Machine learning systems and methods for locating potential patients and / or healthcare providers

Abstract

According to a first aspect of the present disclosure, a computer-implemented method for identifying potential disease diagnoses and / or healthcare providers is described. The method includes receiving geofencing data from one or more healthcare delivery locations, the geofencing data related to a disease and including one or more impressions made at the healthcare delivery locations and / or one or more clicks made at the healthcare delivery locations; extracting a plurality of features from the geofencing data for each of a plurality of entities, each entity associated with a respective healthcare delivery location in the one or more healthcare delivery locations; processing the extracted features for each entity using a machine learning model to determine an indication of whether a patient with the potential disease is present at a respective healthcare delivery location associated with the entity and / or whether a healthcare provider at a respective healthcare delivery location associated with the entity is potentially seeking information related to the disease; and triggering an alert in response to determining an indication of whether a patient with the potential disease is present at a healthcare delivery location in the one or more healthcare delivery locations and / or whether a healthcare provider at a respective healthcare delivery location associated with the entity is potentially seeking information related to the disease, wherein the alert includes identification information of the healthcare delivery locations and / or the entities associated with the healthcare delivery locations.

Description

[Technical Field]

[0001] The present specification relates to systems and methods for locating potential patients and / or healthcare providers from geofencing data using machine learning models. [Background technology]

[0002] Many diseases require timely intervention to improve patient outcomes. Treatment delays and symptom variability can lead to undiagnosed conditions and poor outcomes, such as comorbidities. Early identification of a patient's diagnosis may improve the patient's health outcomes and disease management. Timely educational efforts may reduce the patient's signs and symptoms and potentially reduce other health risks.

[0003] Examples of such diseases include some rare diseases. It is estimated that approximately 300 million people worldwide suffer from a rare disease. Rare diseases typically take longer to diagnose than more common diseases, with the average diagnosis time for a rare disease exceeding four years. This can lead to significant delays in rare disease treatment and a lower likelihood of successful treatment. Many factors contribute to this delay, including physician unfamiliarity with rare diseases, the variable symptoms of a given rare disease, and the masking of rare diseases by symptoms of more common diseases. While traditional diagnostic algorithms can be effective in some cases, they rely on healthcare professional (HCP) awareness and require validation of a myriad of clinical features, including differential diagnoses. In the real world, these conditions are rarely met. Summary of the Invention [Problem to be solved by the invention]

[0004] A machine learning computer-implemented method for identifying healthcare providers seeking potential disease diagnoses and / or education regarding the diagnosis of a particular disease through disease-related geofencing data from one or more healthcare delivery locations, including one or more impressions and / or one or more clicks made at a healthcare delivery location, in combination with historical patterns from healthcare reimbursement statements or clinical test findings and historical sales / prescription sales order data. [Means for solving the problem]

[0005] According to a first aspect of the present specification, a computer-implemented method for identifying a potential disease diagnosis and / or healthcare provider is described. The method includes receiving geofencing data from one or more healthcare delivery locations, the geofencing data related to a disease and including one or more impressions made at the healthcare delivery locations and / or one or more clicks made at the healthcare delivery locations; extracting a plurality of features from the geofencing data for each of a plurality of entities, each entity associated with a respective healthcare delivery location in the one or more healthcare delivery locations; processing the extracted features for each entity using a machine learning model to determine an indication of whether a patient with the potential disease is present at a respective healthcare delivery location associated with the entity and / or whether a healthcare provider at a respective healthcare delivery location associated with the entity is potentially seeking information related to the disease; and triggering an alert in response to determining an indication of whether a patient with the potential disease is present at a healthcare delivery location in the one or more healthcare delivery locations and / or an indication that a healthcare provider at a respective healthcare delivery location associated with the entity is potentially seeking information related to the disease, wherein the alert includes identification information of the healthcare delivery locations and / or the entities associated with the healthcare delivery locations.

[0006] The above and other aspects may further include one or more of the following features, either alone or in combination.

[0007] The plurality of feature amounts may include the number of impressions in the first time period, the number of impressions in the second time period, the number of clicks in the first time period, and the number of clicks in the second time period.

[0008] The method may further include receiving one or more sets of disease-related laboratory test data from one or more of the healthcare delivery locations, each set of laboratory test data indicating laboratory tests performed by an entity at the healthcare delivery location. A plurality of features may be further extracted from the one or more sets of laboratory test data. The plurality of features may include a number of laboratory test orders and / or laboratory tests in a first time period and a number of laboratory test orders and / or laboratory tests in a second time period. The first time period may be 3 to 5 days, and the second time period may be 5 to 14 days (e.g., 10 days).

[0009] The method may be repeated periodically, and in each repetition, extracting the plurality of features includes extracting the plurality of features from geofencing data captured within a predetermined time prior to the time of the repetition.

[0010] The disease may be a rare disease or an ultra-rare disease.

[0011] Processing the extracted features for each entity to determine an indication of whether an individual with a potential disease is present at each healthcare location associated with the entity may include: using a machine learning model to determine, for each entity, a probability that a patient with the disease is present at each healthcare location associated with the entity and / or a probability that a healthcare provider at each healthcare location associated with the entity is potentially seeking information related to the disease; comparing the determined probability with a threshold probability level; and, in response to determining that the determined probability is greater than the threshold probability level, indicating that a patient with the potential disease is present at each healthcare location associated with the entity and / or that a healthcare provider at each healthcare location associated with the entity is potentially seeking information related to the disease.

[0012] Triggering the alert may include sending educational material related to the disease to a medical professional at the identified healthcare delivery location. Triggering the alert may include sending an indication to a sales agent that a potential disease diagnosis or disease-related information is associated with the healthcare facility / entity.

[0013] The method may further include receiving search data associated with one or more search terms related to the disease, and extracting one or more features from the search data for each of the plurality of entities.

[0014] The machine learning model may include a logistic regression model, a random forest model, a neural network, a generalized additive model, and / or an XGBoost model.

[0015] The plurality of entities may include one or more of: one or more customer accounts associated with each healthcare location; one or more departments and / or sub-departments associated with each healthcare location; one or more healthcare professionals associated with each healthcare location; and / or one or more healthcare providers associated with each healthcare location.

[0016] According to a further aspect of the present specification, a computer program product is described that includes computer-readable instructions that, when executed by a computer, cause the computer to perform any one or more of the methods disclosed herein.

[0017] According to a further aspect of the present specification, a system is described that includes one or more processors and a memory, wherein the memory stores computer-readable instructions that, when executed by the one or more processors, cause the computer to perform any one or more of the methods disclosed herein.

[0018] As used herein, the term "rare disease" is preferably used to include diseases that affect fewer than 1 in 2000 people in the general population. Currently, there are more than 6,000 known rare diseases, and new rare diseases are constantly being discovered. The term "rare disease" may also encompass ultra-rare diseases. The term "ultra-rare disease" is preferably used to include diseases that affect fewer than 1 in 50,000 people in the general population. However, the methods, systems, and devices described herein are not limited to use with rare and ultra-rare diseases, but may be applied to any type / rarity of disease.

[0019] Exemplary embodiments will now be described with reference to the accompanying drawings. [Brief explanation of the drawings]

[0020] [Figure 1] 1 shows a schematic diagram of a method for identifying locations where potential disease patients may be present. [Figure 2] 2 shows a schematic diagram of a method 200 for training a machine learning model to identify potential patient locations of disease. [Figure 3] 1 shows a flow diagram of an exemplary method for identifying a potential disease diagnosis. [Figure 4]4 shows a schematic diagram of a system / apparatus 400 for performing any of the methods described herein. DETAILED DESCRIPTION OF THE INVENTION

[0021] The methods and systems described herein use machine learning to identify healthcare locations where individuals with a disease (also referred to herein as patients with the disease) may be located from geofencing data associated with the healthcare locations. This can then issue an alert indicating the likelihood of a patient with the disease and additional actions to be taken accordingly (e.g., providing information about the disease). This results in faster diagnosis and treatment of patients with the disease, increasing the likelihood of successful disease management. In some embodiments, the disease may be a rare or ultra-rare disease, although the methods and systems described herein may be applied to diseases that are not rare as well.

[0022] FIG. 1 shows a schematic diagram of a method 100 for identifying locations where potential disease patients may be present and / or where healthcare providers / healthcare professionals may be seeking information regarding the disease.

[0023] The method includes receiving geofencing data 102 associated with a plurality of healthcare delivery locations and extracting features 104 from the geofencing data for each of a plurality of entities associated with the healthcare delivery location / sub-location. The extracted features are input into one or more machine learning models 106, which process the extracted features to generate predictions 108 related to whether an individual with a disease (e.g., a rare disease) is present at the location / sub-location associated with each entity. If such a patient (or multiple such patients) is present at one or more locations / sub-locations, an alert 110 is issued identifying one or more entities associated with the location. The method 100 may be performed / repeated periodically (e.g., daily).

[0024] A geofence comprises a virtual boundary that defines an area in the real world. A geofence may be based on the footprint of a defined boundary / real-world work location (e.g., a hospital), a distance from a predetermined location (e.g., within 10 meters of a hospital diagnostic department), or a predetermined travel time from a predetermined location (e.g., within 10 minutes of travel time from a hospital). In the methods described herein, geofences are constructed around healthcare delivery locations / sub-locations. Examples of such locations / sub-locations include hospitals, hospital departments, pharmacies, primary care physician locations, etc.

[0025] When a user of a mobile computing device enters a geofenced area, the user's presence in the geofenced area is detected. For example, the user's GPS location may be detected to be within the geofenced area, and / or the location of a wireless access point used by the user may be within the geofenced area. Based on the user's presence within the geofence, advertisements and / or information related to a given disease may be provided to the user.

[0026] Geofencing data 102 includes information related to a user's interactions with advertisements and / or information served by a search engine to a user in a geofenced location.

[0027] The geofencing data may include "impression" data for each geofence location. Each instance of providing an advertisement / information set may be referred to as an "impression." The geofencing data may include the number of times, i.e., the number of impressions, that the advertisement / information set was provided to a user at the geofence location over one or more time periods. For example, the geofencing data 102 may include the number of impressions related to a particular disease for each day in a predetermined number of preceding days.

[0028] The geofencing data 102 may include "click" data for each geofence location. A click is generated each time a user interacts with a provided set of advertisements / information. For example, a "click" is generated each time a user clicks / selects / interacts with a provided set of advertisements or information. The geofencing data may include the number of clicks associated with a set of advertisements / information provided to a user at a geofence location over one or more time periods. For example, the geofencing data 102 may include the number of clicks associated with a particular disease for each day in a predetermined number of preceding days.

[0029] The geofencing data 102 may further include search term data for each geofence location, which may include a count of the number of times one or more disease-related terms are searched within the geofence location.

[0030] The search terms may include, for example, the name(s) of the disease, symptoms of the disease, treatments for the disease, etc. The search terms may be grouped into one or more sets of similar and / or closely related words (e.g., synonyms).

[0031] The method 100 may use geofencing data 102 collected from a predetermined rolling period before the method 100 is performed. For example, the geofencing data 102 may be geofencing data 102 collected over the past 14 days.

[0032] A plurality of features 104 are extracted from the geofencing data 102 for each of a plurality of entities, each corresponding to a geofence location. The plurality of features 104 may include, for example, a number of impressions associated with each entity for each of one or more time periods (e.g., the last 3 days, 5 days, 7 days, 10 days, and / or 14 days). The plurality of features 104 may include, for example, a number of clicks associated with each entity for each of one or more time periods (e.g., the last 3 days, 5 days, 7 days, 10 days, and / or 14 days). The plurality of features 104 may include, for example, a number of times each search term (or each set of similar search terms) was entered at a geofence location associated with each entity for each of one or more time periods (e.g., the last 3 days, 5 days, 7 days, 10 days, and / or 14 days). Additional features may be extracted from further input data, as described below.

[0033] Each geofenced healthcare delivery location may have one or more associated entities. An entity may be, for example, the healthcare delivery location itself (e.g., the identity of the hospital associated with the geofenced location) or a department / sub-department within the hospital. Alternatively or additionally, the one or more entities associated with a location may include one or more accounts (e.g., commercial accounts) associated with the healthcare delivery location. Alternatively or additionally, the one or more entities associated with a healthcare delivery location may include the identity of a healthcare professional or provider associated with the location, e.g., the name of a physician working at the location and / or the healthcare provider's National Provider Identifier (NPI) if opted in to receive advertisements.

[0034] The one or more machine learning models 106 include one or more parameterized models trained to predict, based on the extracted features 104, data 108 indicative of the presence of potential diseased patients at healthcare locations during the time period in which the geofencing data 102 was collected. An example of the training process is described below in conjunction with FIG. 2. The one or more machine learning models 106 take the extracted features 104 as input and process them based on the learned parameters of the models to generate output 108 indicative of the presence of potential diseased patients. The output 108 may include, for example, a probability that a diseased patient is present at each healthcare location / entity and / or a probability that a healthcare provider desires information about the disease (e.g., symptoms, causes, treatment options, etc.). The probability may be a score between 0 and 1 or a percentage score. Alternatively, the output may be a binary indicator of the presence of potential diseased patients or healthcare providers seeking information (e.g., 1 indicating the presence of potential diseased patients and 0 indicating the absence of potential diseased patients).

[0035] Processing the extracted data 104 with one or more machine learning models 106 may include inputting the extracted features for each entity individually in sequence into the machine learning models 106. The machine learning models 106 process the input data for each entity individually to generate predictions for that entity, and then receive input data for the next entity.

[0036] Alternatively, for N entities, N copies of the machine learning model 106 may be used to process the extracted features 104 for the N entities in parallel, with each copy of the machine learning model 106 processing the extracted features 104 from a respective entity and outputting a respective prediction 108 for that entity.

[0037] The one or more machine learning models 106 may include, for example, a random forest model, a logistic regression model, an XGBoost model, a neural network such as a fully connected neural network, and / or a generalized additive model (GAM).

[0038] If the output 108 of one or more machine learning models 106 indicates that a patient with a potential disease is present at a location associated with an entity and / or that healthcare providers at respective healthcare delivery locations associated with the entity are potentially seeking information about the disease, an alert 110 is triggered. The output 108 of the one or more machine learning models 106 may be compared to a threshold condition to determine whether to trigger the alert 110. For example, the probability that a patient with a disease is present at a location may be compared to a threshold probability. If the threshold probability is exceeded, an alarm is triggered for the location or one or more entities associated with the location. The threshold probability may be equal to, for example, a 70%, 80%, or 90% probability.

[0039] The alert 110 includes an identifier for the entity and / or location where the potential disease patient was detected and / or where a healthcare provider at a respective healthcare delivery location associated with that entity is potentially seeking information about the disease. The alert 110 may further include additional data related to the detection, such as the time / date of detection and / or contact details of the entity where the potential disease patient was detected. The alert 110 may be sent to the entity associated with the location where the potential disease patient / interested healthcare provider was detected and / or to one or more third parties, such as research institutions studying the disease and / or manufacturers / distributors of treatments for the disease. In some embodiments, the alert may be provided to sales or marketing representatives of pharmaceutical companies and serve as a form of lead generation.

[0040] The alert 110 may trigger one or more further actions. For example, in some embodiments, the alert 110 may trigger transmission of disease-related information to the identified entity and / or medical professionals associated with that entity. In this way, medical professionals in locations where patients with potential disease are present may be educated or alerted regarding the characteristics of the disease, thereby increasing the likelihood of a correct diagnosis.

[0041] In some implementations, when an alert 110 is triggered for an entity, the method 100 may include determining whether a previous alert for the entity was triggered within a predetermined period of time prior to the current alert. If an alert was triggered within that period, the current alert 110 for that entity may be discarded and / or ignored. The predetermined period may be, for example, seven days.

[0042] In some embodiments, alerts can be stored and analyzed to generate insights and identify trends in disease-related information. For example, a cluster of potential patients with a given disease can be identified. Information and / or marketing activities related to that disease can then be targeted to locations associated with the cluster. The insights / trends can trigger further alerts / actions. For example, if a disease has an environmental underlying cause, a cluster of potential diseased patients can trigger an alert that the environmental condition may be present in a geographic area.

[0043] The extracted features 104 may additionally be extracted from one or more further sets of input data associated with each geofence location. One example of such data may be the number and type of lab tests 112 performed by an entity associated with each geofence location. Such data may be anonymized and may include the number of times each of a number of different lab tests related to a disease originating from (e.g., ordered by) the entity associated with the geofence location was performed. As another example, the one or more further sets of input data associated with each geofence location may alternatively or additionally include lab test orders 114 placed by the entity associated with the geofence location.

[0044] 2 shows a schematic diagram of a method 200 for training a machine learning model to identify locations where patients with potential diseases may be present and / or where healthcare providers are potentially interested in information about a given disease. The method may be performed by one or more computer systems, such as the system described below in connection with FIG. 4.

[0045] The training method 200 uses training examples 218 taken from a set of training data. Each example includes a set of input data including historical geofencing data 202 and a corresponding ground truth classification 210 for the input data. The geofencing data may be the geofencing data described above in connection with FIG. 1. The input data may further include additional data types, such as the number and types of lab tests 112 and / or lab test orders 114 performed by entities associated with each geofenced location.

[0046] The ground truth classification 210 includes data indicating whether a patient with the disease was present at the healthcare location associated with the geofencing data 202 during the time period covered by the geofencing data 202. Such data may include an indication of a positive diagnosis of the disease at the healthcare location within a predetermined period (e.g., within seven days) of the time period covered by the geofencing data.

[0047] Alternatively or additionally, the ground truth classification 210 may include indirect indicators of the presence of patients with the disease, such as orders from / sales to entities associated with the healthcare location for medications / devices to treat the disease within a predetermined time period (e.g., within 7 days) of the time period covered by the geofencing data. Sales / orders for treatments within a predetermined geodesic distance (e.g., 1 km) of a geofence location may be counted as originating from the healthcare location associated with that geofence location. This can increase the amount of available training data, thereby improving the performance of the trained machine learning model 206.

[0048] During training, multiple features 204 for each of multiple entities are extracted from the input data. The extracted features 204 for each entity are processed by one or more machine learning models 206 to generate candidate predictions 208 regarding the presence of patients with the disease at healthcare delivery locations associated with each entity. The candidate predictions 208 are compared to corresponding ground truth classifications 210 using a loss / objective function 216. The loss function 216 may include a measure of area under the curve. Alternatively, a classification loss such as binary cross-entropy may be used. The value of the loss / objective function 216 is used to generate updates to the machine learning models 206 using an optimization routine (e.g., stochastic gradient descent).

[0049] Training may be repeated on the training data until a threshold condition is met, which may be a threshold number of training iterations / epochs and / or a threshold performance reached on a test dataset, which may have the same structure as the training dataset but may also contain different examples.

[0050] 3 shows a flow diagram of an exemplary method for identifying a potential disease diagnosis and / or identifying healthcare providers who may be interested in a given disease, which may be performed by one or more computing systems / devices, such as the system / device described in connection with FIG.

[0051] In operation 3.1, geofencing data is received from one or more care delivery locations, the geofencing data including one or more impressions related to a condition performed at the care delivery location and / or one or more clicks related to a condition performed at the care delivery location.

[0052] In some embodiments, one or more additional inputs are also received. The one or more additional inputs may include search data related to one or more search terms related to the disease, such as the number of times each of the one or more search terms was searched at the healthcare delivery locations. The search terms may include, for example, symptoms of the disease, treatments for the disease, and / or names / types of the disease. The one or more additional inputs may alternatively or additionally include one or more sets of laboratory test data related to the disease from one or more of the healthcare delivery locations. Each set of laboratory test data may indicate which of multiple laboratory tests were performed by an entity at the healthcare delivery location, for example, how many times a particular type of blood test was performed at the healthcare delivery location and / or by the entity.

[0053] In operation 3.2, a plurality of features are extracted from the geofencing data for each of a plurality of entities, each entity being associated with a respective care delivery location in one or more care delivery locations. Features may also be extracted from any additional input to the method.

[0054] The plurality of feature amounts include the number of impressions in a first period, the number of impressions in a second period, the number of clicks in the first period, and the number of clicks in the second period. The first period may be 3 to 5 days (e.g., 3 days), and the second period may be 5 to 14 days (e.g., 5 days).

[0055] The plurality of feature amounts may further include the number of impressions in a third time period and / or the number of clicks in a third time period. The third time period may be 7 to 14 days (e.g., 10 days). The feature amounts may further include the number of impressions in a fourth time period and / or the number of clicks in a fourth time period. The fourth time period may be 12 to 21 days (e.g., 14 days).

[0056] In embodiments where lab test order data and / or lab tests are also received, the extracted features may further include the number of lab test orders and / or lab tests in the first time period, the second time period, the third time period, and / or the fourth time period.

[0057] The plurality of entities may include one or more of: one or more customer accounts associated with each healthcare location; one or more departments and / or sub-departments associated with each healthcare location; one or more healthcare professionals associated with each healthcare location; and / or one or more healthcare providers associated with each healthcare location.

[0058] In operation 3.3, the extracted features for each entity are processed using one or more machine learning models that output data indicating whether there are individuals with potential diseases at each healthcare location associated with the entity and / or whether healthcare providers at each healthcare location associated with the entity are potentially seeking information related to the disease.

[0059] The output of the machine learning model can be the probability of the presence of an individual with the disease for each entity / location.

[0060] The machine learning model may include one or more of a logistic regression model, a random forest model, a GAM, a neural network (e.g., a fully connected neural network), and / or an XGBoost model. The machine learning model may be trained on historical geofencing data, as described above in connection with FIG. 2.

[0061] In operation 3.4, the potential presence of patients with the disease and / or healthcare providers at each healthcare delivery location associated with the entity potentially seeking information about the disease (e.g., symptoms, diagnostic information / tests, and / or treatment options) is evaluated based on the data output by the machine learning model. If the data indicates the potential presence of individuals with the disease and / or healthcare providers at each healthcare delivery location associated with the entity potentially seeking information related to the disease, the method proceeds to operation 3.5. Otherwise, the method returns to operation 3.1 to wait for the next batch of geofencing data.

[0062] The data output by the machine learning model can be compared to one or more thresholds to determine whether there are potential patients with the disease and / or whether healthcare providers at the entity's respective healthcare locations are potentially seeking information related to the disease. For example, the probability of each entity having an individual with the disease can be compared to a threshold probability (e.g., 70%, 80%, or 90%, among others). If the probability of an individual with the disease being present for an entity exceeds the threshold probability, it is determined that there may be potential patients with the disease at the location associated with that entity.

[0063] At operation 3.5, an alert is triggered. The alert includes the identity of the care location and / or an entity associated with the care location. The method then returns to operation 3.1 to wait for the next batch of geofencing data.

[0064] The alert may be sent to a medical professional at the identified healthcare location (i.e., a location associated with the identified entity). The alert may include educational materials related to the disease, such as symptoms of the disease and / or treatments / therapy for the disease. In some embodiments, the alert may include a link to a site where medication for the disease can be ordered.

[0065] Alternatively or additionally, the alert may be sent to a sales or marketing representative of a pharmaceutical company that produces a treatment for the disease.

[0066] Batches of geofencing data (and in some implementations, other input data) may be received periodically (e.g., once daily). The method may be repeated each time a new batch is received. In each repetition, extracting the plurality of features may include extracting the plurality of features from geofencing data captured within a predetermined time period before the new data is received. This predetermined time period may be greater than 10 days (e.g., 14 days).

[0067] 4 shows a schematic diagram of a system / apparatus 400 for performing any of the methods described herein. The illustrated system / apparatus is an example of a computing device. The system / apparatus 400 may form at least a part of a concrete mixer, for example, part of the concrete mixer's ECU.

[0068] The device (or system) 400 includes one or more processors 402. The one or more processors control the operation of the other components of the system / device 400. The one or more processors 402 may include, for example, a general-purpose processor. The one or more processors 402 may be single-core or multi-core devices. The one or more processors 402 may include a central processing unit (CPU) or a graphics processing unit (GPU). Alternatively, the one or more processors 402 may include specialized processing hardware (e.g., a RISC processor or programmable hardware incorporating firmware). Multiple processors may be included.

[0069] The system / device includes a working or volatile memory 404. One or more processors may access the volatile memory 404 to process data and may manage the storage of data in the memory. The volatile memory 404 may include any type of RAM (e.g., static RAM (SRAM) or dynamic RAM (DRAM)), or may include flash memory such as an SD card.

[0070] The system / apparatus includes a non-volatile memory 406. The non-volatile memory 406 stores a set of operating instructions 408 in the form of computer-readable instructions for controlling the operation of the processor 402. The non-volatile memory 406 may be any type of memory, such as read-only memory (ROM), flash memory, or magnetic drive memory.

[0071] The one or more processors 402 are configured to execute operational instructions 408 to cause the system / device to perform any of the methods described herein. The operational instructions 408 may include code (i.e., drivers) associated with hardware components of the system / device 400 and code associated with basic operations of the system / device 400. Generally speaking, the one or more processors 402 use the volatile memory 404 to execute one or more of the operational instructions 408 that are stored persistently or semi-persistently in the non-volatile memory 406, and to temporarily store data generated during execution of the operational instructions 408.

[0072] Any of the mentioned devices and / or other features of a particular mentioned device may be provided by a device arranged such that it is configured to perform a desired operation only when enabled (e.g., switched on). In such a case, a non-enabled state (e.g., switched off) may not necessarily have appropriate software loaded into active memory, but only an enabled state (e.g., on) may have appropriate software loaded. The device may include hardware circuitry and / or firmware. The device may include software loaded into memory. Such software / computer programs may be recorded on the same memory / processor / functional unit and / or on one or more memory / processors / functional units.

[0073] A mentioned device / circuit / element / processor may have other functions in addition to those mentioned, and further, these functions may be performed by the same device / circuit / element / processor. One or more disclosed aspects may encompass an associated computer program and electronic distribution of the computer program (which may be source / transport encoded) recorded on a suitable carrier (e.g., memory, signal).

[0074] Any "computer" described herein may comprise a collection of one or more individual processors / processing elements, whether located on the same circuit board, or in the same region / location of a circuit board, or even on the same device. In some examples, one or more of any mentioned processors may be distributed across multiple devices. The same or different processors / processing elements may perform one or more functions described herein.

[0075] The term "signaling" may refer to one or more signals being transmitted as a series of transmitted and / or received electrical / optical signals. The series of signals may include one, two, three, four, or more individual signal components or separate signals to make up the signaling. Some or all of these individual signals may be transmitted / received simultaneously, sequentially, and / or overlapping in time with one another via wireless or wired communication.

[0076] With reference to any mentioned discussion of computers and / or processors and memory (including, for example, ROM, CD-ROM, etc.), these may include computer processors, application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), and / or other hardware components programmed to perform the functions of the present invention.

[0077] Herein, the applicant hereby discloses each individual feature described in this specification and any combination of two or more such features solely, to the extent that such feature or combination is operable based on the specification as a whole and in light of the general general knowledge of a person skilled in the art, regardless of whether such feature or combination solves the problems disclosed herein, and without limiting the scope of the claims. The applicant indicates that the disclosed aspects / examples may be composed of any such individual feature or combination of features. Upon reviewing the above description, it should be apparent to one skilled in the art that various modifications can be made within the scope of the present disclosure.

[0078] While essential novel features applicable to the examples have been illustrated, described, and pointed out, it will be understood that various omissions, substitutions, and changes in the form and details of the described devices and methods may be made by those skilled in the art without departing from the scope of the present disclosure. For example, all combinations of those elements and / or method steps that perform substantially the same function in substantially the same way to achieve the same results are expressly intended to be within the scope of the present disclosure. Furthermore, it will be recognized that structures and / or elements and / or method steps illustrated and / or described in connection with any disclosed embodiment or example may be incorporated into any other disclosed or described or proposed embodiment or example as a general matter of design choice. Furthermore, in the claims, means-plus-function clauses are intended to encompass structures described herein as performing the recited function and to encompass equivalent structures as well as structural equivalents.

[0079] Implementations of the methods described herein can be realized in digital electronic circuitry, integrated circuits, specially designed ASICs (application-specific integrated circuits), computer hardware, firmware, software, and / or combinations thereof. These can include a computer program product (e.g., software stored on a magnetic or optical disk, memory, programmable logic device, etc.) containing computer-readable instructions that, when executed by a computer such as that described in connection with FIG. 7, cause the computer to perform one or more of the methods described herein.

[0080] Any system features as described herein may also be provided as method features, and vice versa. As used herein, means-plus-function features may alternatively be expressed in terms of their corresponding structure. In particular, method aspects may apply to system aspects, and vice versa.

[0081] Furthermore, any, some, and / or all of the features in one embodiment may be applied to any, some, and / or all of the features in any other embodiment, in any appropriate combination. It is also understood that specific combinations of the various features described and defined in any embodiment of the present invention may be implemented and / or provided and / or used independently.

[0082] While several embodiments have been shown and described, it will be understood by those skilled in the art that changes can be made in these embodiments without departing from the principles of the present disclosure, the scope of which is defined in the claims and their equivalents.

[0083] The terms "drug" or "medicament" are used synonymously herein to refer to a formulation containing one or more active pharmaceutical ingredients or pharmaceutically acceptable salts or solvates thereof, and optionally a pharmaceutically acceptable carrier. An active pharmaceutical ingredient ("API"), in the broadest sense, is a chemical structure that has a biological effect on humans or animals. In pharmacology, drugs or agents are used to treat, cure, prevent, or diagnose disease or otherwise improve physical or mental well-being. Drugs or agents may be used for a limited period of time or periodically for chronic conditions.

[0084] As described below, drugs or pharmaceutical agents can include at least one API or a combination thereof in various types of formulations for the treatment of one or more diseases. Examples of APIs include small molecules having a molecular weight of 500 Da or less, polypeptides, peptides, and proteins (e.g., hormones, growth factors, antibodies, antibody fragments, and enzymes), carbohydrates and polysaccharides, as well as nucleic acids, double-stranded or single-stranded DNA (including naked and cDNA), RNA, antisense nucleic acids such as antisense DNA and RNA, small interfering RNA (siRNA), ribozymes, genes, and oligonucleotides. Nucleic acids can be incorporated into molecular delivery systems such as vectors, plasmids, or liposomes. Mixtures of one or more drugs are also contemplated.

[0085] The drug or agent may be contained within a primary package or "drug container" adapted for use with a drug delivery device. The drug container may be, for example, a cartridge, syringe, reservoir, or other sturdy or flexible vessel configured to provide a suitable chamber for storage (e.g., short-term or long-term storage) of one or more drugs. For example, in some instances, the chamber may be designed to store the drug for at least one day (e.g., from one day to at least 30 days). In some cases, the chamber may be designed to store the drug for about one month to about two years. Storage may occur at room temperature (e.g., about 20°C) or at refrigerated temperatures (e.g., from about -4°C to about 4°C). In some cases, the drug container may be or include a dual-chamber cartridge configured to separately store two or more components of a pharmaceutical formulation to be administered (e.g., an API and a diluent, or two different drugs), one in each chamber. In such cases, the two chambers of the dual-chamber cartridge may be configured to allow mixing between the two or more components prior to and / or during administration to a human or animal body. For example, the two chambers may be configured so that they are in fluid communication with each other (e.g., by a conduit between the two chambers), allowing the user to mix the two components if desired prior to administration. Alternatively, or additionally, the two chambers may be configured to allow mixing of the components as they are administered into the human or animal body.

[0086] Drugs or agents contained in drug delivery devices as described herein can be used to treat and / or prevent many different types of medical disorders. Examples of disorders include, for example, diabetes or complications related to diabetes, such as diabetic retinopathy, and thromboembolic disorders, such as deep vein thromboembolism or pulmonary thromboembolism. Further examples of disorders include acute coronary syndrome (ACS), angina pectoris, myocardial infarction, cancer, macular degeneration, inflammation, hay fever, atherosclerosis, and / or rheumatoid arthritis. Examples of APIs and drugs are those listed in handbooks such as the Rote Liste 2014, including, but not limited to, Main Group 12 (antidiabetic drugs) or 86 (oncology drugs), and the Merck Index, 15th Edition.

[0087] Examples of APIs for the treatment and / or prevention of type 1 or type 2 diabetes mellitus or complications associated with type 1 or type 2 diabetes mellitus include insulin, e.g., human insulin or a human insulin analog or derivative; glucagon-like peptide (GLP-1); GLP-1 analog or GLP-1 receptor agonist or analog or derivative thereof; dipeptidyl peptidase-4 (DPP4) inhibitor or a pharmaceutically acceptable salt or solvate thereof, or any mixture thereof. As used herein, the terms "analog" and "derivative" refer to a polypeptide having a molecular structure that is formally derivable from the structure of a naturally occurring peptide, e.g., the structure of human insulin, by deletion and / or replacement of at least one amino acid residue occurring in the naturally occurring peptide and / or by addition of at least one amino acid residue. The added and / or substituted amino acid residues can be either codable amino acid residues or other naturally occurring residues, or purely synthetic amino acid residues. Insulin analogs are also referred to as "insulin receptor ligands." In particular, the term "derivative" refers to a polypeptide having a molecular structure formally derivable from that of a naturally occurring peptide, such as that of human insulin, in which one or more organic substituents (e.g., fatty acids) are attached to one or more of the amino acids. Optionally, one or more amino acids found in the naturally occurring peptide may be deleted and / or substituted with other amino acids, including non-codable amino acids, or amino acids, including non-codable amino acids, may be added to the naturally occurring peptide.

[0088] Examples of insulin analogues are Gly(A21), Arg(B31), Arg(B32) human insulin (insulin glargine); Lys(B3), Glu(B29) human insulin (insulin glulisine); Lys(B28), Pro(B29) human insulin (insulin lispro); Asp(B28) human insulin (insulin aspart); human insulin in which the proline in position B28 can be replaced by Asp, Lys, Leu, Val or Ala and in position B29 Lys can be replaced by Pro; Ala(B26) human insulin; Des(B28-B30) human insulin; Des(B27) human insulin and Des(B30) human insulin.

[0089] Examples of insulin derivatives are, for example, B29-N-myristoyl-des(B30) human insulin, Lys(B29)(N-tetradecanoyl)-des(B30) human insulin (insulin detemir, Levemir®); B29-N-palmitoyl-des(B30) human insulin; B29-N-myristoyl human insulin; B29-N-palmitoyl human insulin; B28-N-myristoylLysB28ProB29 human insulin; B28-N-palmitoyl-LysB28ProB29 human insulin; B30-N-myristoyl-ThrB29LysB30 human insulin. B29-N-palmitoyl-ThrB29LysB30 human insulin; B29-N-(N-palmitoyl-gamma-glutamyl)-des(B30) human insulin, B29-N-omega-carboxypentadecanoyl-gamma-L-glutamyl-des(B30) human insulin (insulin degludec, Tresiba®); B29-N-(N-lithocholyl-gamma-glutamyl)-des(B30) human insulin; B29-N-(ω-carboxyheptadecanoyl)-des(B30) human insulin, and B29-N-(ω-carboxyheptadecanoyl) human insulin.

[0090] Examples of GLP-1, GLP-1 analogs and GLP-1 receptor agonists include, for example, lixisenatide (Lyxumia®), exenatide (exendin-4, Byetta®, Bydureon®, a 39 amino acid peptide produced by the salivary glands of the flathead sea urchin), liraglutide (Victoza®), semaglutide, taspoglutide, albiglutide (Syncria®), dulaglutide (Trulicity®), rexendin-4, CJC-1134-PC, PB-1023, TTP-054, langrenatide / HM-11260C (efpegrenatide), HM-15211, C M-3, GLP-1 Erigen, ORMD-0901, NN-9423, NN-9709, NN-9924, NN-9926, NN-9927, Nodexene, Viador-GLP-1, CVX-096, ZYOG-1, ZYD-1, GSK-2374697, DA-3091, MAR-701, MAR709, ZP-2929, Z P-3022, ZP-DI-70, TT-401 (pegapamodtide), BHM-034, MOD-6030, CAM-2036, DA-15864, ARI-2651, ARI-2255, tirzepatide (LY3298176), bamadutide (SAR425899), exenatide-XTEN, and glucagon-Xten.

[0091] An example of an oligonucleotide is mipomersen sodium (Kynamro®), a cholesterol-lowering antisense therapeutic agent for the treatment of, for example, familial hypercholesterolemia, or RG012 for the treatment of Alport syndrome.

[0092] Examples of DPP4 inhibitors are linagliptin, vildagliptin, sitagliptin, denagliptin, saxagliptin, berberine.

[0093] Examples of hormones include pituitary or hypothalamic hormones or regulatory active peptides such as gonadotropins (follitropin, lutropin, chorion gonadotropin, menotropin), somatropin (somatropin), desmopressin, terlipressin, gonadorelin, triptorelin, leuprorelin, buserelin, nafarelin, and goserelin, and their antagonists.

[0094] Examples of polysaccharides include glycosaminoglycans, hyaluronic acid, heparin, low molecular weight heparin, or ultra-low molecular weight heparin, or derivatives thereof, or polysulfated forms thereof, such as sulfated forms of the above polysaccharides, and / or pharmaceutically acceptable salts thereof. An example of a pharmaceutically acceptable salt of polysulfated low molecular weight heparin is enoxaparin sodium. An example of a hyaluronic acid derivative is Hylan GF 20 (Synvisc®), sodium hyaluronate.

[0095] As used herein, the term "antibody" refers to an immunoglobulin molecule or an antigen-binding portion thereof. Examples of antigen-binding portions of immunoglobulin molecules include F(ab) and F(ab')2 fragments that retain antigen-binding ability. An antibody can be a polyclonal antibody, a monoclonal antibody, a recombinant antibody, a chimeric antibody, a deimmunized or humanized antibody, a fully human antibody, a non-human (e.g., murine) antibody, or a single-chain antibody. In some embodiments, an antibody has effector function and is capable of fixing complement. In some embodiments, an antibody has reduced or no binding ability to Fc receptors. For example, an antibody can be an isotype or subtype, antibody fragment, or mutant that does not support Fc receptor binding, e.g., with a mutation or deletion of the Fc receptor binding region. The term antibody also includes antigen-binding molecules based on tetravalent bispecific tandem immunoglobulins (TBTIs) and / or dual variable region antibody-like binding proteins with a crossover binding region orientation (CODV).

[0096] The term "fragment" or "antibody fragment" refers to a polypeptide derived from an antibody polypeptide molecule (e.g., an antibody heavy and / or light chain polypeptide) that does not include the full-length antibody polypeptide but comprises at least a portion of the full-length antibody polypeptide that is still capable of binding to antigen. Antibody fragments may include truncations of the full-length antibody polypeptide, although the term is not limited to such truncated fragments. Antibody fragments useful in the present invention include, for example, Fab fragments, F(ab')2 fragments, scFv (single-chain Fv) fragments, linear antibodies, monospecific or multispecific antibody fragments, such as bispecific, trispecific, tetraspecific, and multispecific antibodies (e.g., diabodies, triabodies, tetrabodies), monovalent or multivalent antibody fragments, such as bivalent, trivalent, tetravalent, and multivalent antibodies, minibodies, chelating recombinant antibodies, tribodies or bibodies, intrabodies, nanobodies, small modular immunopharmaceuticals (SMIPs), binding domain immunoglobulin fusion proteins, camelized antibodies, and VHH-containing antibodies. Additional examples of antigen-binding antibody fragments are known in the art.

[0097] The term "complementarity determining region" or "CDR" refers to short polypeptide sequences within the variable regions of both heavy and light chain polypeptides that are primarily responsible for mediating specific antigen recognition. The term "framework region" refers to amino acid sequences within the variable regions of both heavy and light chain polypeptides that are not CDR sequences and that are primarily responsible for maintaining the proper orientation of the CDR sequences to enable antigen binding. Although the framework regions themselves typically do not directly participate in antigen binding, as is known in the art, certain residues within the framework regions of a particular antibody may be directly involved in antigen binding or may affect the ability of one or more amino acids within the CDRs to interact with the antigen.

[0098] Examples of antibodies are anti-PCSK-9 mAb (e.g., alirocumab), anti-IL-6 mAb (e.g., sarilumab), and anti-IL-4 mAb (e.g., dupilumab).

[0099] Pharmaceutically acceptable salts of any of the APIs described herein are also contemplated for use in the drug or medicament within the drug delivery device. Pharmaceutically acceptable salts include, for example, acid addition salts and base salts.

[0100] It will be understood by those skilled in the art that modifications (addition and / or deletion) of various components of the APIs, formulations, devices, methods, systems, and embodiments described herein may be made without departing from the full scope and spirit of the invention, and that the invention encompasses such modifications and any and all equivalents thereof.

[0101] An exemplary drug delivery device may include a needle-based injection system as described in Table 1 of Section 5.2 of ISO 11608-1:2014(E). As described in ISO 11608-1:2014(E), needle-based injection systems may be broadly divided into multi-dose container systems and single-dose (with partial or complete evacuation) container systems. The container may be an exchangeable container or an integrated, non-exchangeable container.

[0102] As further described in ISO 11608-1:2014(E), a multi-dose container system may include a needle-based injection device with replaceable containers. In such a system, each container holds multiple doses and the size may be fixed or variable (pre-set by the user). Another multi-dose container system may include a needle-based injection device integrated with a non-replaceable container. In such a system, each container holds multiple doses and the size may be fixed or variable (pre-set by the user).

[0103] As further described in ISO 11608-1:2014(E), a single-dose container system may include a needle-based injection device with replaceable containers. In one example of such a system, each container holds a single dose, thereby dispensing the entire deliverable amount (full discharge). In a further example, each container holds a single dose, thereby dispensing a portion of the deliverable amount (partial discharge). Also as described in ISO 11608-1:2014(E), a single-dose container system may include a needle-based injection device with integrated non-replaceable containers. In one example of such a system, each container holds a single dose, thereby dispensing the entire deliverable amount (full discharge). In a further example, each container holds a single dose, thereby dispensing a portion of the deliverable amount (partial discharge).

Claims

1. 1. A computer-implemented method for identifying a potential disease diagnosis and / or healthcare provider, the method comprising: receiving geofencing data from one or more healthcare delivery locations, the geofencing data being related to the disease and including one or more impressions made at the healthcare delivery locations and / or one or more clicks made at the healthcare delivery locations; extracting a plurality of features from the geofencing data for each of a plurality of entities, each entity associated with a respective care delivery location in the one or more care delivery locations; processing the extracted features for each entity using a machine learning model to determine an indication of whether a patient with a potential disease is present at each healthcare location associated with the entity and / or whether a healthcare provider at each healthcare location associated with the entity is potentially seeking information related to the disease; triggering an alert in response to determining an indication that a patient with a potential disease is present at a care delivery location in the one or more care delivery locations and / or an indication that a health care provider at each of the care delivery locations associated with the entity is potentially seeking information related to the disease, the alert including identification information of the care delivery location and / or the entity associated with the care delivery location; A method comprising:

2. The method of claim 1 , wherein the plurality of features comprises a number of impressions in a first time period, a number of impressions in a second time period, a number of clicks in the first time period, and a number of clicks in the second time period.

3. the method further comprising receiving one or more sets of laboratory test data related to the disease from one or more of the healthcare delivery locations, each set of laboratory test data indicating a laboratory test performed by an entity at the healthcare delivery location; The method of claim 1 or 2, wherein the plurality of features are further extracted from the one or more sets of clinical test data.

4. The method of claim 3 , wherein the plurality of features include a number of laboratory test orders and / or laboratory tests in the first time period and a number of laboratory test orders and / or laboratory tests in the second time period.

5. 5. The method of claim 2 or 4, wherein the first period of time is between 3 days and 5 days, and the second period of time is between 5 days and 14 days.

6. 6. The method of claim 1, wherein the method is repeated periodically, and in each repetition, extracting the plurality of features comprises extracting a plurality of features from geofencing data captured within a predetermined time prior to the time of the repetition.

7. The method of any one of claims 1 to 6, wherein the disease is a rare disease or an ultra-rare disease.

8. processing the extracted features for each entity to determine the indication of whether an individual with a potential disease is present at each healthcare location associated with the entity; using the machine learning model to determine the probability that a patient with a disease is present at each of the healthcare locations associated with the entity and / or the probability that a healthcare provider at each of the healthcare locations associated with the entity is potentially seeking information related to the disease; comparing the determined probability to a threshold probability level; In response to the determined probability being greater than the threshold probability level, indicating that a patient with a potential disease is present at the respective healthcare delivery location associated with the entity and / or that a healthcare provider at the respective healthcare delivery location associated with the entity is potentially seeking information related to the disease; The method according to any one of claims 1 to 7, comprising:

9. The method of any one of claims 1 to 8, wherein triggering the alert comprises sending educational material related to the disease to a medical professional at the identified medical delivery location.

10. 10. The method of claim 1, wherein triggering the alert comprises sending to a sales agent an indication that a diagnosis of a potential disease or information about the disease is associated with a medical facility / entity.

11. receiving search data related to one or more search terms related to the disease; extracting one or more features from the search data for each of the plurality of entities; The method of any one of claims 1 to 10, further comprising:

12. 12. The method of claim 1, wherein the machine learning model comprises a logistic regression model, a random forest model, a neural network, a generalized additive model, and / or an XGBoost model.

13. 13. The method of any one of claims 1-12, wherein the plurality of entities includes one or more of: one or more customer accounts associated with each healthcare delivery location; one or more departments and / or sub-departments associated with each healthcare delivery location; one or more healthcare professionals associated with each healthcare delivery location; and / or one or more healthcare providers associated with each healthcare delivery location.

14. A computer program product comprising computer readable instructions which, when executed by a computer, cause the computer to perform the method of any one of claims 1 to 13.

15. A system comprising one or more processors and a memory, said memory storing computer readable instructions that, when executed by said one or more processors, cause said computer to perform the method of any one of claims 1 to 13.

Images