct-wired connection, and wireless media such as acoustic, RF, infrared, and other wireless media. Combinations of any of the above also may be included within the scope of computer readable media.
(14) The computer storage media discussed above and illustrated in
(15) The server 22 may operate in a computer network 26 using logical connections to one or more remote computers 28. Remote computers 28 may be located at a variety of locations in a medical or research environment, for example, but not limited to, clinical laboratories, hospitals and other inpatient settings, veterinary environments, ambulatory settings, medical billing and financial offices, hospital administration settings, home health care environments, and clinicians' offices. Clinicians may include, but are not limited to, a treating physician or physicians, specialists such as surgeons, radiologists, cardiologists, and oncologists, emergency medical technicians, physicians' assistants, nurse practitioners, nurses, nurses' aides, pharmacists, dieticians, microbiologists, laboratory experts, genetic counselors, researchers, veterinarians, students, and the like. The remote computers 28 may also be physically located in non-traditional medical care environments so that the entire health care community may be capable of integration on the network. The remote computers 28 may be personal computers, servers, routers, network PCs, peer devices, other common network nodes, or the like, and may include some or all of the components described above in relation to the server 22. The devices can be personal digital assistants or other like devices.
(16) Exemplary computer networks 26 may include, without limitation, local area networks (LANs) and/or wide area networks (WANs). Such networking environments are commonplace in offices, enterprise-wide computer networks, intranets, and the Internet. When utilized in a WAN networking environment, the server 22 may include a modem or other means for establishing communications over the WAN, such as the Internet. In a networked environment, program modules or portions thereof may be stored in the server 22, in the database cluster 24, or on any of the remote computers 28. For example, and not by way of limitation, various application programs may reside on the memory associated with any one or more of the remote computers 28. It will be appreciated by those of ordinary skill in the art that the network connections shown are exemplary and other means of establishing a communications link between the computers (e.g., server 22 and remote computers 28) may be utilized.
(17) In operation, a user may enter commands and information into the server 22 or convey the commands and information to the server 22 via one or more of the remote computers 28 through input devices, such as a keyboard, a pointing device (commonly referred to as a mouse), a trackball, or a touch pad. Other input devices may include, without limitation, microphones, satellite dishes, scanners, or the like. Commands and information may also be sent directly from a remote healthcare device to the server 22. In addition to a monitor, the server 22 and/or remote computers 28 may include other peripheral output devices, such as speakers and a printer.
(18) Although many other internal components of the server 22 and the remote computers 28 are not shown, those of ordinary skill in the art will appreciate that such components and their interconnection are well known. Accordingly, additional details concerning the internal construction of the server 22 and the remote computers 28 are not further disclosed herein.
(19) Health data is often stored in person-centric health records, such that each individual has a health record. The term “health record” is not meant to be limited to any particular type of unified record that stores health data for an individual, examples of which include an electronic medical record (EMR), electronic health record (EHR), personal health record (PHR), continuity of care record (CCR), among others. Furthermore, the terms “patient”, “individual”, “person” and “subject” may be used interchangeably herein to refer to someone who has an associated health record (patient record).
(20) As previously mentioned, embodiments of the present invention relate to optimization of the matching of vibratory inhaler devices to the particular anatomical and acoustic properties of the patient's mouth, pharynx, and upper respiratory tract, in order to achieve the most effective dispersion of aerosolized active drug particles (i.e., powder) in a respirable form, using inspiratory effort of a user of the inhaler. By respirable form, it is meant that the respirable particles are in a deaggregated form and that, upon being dispensed by an inhaler device, the particles will be dispensed so that the active agent present in the dose is in the form of particles or particle agglomerates of respirable size (preferably having a diameter between 0.5 μm and 10 μm) at the time of aerosolization and, furthermore, that the particle size distribution remains predominantly in this size range during transit through the patient's airways. The aerosolized drug in respirable form may include small agglomerates of particles, as are present in an ordered mixture, for example where the carrier material, to which the fine particles adhere. Any agglomerates present in the powder in respirable form may be broken up by the turbulence created upon actuation of the inhaler or created by the flow of the inspired air.
(21) The respirable nature of the resultant powder may be ascertained by ACI (Anderson Cascade Impactor). In some embodiments, it is optimal for the respirable powder to have a Fine Particle Fraction (FPF) of greater than 50%. Utilizing embodiments of the present invention, a significant improvement of the FPF as compared to the FPF prior to the selection of the proper size vibratory inhaler (or flow-path modifying insert or attachment to the vibratory inhaler designed to maximize deep lung drug particle deposition. The regional deposition of drug in deep pulmonary tissues may be measured through the use of radiolabeled drug powder and quantitative gamma scintigraphy, SPECT (usually by co-spray drying with .sup.99mTc-DTPA), or PET (using .sup.11C- or .sup.18F-labeled drug), as are familiar to those practiced in the art. The latter two techniques involve short half-lives that are typically not well suited to studying dry powder inhalers, although they may be used in nebulizer and MDI and other inhaler types.
(22) For optimal vibratory inhaler function, the step of breaking up compacted powder involves agitating the powder with force sufficient to break up the compacted powder. The force is preferably applied once the drug has been placed into a proper receptacle of the vibratory inhaler and more preferably once the receptacle has been sealed, thereby ensuring that the agitation step does not result in any appreciable loss of the measured dose of drug. Receptacles are containers, such as blisters and capsules, into which measured or metered doses of dry powder formulations are placed for storage and from which the dose of drug may be dispensed by an inhaler device.
(23) Agitating a measured dose of a drug formulation in a sealed receptacle is accomplished by indirectly applying a vibrational force to the powder, for example, via the air within the sealed receptacle. In selecting an optimal vibratory inhaler configuration, it is contemplated that extensions or attachments or ‘insert components’ that can alter the geometry of the inhaler device may be considered, so as to accomplish more complete deagglomeration of the drug formulation upon inspiration by a user as well as minimize the rate of subsequent coagulation and precipitation or impaction of the particles in the upper airway and mouth.
(24) In certain aspects, the present invention involves collecting acoustic frequency spectrum properties and anatomical/geometric measurements of the structures comprising a particular patient's mouth, pharynx, and upper airway. Time-domain acoustic reflectometry is utilized, a method that is capable not only of acquiring the necessary anatomical measurements but can assess the sensitivity to acoustic frequency spectral properties as well.
(25) As one option, optimization of vibratory inhaler properties may be carried out using a statically configured ‘spacer’ or ‘insert’ attached to the exterior of the inhaler device so as to modify and augment the inhaler's enclosed cavity size and the geometry of the inhaler device's inspiratory airflow path, to create a flow pattern within the device that does not result in undue loss of particles within the device itself and is not associated with acoustically exacerbated agglomeration or excessive deposition of particles in the mouth or upper airway. Additionally, deagglomerating drug formulation in an inhaler device is preferably done so as to generate some acoustic emissions into the patient's airway, either directly or indirectly. For instance, employing an inhaler with a vibratory dispersal mechanism extending across at least a portion of an inspiratory flow path enables the generation of a respirable aerosol plume.
(26) Decision-support methods, in another aspect, are also employed. The decision-support methods utilize an associative neural network (ANN), or alternatively a logistic regression statistical model, arranged so as to ascertain which inhaler device or inhaler configuration best matches the device to the patient's anatomical and upper-airway acoustic properties and thereby achieve the optimal alveolar deposition of drug.
(27) In certain embodiments, the airflow path has a deagglomerating portion possessing a cross-sectional area that is about two centimeters squared (2 cm.sup.2) or less. Additionally, in certain embodiments, the deagglomerating portion of the inspiratory airflow path can have a cross-sectional width that is about 2 cm or less and may have a length that is less than about 3 cm.
(28) As another configuration option, a vibratory inhaler designed to maximize deep lung deposition of drug particles may include: (a) an inhaler body with an inspiratory flow path therein; and (b) at least one turbulence promoter residing in the inspiratory flow path, creating flow vortices have an axis of rotation that is substantially orthogonal to the inspiratory flow direction. Ideally, when viewed in transverse cross section, an attachment or insert extends from a bounding surface a distance into the inspiratory airflow path, the distance being a sub-portion of a cross-sectional width of the inspiratory path whereby the inhaler's vibratory mechanism deagglomerates the dry powder in response to inspiratory effort by a user.
(29) Examples of diseases, conditions or disorders that may be treated with vibratory inhalers optimally matched to a patient's acoustic and anatomical configuration include, but are not limited to, asthma, COPD (chronic obstructive pulmonary disease), viral or bacterial infections, influenza, allergies, cystic fibrosis, and other respiratory ailments as well as diabetes and other insulin resistance disorders. Vibratory inhalers, particularly in delivering dry powder medicaments, may be used to deliver locally-acting agents such as antimicrobials, protease inhibitors, and nucleic acids/oligionucleotides as well as systemic agents such as peptides like leuprolide and proteins such as insulin. For example, inhaler-based delivery of antimicrobial agents such as antitubercular compounds, proteins such as insulin for diabetes therapy or other insulin-resistance related disorders, peptides such as leuprolide acetate for treatment of prostate cancer and/or endometriosis and nucleic acids or oligonucleotides for cystic fibrosis gene therapy may be performed.
(30) Typical dose amounts of the unitized dry powder mixture dispersed in the inhalers may vary depending on the patient size, the systemic target, and the particular drug(s). A conventional exemplary dry powder dose amount for an average adult is less than about 50 mg, typically between about 10-30 mg and for an average adolescent pediatric subject is typically from about 5-10 mg. A typical dose concentration may be between about 1-2%. Exemplary dry powder drugs include, but are not limited to, albuterol, fluticasone, beclamethasone, cromolyn, terbutaline, fenoterol, β-agonists (including long-acting β-agonists), salmeterol, formoterol, cortico-steroids and glucocorticoids. In certain embodiments, the administered bolus or dose can be formulated with an increase in concentration (an increased percentage of active constituents) over conventional blends. Further, the dry powder formulations may be configured as a smaller administrable dose compared to the conventional 10-25 mg doses. For example, each administrable dry powder dose may be on the order of less than about 60-70% of that of conventional doses. As one example, using the active dispersal systems provided by certain configurations of vibratory inhalers suggested by embodiments of the present invention, an adult dose may be reduced to under about 15 mg, such as between about 10 μg-10 mg, and more typically between about 50 μg-10 mg per puff. The active constituent(s) concentration may be between about 5-10%. In other embodiments, active constituent concentrations can be in the range of between about 10-20%, 20-40%, or even larger. In the case of nebulizer devices, the amount of drug delivered can be much larger, often hundreds of milligrams, and is achieved over a plurality of respiratory cycles or breaths instead of in just one cycle or puff.
(31) In certain embodiments, during dose dispensing, the dry powder in a particular drug compartment or blister may be formulated in high concentrations of an active pharmaceutical constituent(s) substantially without additives (such as excipients). As used herein, ‘substantially without additives’ means that the dry powder is in a substantially pure active formulation with only minimal amounts of other non-biopharmacological active ingredients. The term ‘minimal amounts’ means that the non-active ingredients may be present, but are present in greatly reduced amounts, relative to the active ingredient(s), such that they comprise less than about 10%, and preferably less than about 5%, of the dispensed dry powder formulation, and, in certain embodiments, the non-active ingredients are present in only trace amounts.
(32) Utilizing the methodologies of various embodiments of the present invention, vibratory inhalers may be selected to have, when analyzed in a steady state flow, at least one acoustic vibrational mode can be generated in an inspiratory flow direction in an inspiratory airflow path, as an amount of dry powder travels through the inhaler and upper respiratory tract upon patient inspiration. This facilitates deagglomeration of inhaled particles and inhibits spontaneous agglomeration (coagulation) of particles, thereby preventing trapping of undue amounts of the particles in the inhaler or in the upper respiratory tract or in the mouth during inhalation.
(33) Accordingly, the vibratory signal can include a carrier frequency that may be between about 50 Hz to about 2000 Hz, and typically is between about 100 Hz-1000 Hz. The carrier frequency may be modified by one or more low modulating frequencies (typically between about 10-200 Hz). The frequency of the vibration can be modified to match or correspond to the flow characteristics of the dry powder substance held in a package to attempt to reach a resonant frequency(s) to promote uniform drug dispersion into the body. In some embodiments, a non-linear powder-specific dry powder vibratory energy signal comprises a plurality of selected frequencies that can be generated (corresponding to the particular dry powder(s) being currently dispensed) to output the particular signal corresponding to the dry powder(s) then being dispensed. As used herein, the term ‘non-linear’ means that the vibratory action or signal applied to the package to deliver a dose of dry powder to a user has an irregular shape or cycle, typically employing multiple superimposed frequencies, and/or a vibratory frequency line shape that has varying amplitudes (peaks) and peak widths over typical standard intervals (per second, minute, etc.) over time. In contrast to conventional systems, the non-linear vibratory signal input can operate without a fixed single or steady state repeating amplitude at a fixed frequency or cycle. This non-linear vibratory input can be applied to the blister to generate a variable amplitude motion (in either a one, two and/or three-dimensional vibratory motion). The non-linear signal fluidizes the powder in such a way that a powder ‘flow resonance’ is generated allowing active flow able dispensing. For instance, in one arrangement, a non-linear vibratory signal can include a carrier frequency that is between 1000 Hz and 40,000 Hz.
(34) In certain embodiments, a signal of combined frequencies can be generated to provide a non-linear signal to improve fluidic flow performance. Selected frequencies can be superimposed to generate a single superposition signal (that may also include weighted amplitudes for certain of the selected frequencies or adjustments of relative amplitudes according to the observed frequency distribution). Thus, the vibratory signal can be a derived non-linear oscillatory or vibratory energy signal used to dispense a particular dry powder. In certain embodiments, the output signal used to activate the piezoelectric blister channel may include a plurality of superpositioned modulating frequencies (‘overtones’; ‘harmonics’) and a selected carrier frequency.
(35) Exemplary vibratory elements for an inhaler to accomplish moving dry powder into an airflow path include, but are not limited to, one or more of: (a) ultrasound or other acoustic or sound-based sources (above, below or at audible wavelengths) that can be used to instantaneously apply non-linear pressure signals onto the dry powder; (b) electrical or mechanical deflection of the sidewalls and/or floor of the inhalation flow channel and/or drug compartment, which can include magnetically induced or caused vibrations and/or deflections (which can use electro or permanent field magnets); (c) solenoids, piezoelectrically active portions and the like; (d) oscillating or pulsed gas (airstreams), which can introduce changes in one or more of volume flow, linear velocity, and/or pressure; and (e) aeroelastic films, which can apply non-linear pressure signals onto the dry powder as vibrations are passively generated by energy transferred from the patient's inspired airstream. In some particular embodiments, the vibrator may be ‘active’ and include one or more electrically-powered piezoelectric elements (such as a piezoceramic component, or a piezoelectric polymer film), or it may be ‘passive’ and include an aeroelastic element or polymer film whose motion is generated by energy of the user's inhalation inspiratory effort. Furthermore, in some embodiments, the vibratory element can be configured to vibrate the drug compartment holding the dry powder.
(36) In one aspect, the agitation involves applying a vibrational force using a vibrational means, wherein the vibrational force is not applied directly to the receptacle at the point where the powder contacts the receptacle.
(37) The vibrational force may be provided in the form of sonication, including acoustic and ultrasound agitation (including resonant frequency matching), shaking, impacts and percussion effects. In each case, the vibration may be applied to the outside of the sealed receptacle, and is communicated to the powder compact held inside the receptacle, preferably through the air in the receptacle (in preference to the vibration being communicated through the body of the receptacle, at least part of which will be in direct contact with the powder compact.
(38) As those of skill in the art appreciate, various elements may be used to focus or transfer the vibrational forces, such as an acoustic lens, or transmitting media to improve contact with the sealed receptacle. Additionally, various elements may be used to apply the vibrational force may be an acoustic lens or a piezoelectric element or an aeroelastic film. Exemplary agitation techniques include applying a vibrational force to the powder in the receptacles at frequencies of less than about 1 MHz. Preferably, the frequency is from about 100 Hz to about 500 KHz, from about 1 KHz to about 200 KHz, from about 4 KHz to about 70 KHz, or from about 20 KHz to about 40 KHz. As a matter of illustration and not limitation, the agitation may be provided by contacting the filled receptacle with an ultrasonic probe, for example a probe which is operating at a frequency range of between about 1 KHz and about 200 KHz.
(39) The amplitude of oscillation of the vibrational force is also relevant to particle movement. It has been found that application of a vibrational force with a particular amplitude to a receptacle improves the breakup of powder compacts and can assist in subsequent emptying of the receptacle.
(40) In one aspect, the amplitude of the vibrational force should be between about 10 to about 100%, more preferably from about 50 to about 100%. Especially preferred amplitudes range from about 75 to about 100%, from about 80 to about 100%, from about 85 to about 100% or from about 90 to about 100%. Those of skill in the art would have no difficulty in ascertaining the suitable amplitude in order to break up powder compact, based upon the nature of the powder, the nature of the compacts formed during receptacle filling and the nature of the receptacle.
(41) The pressure with which the vibrational force is applied to the receptacle has been found to be an important factor in achieving effective break up of powder compacts and improving the emptying of the receptacle upon actuation of the inhaler. Exemplary pressures include ranges from about 0.1 to about 1.5 bar, from about 0.2 to about 1.2 bar and most preferably from about 0.2 to about 0.6 bar. The value of the pressure parameter lies in its application to the enclosed environment with the patient's lips sealing the orifice of the inhaler, through which the inspired jet of air is drawn.
(42) Finally, the duration for which the vibrational force applied to the receptacle has been found to also be an important factor in achieving effective break up of powder compacts and improving the emptying of the receptacle.
(43) It has been found that optimal results are achieved if the vibrational force is applied for between about 0.10 and 5 seconds. The vibrational force should be applied for long enough to allow complete break up or deaggregation of the compacted powder, but must not be so long as to have any detrimental effects on the powder, for example by causing ordered mixtures to segregate. Preferably, the vibrational force is applied for between about 0.25 to about 1 second. The duration may need to be adjusted, depending upon the other parameters. For example, a lower frequency vibrational force may need to be applied for longer in order to have the desired effect.
(44) Turning now to
(45) With respect to
(46) Thereafter, at block 306, quantitative spirometry is performed on the subject and, at a minimum, forced expiratory volume 1-second (FEV1), forced vital capacity (FVC), and FEV1 as a percent of FVC are measured or calculated. At block 308, time-domain acoustic reflectometric laryngometry is performed on the subject, and, at a minimum, the distance to the point of maximal narrowing (Lmin), the distance to the glottis (Lglot), and the maximal airway area (Amax) are measured, preferably with time-domain reflectometry acoustic frequency spectrum in the same range as frequencies emitted by the relevant alternative inhaler/nebulizer (vibratory inhaler) devices. For instance a TDR apparatus available from Sleep Group Solutions Inc. of North Miami Beach, Fla. was employed in one implementation. However, similar clinical acoustic TDR devices could likewise be used.
(47) At block 310, one of a plurality of relevant inhaler types and configurations (e.g., with or without spacer, with spacers of different length and bore, etc.) is selected for testing. At block 312, quantitative 2-dimensional or 3-dimensional scintigraphy is performed following the subject's inhaling radiolabelled test aerosol using the selected inhaler device. Associated scattering corrections and integration of pixel/voxel intensities is performed to calculate percentage dose deposition in deep lung locations. Data is captured from blocks 304 to 312 in the knowledge base.
(48) At block 314, after a suitable time has elapsed (depending on the half-life of the radionuclide used for the foregoing scintigraphy) a determination is made as to whether to repeat the study on the same subject using a different inhaler type or inhaler configuration. If so, then the method returns to block 310 and the process of selecting an inhaler type/configuration is repeated. Otherwise, if the study is not to be repeated on the subject at this time, then it is determined at block 316 whether additional study subjects should be examined in building the knowledge base. If so, then the method returns to block 302 where a candidate study subject is selected and the process repeated. Otherwise, if no additional study subjects are to be examined, then, at block 318, the knowledge base of correlations between patient and inhaler attributes and drug particular deposition is established.
(49) Turning to
(50) Initially, inclusion-exclusion criteria is defined, as shown at block 402, as well as problem specification in terms of available input and output variables, at block 404. producing and validating a statistical model for accurately predicting patient attributes that match given inhaler attributes to achieve deep lung drug deposition
(51) Thereafter, as shown at block 406, training data is collected. Training data comprises a set of data points having known characteristics. This data may come from research facilities, academic institutions, commercial entities, and/or other public or confidential sources The collection of training data may be accomplished manually or by way of an automated process, such as known electronic data transfer methods. Accordingly, an exemplary embodiment of the learning machine for use in conjunction with the present invention may be implemented in a networked computer environment.
(52) With reference again to block 402, it is known to those practiced in the art that to construct an effective classifier, appropriate inclusion-exclusion criteria is first defined in sufficient detail that the cases acquired for the purpose of classifier design accurately represent the population to which the classifier is intended to be applied.
(53) With reference again to block 404, for the cohort meeting the applicable inclusion-exclusion criteria, database retrieval of extant electronic medical records is performed. This serves to define the available input and output clinical and laboratory variables and characterize the descriptive statistics of each variable and assess the degree of “missingness” of information for each variable. In one embodiment, variables whose values are missing at a greater than 20% rate are excluded from subsequent consideration in classifier construction and development. It should be understood that although database retrieval of electronic medical records is described, any type of patient medical or health record may be utilized within the various embodiments of the present invention (in the context of method 400 or in other contexts of embodiments of the present invention).
(54) Next, at block 406, information for the qualifying cases for each of the selected variables is extracted from the electronic medical record or other data source, including the date-time stamp for each item. As shown at block 408, the retrieved cases and case information are partitioned into two subsets—a first subset that is to be utilized for classifier construction and training (training data subset), and a second subset that is to be used for classifier validation testing (test data subset). Any of a variety of partitioning methods can be employed such as are well-known to statisticians practiced in the art. Randomized ‘bootstrap’ sampling without replacement, for example, may be used to insure that the subsets that are generated are not biased with regard to time, source institutions, or other factors. In some embodiments, the partitioning is made into two subsets of equal size (50%-50%). However, there is no requirement that this be the case. The subsets can be of different sizes. In some embodiments, the sample size of each subset is sufficient to achieve a desired 80% or greater statistical power for classification of the cases.
(55) As shown at block 410, statistical pre-processing is performed, including calculation of mean, median, standard deviation, skewness, and kurtosis for each of the numerical variables and frequency tables for each of the categorical variables. In instances where the statistical distribution of a numerical variable is markedly skewed, then logarithmic or power-law or other transformation of that variable is performed by methods that are well-known to statisticians, so as to produce a distribution for the transformed variable that is symmetrical and more nearly Gaussian in shape than that of the raw variable. The collected training data is optionally pre-processed in order to allow the learning machine to be applied most advantageously toward extraction of the knowledge inherent in the training data. During this preprocessing stage, a variety of different transformations can be performed on the data to enhance its usefulness. Such transformations, examples of which include addition of expert information, spline conversion, logarithmic or power-law transformations, etc., will be readily apparent to those of skill in the art. However, the preprocessing of interest in an embodiment of the present invention is the reduction of dimensionality by way of feature selection.
(56) The resulting dataset is processed with associative neural network (ANN), as shown at block 412. The training data subset is used to condition the ANN coefficients and generate ANN layers, nodes and weights (at block 416) that optimally distinguish the cases according to the dependent variable.
(57) As shown at block 414, the ANN is trained using the pre-processed data from the training data subset. Accordingly, the ANN is trained by adjusting its operating parameters until a desirable training output is achieved. The determination of whether a training output is desirable may be accomplished either manually or automatically by comparing the training output to the known characteristics of the training data. The ANN is considered to be trained when its training output is within a predetermined error threshold from the known characteristics of the training data.
(58) At block 418, the resulting classification table is examined by available receiver-operating characteristic (ROC) statistical methods, to assess whether the classifier generated by the ANN meets the design requirements established for the predictive model. In the event that ROC is lower than the acceptable minimum (e.g., a minimum ROC area-under-the-curve (C-statistic) of a particular value was not met), then additional iterations through blocks 310-318 are performed. Alternatively, if ROC is acceptable at block 418, then the ANN is accepted and the model is validation-tested, as shown at block 420, using the test data subset that was previously prepared and reserved at block 408.
(59) Based on the post-processed test output, it is determined through block 422 whether an optimal minimum was achieved by the trained ANN. If it is determined that the optimal minimum has not been achieved by the trained ANN at block 322, then the method returns to block 410, and ANN selection is readjusted (blocks 410-420). When it is determined in step 422 that the optimal minimum has been achieved, the validated classifier model is accepted for implementation, at block 424.
(60) With respect to
(61) At block 506, if spirometry data have not already been contemporaneously acquired, then a new spirometry procedure is conducted and data entered. Otherwise, recent FEV1, FVC, and FEV1% FVC values can be retrieved from the medical record and loaded into the DSS. At block 508, if acoustic reflectometry data (TDR) have not already been contemporaneously acquired, then a new acoustic reflectometry (TDR) procedure is performed and data entered. Otherwise, recent Lmin, Lglot, and Amax values can be retrieved from the medical record and loaded into the DSS. At block 510, any newly-acquired spirometry and/or acoustic reflectometry data are stored in the medical record, and the set of patient variables is presented to the ANN. At block 512, the ANN algorithm computes the matching of the current patient's data to the previously acquired knowledge base of test subject demographic, clinical, spirometric, acoustic reflectometry (TDR), and scintigraphic data. A Hopfield network or other pattern-recognition neural network may be utilized, as will be apparent to those knowledgeable in the field of the invention. As one example, an ANN implemented in Java using the open-source JOONE software environment was utilized. However, any of a variety of neural network software packages would serve equally well for this purpose. The output of the ANN algorithm identifies one or a plurality of the closest matches and the associated percent-deep-lung deposition values. A determination is then made in step 514 as to whether a preferable match of patient to inhaler is found, based on the output. If a preferable match is found, then at block 518, the clinician (e.g., prescribing physician) is then able to initiate an order or prescription for the device or configuration that is optimal for the individual and likeliest to achieve an optimal delivery of aerosolized drug to the alveoli. Alternatively, if a preferable match is not found, then at block 516, routine care is continued as clinically indicated.
(62) An embodiment of the present invention will now be described with reference to
(63) As shown in
(64) The screen display also includes a decision support patient criteria area 604. The patient criteria area 604 allows a clinician to enter the input data set that is used in conjunction with the statistical model developed, for instance, using the method 400 described above with reference to
(65) The screen display 600 also displays information 606 regarding the recommended vibratory inhaler, in this case a dry powder inhaler (DPI), determined based on the input data in the patient criteria area 604 and utilizing the statistical model, in this case the associative neural network (ANN). The recommended inhaler in this case, having certain physical and acoustic properties, is associated with an integer number. Each variation in vibratory inhaler type and configuration recommended is therefore associated with a unique integer number. However, those of skill in the art appreciate that other identification systems (besides integers) may be utilized to reference various inhaler types and configurations suitable for recommending to persons with a variety of respiratory tract acoustic and physiological configurations.
(66) Exemplary System Implementation
(67) Multi-variable pattern matching, logistic regression, discriminant analysis, cluster analysis, decision-tree induction, or other methods known to those skilled in the art are utilized to establish, for each inhaler device model and configuration, the correlations between upper respiratory tract acoustic variables, spirometry respiratory physiologic variables, patient demographic variables, and the percentage deposition of drug powder particles in the deep lung as measured by quantitative nuclear medicine imaging such as gamma scintigraphy, SPECT, or PET. In one embodiment, the statistical predictive model utilizes the following variables: patient age in years, gender, height in cm, FEV1 in L, FVC in L, FEV1/FVC ratio as %, the distance from the incisor teeth to the first minimum airway cross-section in cm, and the maximum areal dimension in the mouth/oropharynx in square centimeters (cm.sup.2) as measured by acoustic reflectometry data.
(68) From the foregoing, it will be seen that this invention is one well adapted to attain all the ends and objects set forth above, together with other advantages which are obvious and inherent to the system and method. It will be understood that certain features and subcombinations are of utility and may be employed without reference to other features and subcombinations. This is contemplated and within the scope of the claims.