EP4669203A1 - Methods and procedures for the treatment of migraine and post-traumatic headaches - Google Patents

Abstract

This document provides methods and processes involved in receiving one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time, extracting one or more features from the one or more inputs, analyzing the one or more features by determining at least the slope, intercept, and frequency of the one or more features, recommending a treatment plan for the subject based on the analysis of the one or one or more features, and providing one or more treatments to the subject based on the recommendation are provided.

Description

Methods and Processes for Treating Migraine and Post-Traumatic Headache

CROSS-REFERENCE TO RELATED APPLICATIONS

This application claims the benefit of U.S. Patent Application Serial No. 63/447,787, filed on February 23, 2023. The disclosure of the prior application is considered part of, and is incorporated by reference in, the disclosure of this application.

FEDERAL FUNDING

This invention was made with government support under NS 113315 awarded by the National Institutes of Health and W81XWH-19-1-0534 awarded by the Department of Defense. The government has certain rights in the invention.

TECHNICAL FIELD

This document relates to methods and processes for determining outcomes for subjects having a disease or disorder associated with head pain (e.g., migraine or post-traumatic headache). For example, this document provides methods and processes for determining the susceptibility of a subject to develop a disease or disorder associated with head pain, determining a therapeutic outcome of a subject to develop a disease or disorder associated with head pain (e.g., persistent head pain), and methods and process for recommending treatment in a subject having a disease or disorder associated with head pain.

BACKGROUND INFORMATION

Migraine is a neurologic disorder that involves moderate to severe headache, hypersensitivity to light and sound, nausea, and vomiting, and worsening of symptoms with routine physical activity. Migraine is a chronic condition that can have a significant impact on a person's quality of life. Migraine is thought to be related to aberrant function of central and peripheral nervous system structures such as certain brain and brainstem regions and the trigeminocervical complex.

Post -traumatic headache (PTH) is a type of headache that can occur after a traumatic brain injury (TBI). PTH can be caused by mild, moderate, or severe TBIs. PTH is often accompanied by other post- TBI symptoms such as neck pain, sensitivity to light and sound, and mild cognitive dysfunction. PTH can last for days to many years. PTH treatment includes medication and non-medication therapies (e.g., biobehavioral therapy, physical therapy).

SUMMARY

This document provides methods and processes involved in determining outcomes (e.g., future outcomes) for subjects having a disease or disorder associated with head pain (e.g., migraine or post-traumatic headache (PTH)). Migraine is a chronic and debilitating headache disorder that affects over 12% of the United States population and approximately 1 billion people worldwide. Migraine is the second leading cause of disability adjusted life years, and it accounts for over 50% of the disability attributed to neurological disorders. It is estimated to cost the United States about $28 billion annually due to direct and indirect expenses. Migraine is subclassified into episodic migraine, indicating that there are fewer than 15 headache days per month, and chronic migraine, indicating that there are at least 15 headache days per month including at least eight days per month on which the headaches and associated symptoms meet International Classification of Headache Disorders, 3^rd edition (ICHD-3) criteria for migraine with or without aura. Migraine treatment consists of as-needed therapy (treatments taken during a migraine attack to relieve symptoms) and mitigation therapy (treatments that reduce migraine attack frequency). Due to the severity of symptoms and functional impairment associated with migraine attacks, symptomatic therapy is indicated for most individuals with migraine. Mitigation therapy is indicated in up to 40% of those with episodic migraine and most individuals with chronic migraine, meaning that over 21 million people in the United States alone should be offered mitigation treatment each year.

Migraine mitigation treatment includes medications, nutraceuticals, and neuromodulation devices that are intended to reduce the frequency of migraine attacks and the associated disability. However, each migraine mitigation therapy has only a 30%-50% likelihood of providing such benefit for an individual patient. Thus, most of the time, subjects have to work through a lengthy process of trial-and-error until they can find a mitigation therapy that works for them. It is typical for a patient with migraine to go through a long period of trial-and-error, switching from one migraine mitigation medication to the next, until an effective treatment is found. This process can last from many months to years, a duration during which the patient continues to suffer with migraine. Recent data have demonstrated that nearly 80% of people with migraine who have taken at least one migraine mitigation therapy have needed to change their migraine mitigation therapy at least once, with nearly 90% of those individuals needing to change their migraine mitigation therapy two or more times. Earlier identification of whether a treatment will be effective would shorten the time for a patient to find effective migraine mitigation treatment.

Each year in the United States there are approximately 2.8 million traumatic brain injury (TBI)-related emergency department visits, hospitalizations, and deaths, and over 2 million individuals are diagnosed with mild TBI (mTBI). PTH is the most common symptom following a mTBI. Although some PTHs resolve within the first few days after onset, for many with PTH the headaches become “persistent” and last for months or years. This uncertainty about whether PTH will resolve or persist typically leads to a "wait-and-see" approach, in which PTH treatments are first initiated after the PTHs have already become persistent. Unfortunately, this approach can be unsuccessful. Individuals with persistent PTH have high levels of disability and high medical resource utilization. Earlier intervention would be more effective and individuals who are at high risk for persistence of PTH would be identified and treated relatively early after PTH onset, with the intent of reducing headache pain and interrupting the persistence of PTH.

However, intervention is not indicated for subjects who will have natural resolution of their PTH during the first few days or weeks. In clinical practice, there are not sufficient methods for determining which subjects with acute PTH will have resolution of their headaches during the acute phase vs. those who will have PTH persistence. The ability to determine who will develop PTH persistence can guide the clinician when determining which subjects should receive earlier treatment for PTH, preventing its persistence. Determination of persistence will also enhance the ability to enroll appropriate participants into clinical trials of therapeutics for PTH, enrolling those individuals who are likely to have persistence and excluding those that are likely to have quick natural resolution.

Some embodiments of the methods and processes described herein may provide one or more of the following advantages. First, some embodiments described disclose a statistical algorithm that uses subject-provided responses during a predetermined amount of time (e.g., about 1 to about 30 days) to determine outcomes for the subject identified as having migraine and/or PTH. The subject-provided responses do not require an extensive commitment, are non- invasive, and can give subjects and health care providers insight into the efficacy of treatment quickly and non-invasively. For example, the relatively non-extensive time commitment and non-invasive methodology can increase adherence and participation of a subject.

Second, embodiments of the methods and processes described herein can determine, with high specificity and sensitivity, if a migraine mitigation treatment will be effective for an individual patient. In some examples, this determination is made within 1-30 days of a patient starting or increasing the dose of a migraine mitigation medication. This can reduce the amount of time that a subject spends taking a medication that is ineffective. In contrast, the processes and methods described herein can confirm the effectiveness of a treatment or medication faster than conventional methods alone.

Third, embodiments of the methods and processes described herein can determine which subjects with PTH due to traumatic brain injury and/or concussion will develop PTH persistence. This can help clinicians and health care providers determine the management of care for subjects with acute PTH (e.g., start treatment vs. observation, change in medication, and frequency of follow-up visits).

Fourth, embodiments of the methods and processes described herein can determine which subjects with migraine or PTH can be beneficially enrolled and/or randomized in clinical trials of PTH. For example, a four-week baseline diary run-in phase is customary practice in headache clinical trials. The diary data from the run-in phase is used to determine if a subject continues to meet eligibility criteria and should be randomized. The methods and processes described herein can substantially enhance this process in PTH clinical trials by identifying those subjects who are most likely to have continued PTH and therefore are most appropriate for clinical trial participation.

One aspect of this document features methods for treating a disease or disorder associated with head pain. The methods can include, or consist essentially of, (a) receiving one or more inputs from a subject having migraine, wherein the one or more inputs are received over a predetermined amount of time; (b) extracting one or more features from the one or more inputs; (c) analyzing the one or more features by determining at least the slope, intercept, and frequency of the one or more features; (d) recommending a treatment plan for the subject based on the analysis of the one or one or more features; and (e) providing one or more treatments to the subject based on the recommendation. In some embodiments, this document features methods for determining a therapeutic outcome for a disease or disorder associated with head pain. The methods can include or consist essentially of, (a) receiving one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time; (b) extracting one or more features from the one or more inputs; (c) analyzing the one or more features by determining at least the slope, intercept, and frequency of the one or more features; (d) determining a therapeutic outcome based on the analysis of the one or more features; and (e) providing one or more treatments to the subject based on the determination of the therapeutic outcome.

In some embodiments, methods can include prior to (a) administering an initial therapy for the treatment of the disease or disorder associated with head pain. In some embodiments, methods can include (e) providing one or more treatments that are different from the initial therapy. In some embodiments, (e) comprises providing one or more treatments that comprise an altered dose or regimen of the initial therapy. In some embodiments, the initial therapy is a drug compound.

In some embodiments, this document features methods of determining a susceptibility of a subject to a disease or disorder associated with head pain. The methods can include or consist essentially of, (a) receiving one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time; (b) extracting one or more features from the one or more inputs; (c) analyzing the one or more features by determining at least the slope, intercept, and frequency of the one or more features; (d) determining a susceptibility of the subject to the disease or disorder associated with head pain based on the analysis of the one or more features; and (e) providing one or more treatments to the subject based on the determined susceptibility. In some embodiments, the one or more inputs comprise answers to one or more predetermined questions. In some embodiments, the one or more predetermined questions comprise questions directed to a presence or absence of a headache, a severity of a headache, a degree of debilitation caused by a headache, a presence or absence of a medication taken for a headache, a peak of severity of a headache, and combinations thereof. In some embodiments, the one or more inputs are received by a diary that is associated with the subject. In some embodiments, the diary is a computer based diary. In some embodiments, the computer based diary is an application on a mobile device. In some embodiments, the one or more treatments comprise a drug compound.

In some embodiments, the drug compound is one or more of antihypertensive, an anticonvulsant, an anti-calcitonin gene-related peptide (CGRP) monoclonal antibody, anti-CGRP small molecule antagonists, an antidepressant, botulinum toxin, nitric oxide, orexin, glutamate, N SAIDS, an antihistamine, memantine, or monoclonal antibodies, or a combination thereof. In some embodiments, the one or more treatments comprise an enrollment to a study. In some embodiments, the treatment is vitamins, supplements, use of neuromodulation devices, lifestyle change, or acupuncture or a combination thereof. In some embodiments, the predetermined amount of time is about 1 day to about 30 days. In some embodiments, the predetermined amount of time is about 7 days. In some embodiments, the predetermined amount of time is about 28 days. In some embodiments, the disease or disorder associated with head pain is migraine. In some embodiments, the disease or disorder associated with head pain is traumatic brain injury. In some embodiments, the subject is a human that has been identified as having migraine headaches. In some embodiments, the subject is a human that has been identified as having post-traumatic headache. In some embodiments, the one or more inputs are received by a computing device. In some embodiments, the computing device extracts the one or more features of the one or more inputs. In some embodiments, the computing device comprises a processor and a memory device storing instructions that when executed by the processor analyze the one or more features by determining at least the slope, intercept, and frequency of the one or more features. In some embodiments, the analysis is performed by executing a regression algorithm.

In another aspect, this document features a non-transitory, computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations. In some embodiments, the executable instructions can cause one or more computers to (a) receive one or more inputs from a subject having migraine, wherein the one or more inputs are received over a predetermined amount of time; (b) extract one or more features from the one or more inputs; (c) analyze the one or more features by determining at least the slope, intercept, and frequency of the one or more features; and (d) generate a recommendation for a treatment plan for the subject based on the analysis of the one or one or more features, wherein the subject is treated based on the generated recommendation. In another aspect, this document features a non-transitory, computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations. In some embodiments, the executable instructions can cause one or more computers to (a) receive one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time; (b) extract one or more features from the one or more inputs; (c) analyze the one or more features by determining at least the slope, intercept, and frequency of the one or more features; and (d) generate a determination of a therapeutic outcome based on the analysis of the one or more features, wherein the subject is treated based on the generated determination.

In some embodiments, prior to (a), the instructions further comprise receiving information about an initial therapy for the treatment of the disease or disorder associated with head pain. In some embodiments, (d) comprises providing one or more treatments that are different from the initial therapy. In some embodiments, (d) comprises providing one or more treatments that comprise an altered dose or regimen of the initial therapy. In some embodiments, the initial therapy is a drug compound. In some embodiments, the instructions upon such execution, cause the one or more computers to perform operations including (e) determine a susceptibility of the subject to the disease or disorder associated with head pain based on the determination of the therapeutic outcome; and (f) generate a recommendation of one or more treatments to the subject based on the determined susceptibility. In some embodiments, the one or more inputs comprise answers to one or more predetermined questions. In some embodiments, the one or more predetermined questions are directed to a presence or absence of a headache, a severity of a headache, a degree of debilitation caused by a headache, a presence or absence of a medication taken for a headache, a peak of severity of a headache, and combinations thereof.

In some embodiments, the one or more inputs are received by a diary that is associated with the subject. In some embodiments, the diary is a computer based diary. In some embodiments, the computer based diary is an application on a mobile device. In some embodiments, the subject is treated with one or more drug compounds. In some embodiments, the one or more drug compounds comprise one or more of antihypertensive, an anticonvulsant, an anti-CGRP monoclonal antibody, anti-CGRP small molecule antagonists, an antidepressant, botulinum toxin, nitric oxide, orexin, glutamate, NSAIDS, an antihistamine, memantine, or monoclonal antibodies, or a combination thereof. In some embodiments, the subject is treated by being enrolled to a study. In some embodiments, the subject is treated with vitamins, supplements, use of neuromodulation devices, lifestyle change, or acupuncture or a combination thereof. In some embodiments, the predetermined amount of time is about 1 day to about 30 days. In some embodiments, the predetermined amount of time is about 7 days. In some embodiments, the predetermined amount of time is about 28 days.

In some embodiments, the analyzing of the one or more features include a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope or a combination thereof. In some embodiments, the disease or disorder associated with head pain is migraine. In some embodiments, the subject is a human that has been identified as having migraine headaches. In some embodiments, the disease or disorder associated with head pain is traumatic brain injury. In some embodiments, the subject is a human that has been identified as having post-traumatic headache.

Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Although methods and materials similar or equivalent to those described herein can be used to practice the invention, suitable methods and materials are described below. All publications, patent applications, patents, and other references mentioned herein are incorporated by reference in their entirety. In case of conflict, the present specification, including definitions, will control. In addition, the materials, methods, and examples are illustrative only and not intended to be limiting.

The details of one or more embodiments of the invention are set forth in the accompanying drawings and the description below. Other features, objects, and advantages of the invention will be apparent from the description and drawings, and from the claims.

DESCRIPTION OF THE DRAWINGS

FIG. 1 shows an example schematic of embodiments of the methods and processes described herein to determine if a migraine mitigation treatment will be effective for an individual patient. FIG. 2 shows an example diary with questions programmed for a predetermined amount of time of 28 days.

FIG. 3 shows an example diary with questions programmed for a predetermined amount of time of 7 days.

FIGS. 4A-4B are box plots showing the distribution of headache frequency in the sample data set over time. The sample data set shown here was taken from a research study and tracks the response of subjects with migraine to new treatment regimens. This data set was used to train and test the initial migraine improvement models. The bar of each box indicates the median of the group at that time-point. The decreasing trend indicates the median headache frequency of the population was decreasing over time, shows headache frequency by week. FIG. 4 A shows imputed headache frequency of the data set used to create the migraine improvement model frequency by week. FIG. 4B shows the imputed headache frequency of the data set used to create the migraine improvement model by month. FIG. 4C shows the imputed moderate or severe headache frequency of the data set used to create the migraine improvement model by month.

FIG. 5 shows an exemplary method of treating a disease or disorder associated with head pain.

FIG. 6 shows an exemplary method of determining a therapeutic outcome for a disease or disorder associated with head pain.

FIG. 7 shows an exemplary method of determining a susceptibility of a subject to a disease or disorder associated with head pain.

FIG. 8 schematically depicts an exemplary non-transitory, computer-readable medium storing instruction than when executed by a computer cause the computer to perform operations that generate a recommendation for a treatment for a subject.

FIG. 9 schematically depicts an exemplary non-transitory, computer-readable medium storing instruction that when executed by a computer cause the computer to perform operations that generate a determination for a therapeutic outcome.

DETAILED DESCRIPTION

As described herein, examples of a disease or disorder associated with head pain described herein include without limitation, migraine, PTH, and other head pain associated disease or disorder (e g., tension headache). Examples of a subject described herein include, but are not limited to, a human that has been identified as having a disease or disorder associated with head pain, diagnosed as having a disease or disorder associated with head pain, suspected as having a disease or disorder associated with head pain.

FIG. 1 shows an example schematic of embodiments of the methods and processes described herein to determine if a migraine mitigation treatment will be effective for an individual patient. For example, a patient can utilize a prediction tool (e.g., a computer diary) to record symptoms over about one month. If the prediction tool determines that a migraine mitigation treatment will be effective the tool can recommend that the treatment plan remains about the same. If the prediction tool determines that a migraine mitigation treatment will not be effective, the tool can recommend a different treatment plan.

In an example, workflow, a patient will visit with a headache clinician and receive an initial treatment plan. In some examples, a treatment plan can include a protocol, a therapy, and/or a medication (e.g., a prescription). For instance, the patient can receive an initial treatment plan that includes an initial migraine mitigation medication. Following the visit and for the next month (e.g., about 28 days), the patient can complete a headache diary. Following the 28 days, a prediction tool will use the headache diary entries and apply a statistical algorithm. The algorithm can include diary data preparation (e.g., diary data processed for missing or illogical inputs, and diary responses recoded), a feature extraction (e.g., linear regression equation is fit on each variable from the diary, slope and intercept of each variable-specific linear regression are features in the decision function, and headache frequency over time periods calculated), and a decision function (e.g., extracted features are inputs to the decision function), produces a numeric value 0-1, numeric value may be converted to a binary prediction.

Based on the binary prediction, a clinician can provide a recommendation for a treatment plan (e.g., to change treatment, begin treatment, remain on the current treatment, change dosages, and/or stop treatment). For example, a patient’s headache clinician can use a statistical algorithm output to make a recommendation of whether the patient should remain on the same mitigation medication or try a new mitigation medication. For example, the output can indicate that the patient has about an 80% chance of not responding to the initial medication and the clinician can determine to switch a treatment plan. In another example, the output can indicate that the patient has about an 80% chance of responding to the in initial medication and the clinician can determine to remain on the treatment plan. As shown in FIG. 1, this determination is made within 1-30 days of a patient starting or increasing the dose of a migraine mitigation medication. This can reduce the amount of time that a subject spends on a treatment plan taking a medication that is ineffective. In contrast, the processes and methods described herein can confirm the effectiveness of a treatment plan or medication faster than conventional methods alone.

In some embodiments, a diary as described herein is a computer-based diary that is associated with the subject. For example, the diary can be an application that is used by the subject and includes demographic and/or medical data corresponding to the subject. In some instances, the diary can include historical medical data about the subject.

The diary can be deployed on a computing device. Examples of computing devices described herein include without limitation, a desktop computer, a laptop, a tablet, a phablet, a smartwatch, a smartphone, or an internet of things device. In some examples, the diary can exist across multiple computing devices simultaneously. For example, a user can have a profile (e.g., demographic data, medical data, etc.) that is accessible via a web program, or an application installed on a computing device or across multiple computing devices (e.g., a smart phone and a laptop of a user). The computing device can include one or more processors that are configured to perform operations by executing one or more software instructions stored in one or more memory devices. The computing device can include one or more processing units that are configured to perform operations by executing one or more software instructions. The one or more processing units can include one or more central processing units (CPUs), one or more graphical processing units (GPUs), or any combination thereof. The computer can be configured to interact with the memory, or a programmable circuit (e.g., field programmable gate array or application specific integrated circuit) directly via a direct connection such one or more busses, one or more USB cables, one or more USB-C cables, the like, or any combination thereof. Alternatively, or in addition, the computing device can be configured to interact with the memory, or the programmable circuit via one or more networks. The one or more networks can include a wired Ethernet network, a wireless network, an optical network, a LAN, a WAN, a cellular network, the Internet, or any combination thereof.

In some instances, a subject is provided access to the diary by a medical provider to collect one or more inputs from the subject. For example, a medical provider can configure the diary to include one or more predetermined questions and provide the diary to the subject. For example, the diary can be configured to include one or more predetermined questions for a subject to answer at predetermined intervals (e.g., daily) for a predetermined amount of time (e.g., 7, 14, or 28 days). An input, as described herein, can include an answer to one or more questions that are presented to the subject via the diary.

One or more inputs can be collected on a daily or other frequency basis for a predetermined amount of time. An input as described herein can include an alphanumerical value that a user can enter to the diary. Inputs can represent answers to one or more predetermined questions over a predetermined amount of time. In some instances, a predetermined amount of time is from about 1 day to about 30 days. For example, one or more inputs can be collected by the diary in response to one or more questions for a predetermined amount of time of about 1 day to about 7 days, about 1 day to about 14 days, about 1 day to about 28 days, or about 1 day to about 30 days, without limitation. In some embodiments, one or more inputs can be collected daily by the diary in response to one or more questions for a predetermined amount of time of, for example, 7 days. In some embodiments, one or more inputs can be collected daily by the diary in response to one or more questions for a predetermined amount of time of about 14 days. In some embodiments, one or more inputs can be collected daily by the diary in response to one or more questions for a predetermined amount of time of about 28 days.

When the predetermined amount of time expires, one or more features can be extracted from the one or more inputs provided by the user via the diary. As used herein, a feature is an attribute of an input. An input can be binary (e.g., yes or no, presence or absence) or qualitative (e.g., a severity). For example, the presence or absence of a headache, the severity of the headache, etc. In another example, one or more features can be extracted from the one or more inputs collected over the predetermined amount of time. In some incidences, extracted features are data that has been identified from the one or more inputs collected by the diary during the predetermined amount of time. In some instances, one or more features are extracted from the one or more inputs after the predetermined amount of time has expired. For example, the one or more inputs can be collected over a predetermined amount of time and the one or more features can be extracted based on the one or more inputs after the predetermined amount of time expires. In some instances, one or more features are extracted from the one or more inputs periodically during the predetermined amount of time. For example, the one or more inputs can be collected for a predetermined amount of time and the one or more features can be extracted at predetermined intervals (e.g., daily, weekly, every other day, etc.) during the predetermine amount of time.

The extracted features can be analyzed. Examples of questions, input options, extracted features, and extracted feature analysis include those presented in Table 1. Table 1: Example questions presented via a diary, input options, extracted features, and extracted feature analysis. The Example question response column shows different representations of the answers that are given to the question in the “Example Questions” column. For example, if a subject said "Yes" they had a headache then a "1" can be recorded in the diary, if "No" then a "0" can be recorded in the diary. In some embodiments, the one or more inputs are answers to one or more predetermined questions. For example, the one or more predetermined questions can be directed to a presence or absence of a headache, a severity of a headache, a degree of debilitation caused by a headache, a presence or absence of a medication taken for a headache, a peak of severity of a headache, and combinations thereof.

In some examples, example questions other than those listed in Table 1 can include, without limitation: What time did the headache begin? What time did the headache end? How long did the headache remain at its maximum intensity? Headache location? In some embodiments, the questions can be in relation to a treatment response. For example, if a degree of debilitation is indicated further questions can include, without limitation: How disabling was the migraine? and; How difficult it was to do normal activity?

Examples of an analysis of the one or more features described herein without limitation include a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope.

In some examples, methods of extraction of the features can include, without limitation frequency domain representation as determined by Fourier transform, means, medians, minimums, maximums, percentage change, change between two sequential values of analyzed features, and combinations thereof.

In some examples, example questions other than those listed in Table 1 can include, without limitation: What time did the headache begin? What time did the headache end? How long did the headache remain at its maximum intensity? Headache location?

In some embodiments, the analysis of the extracted features can be extrapolated to a decision function within the diary. For example, the decision function of the diary can provide a determined therapeutic outcome, a recommended treatment, and/or determine a susceptibility to a disease or disorder associated with head pain. The decision function includes an algorithm of logistic regression equations. The equations include one or more of the extracted features (e.g., an analysis of the extracted features).

Below is an example of a regression equation that can be used to generate a decision function. Migraine treatment response model equation M4

The output of the equation is a number between 0 and 1. This number can be interpreted as a probability. It can also be used in conjunction with a threshold value in order to make a binary prediction of whether an individual will, for example, respond to a new migraine treatment. In this use, if the numeric output of the equation is greater than the threshold then the individual will be predicted to respond to the treatment (e.g., patient is a treatment responder). If the numeric output is less than the threshold, the individual would be predicted not to respond to the treatment (e.g., patient is a non-responder). The threshold value can be used to tune the algorithm to have the performance characteristics desirable in the specific application.

For example, a clinical application of this invention is to quickly determine whether a patient will respond to a newly initiated migraine mitigation medication. This determination will help the patient and their clinician decide whether to continue on that medication or stop it and switch to another However, due to the limited number of migraine mitigation treatment options, it is important that medication treatment trials not be ended prematurely, if there is a reasonable chance of the medication providing meaningful benefits. To address this, the negative predictive value of these models can be maximized, to reduce the occurrence of predicting that a medication will not provide benefits when in fact it likely will Ideally, a person who is predicted to not have treatment benefit will in fact be a non-responder. By adjusting the probability threshold that can determine how an individual is classified, the predictive model can be adjusted to maximize its negative predictive value. In the study data (n=^;:15), adjusting the threshold to 40% increased the negative predictive value to 100%, thus indicating that patients who were predicted to be medication non-responders were in fact medication non-responders.

The models can also be used to predict alternative outcomes, such as a 30% reduction in headache days in response to a new migraine treatment or a 20% reduction in migraine days in response to a new⁷ migraine treatment.

In some examples, the decision function of the diary can output a determination of a therapeutic outcome, a determination of a susceptibility of a subject to a disease or disorder associated with head pain, recommendation of a treatment alteration, or a recommendation for a treatment for the subject. In some embodiments, the output of the decision function can be shared with the subject. In some embodiments, the output of the decision function can be shared with the medical provider. In some embodiments, the output of the decision function can be shared with both the subject and the medical provider. In some embodiments, the output of the decision function can be automatically shared with the medical provider and/or the subject upon completion of the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider upon completion of the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time.

The methods and processes described herein can determine a therapeutic outcome for subjects having a disease or disorder associated with head pain (e.g., migraine or PTH). In some embodiments, a therapeutic outcome can be determined based on the analysis of the one or more extracted features. As described herein, the diary can include the decision function that can determine a therapeutic outcome based on analysis of one or more of the extracted features (e.g., a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope). Examples of a therapeutic outcome described herein include without limitation, a determination of whether or not a subject will recover after an amount of time (e.g., 1 week, 2 weeks, 3, weeks, 4 weeks, 2 months, 3 months, etc.), a determination of a subject’s response to an initial treatment, and a determination of a subject’s susceptibility to a persistence to a disease or disorder associated with head pain.

The decision function included in the diary can provide subjects and health care providers a determination and/or a recommendation for a treatment plan (e.g., start a treatment, change dosage, stop a treatment, continue a treatment, change a treatment) based on the analysis of one or more features extracted from one or more inputs collected from the subject in response to questions posed by the diary. In some embodiments, a subject can be provided one or more treatments based on the determination and/or recommendation for a treatment plan. Examples of treatments described herein include without limitation include an administration of a medication (e g., a medicinal compound), a change in medication (e.g., a medicinal compound), enrollment in a clinical study, recommendation of lifestyle change (e.g., behavior, dietary, etc.), recommendation for acupuncture, recommendation for neuromodulation, recommendation for physical therapy, and recommendation for use in disability claims.

Non-limiting categories of medicinal treatment are headache mitigation treatment and headache abortive treatment. Mitigation treatment is taken whether or not a headache is present at the time, with the intention of reducing the frequency, severity and disability from headache. Abortive treatment is taken when there is a headache at that time, with the intention of reducing headache intensity, shortening the duration of the headache, and reducing disability. Examples of medications described herein include without limitation an antihypertensive, an anticonvulsant, an anti-CGRP monoclonal antibody, anti-CGRP small molecule antagonists, an antidepressant, botulinum toxin, nitric oxide, orexin, glutamate, NSAIDS an antihistamine, memantine, or monoclonal antibodies. In some embodiments, the drug compound is one or more of a medication targeting: pituitary adenylate cyclase-activating polypeptide (PACAP), adenosine, the delta-opioid receptor, potassium channels, transient receptor potential (TRP) ion channels, acid sensing ion channels, the N-Methyl-D-Aspartate (NMD A) receptor, or a combination thereof.

Examples of an antihypertensive medication include without limitation propranolol, timolol, and metoprolol. Examples of an anticonvulsant medication include without limitation divalproex sodium, and topiramate. Examples of a CGRP monoclonal antibody medication include without limitation erenumab, fremanezumab, galcanezumab, and etpinezumab. Examples of an anti-CGRP small molecule antagonist medication include without limitation include gepants (e.g., rimegepant, atogepant). Examples of an antidepressant medication include without limitation amitriptyline, nortripytline, venlafaxine, and duloxetine. Other medications that can be administered include without limitation amitriptyline and onabotulinumtoxinA.

Examples of non-medication can include, vitamins (riboflavin), supplements (magnesium, CoQlO), neuromodulation devices (supraorbital nerve stimulation, occipital nerve stimulation, transcranial magnetic stimulation, remote electrical neuromodulation, vagus nerve stimulation) and combinations thereof. In some examples, the extracted feature can correspond to a predetermined amount of time. For example, FIG. 2 shows an example diary with questions programmed for a predetermined amount of time of 28 days. FIG. 2 also shows the extracted features and variables as corresponding to various amounts of time within the predetermined time period of 28 days.

In another example, FIG. 3 shows an example diary with questions programmed for a predetermined amount of time of 7 days. FIG. 3 shows the extracted features and variables as corresponding to an amount of time within the predetermined time period of 7 days with a desired determination of recovery (yes or no) at 3 months.

In some embodiments, the predetermined amount of time that the one or more inputs are collected via the diary, the specific questions provided to the subject, the one or more extracted features, and the decision function (e.g., the algorithm) are dependent on the disease or disorder associated with head pain being evaluated and the time-point a determination is wanted for.

For example, the predetermined amount of time that the one or more inputs are collected via the diary, the specific questions provided to the subject, the one or more extracted features, and/or the decision function (e.g., the algorithm) can be different depending on the disease or disorder associated with head pain and/or amount of time that a determination is configured to be accurate. Said differently, a decision function that determines an outcome at 3 months (e.g., FIG. 3) can have a different predetermined amount of time that the one or more inputs are collected via the diary, different specific questions provided to the subject, and different extracted features, than a decision function that determines an outcome at 9-12 weeks (e.g., FIG. 2).

The methods and process disclosed herein can advantageously reduce the amount of time that a patient and their clinician are required to spend to find effective treatment for diseases or disorders associated with head pain (e.g., migraine or PTH) treatment (e.g., mitigation treatment) since it can provide an accurate determination of migraine mitigation treatment response within four weeks of starting a treatment. The information provided by the methods and process disclosed herein can be used in clinical practice when the clinician and patient are determining whether to continue with a migraine mitigation therapy or to switch to a different migraine mitigation therapy. For example, subjects can provide data within an electronic diary during the four weeks following the initiation of a particular migraine mitigation therapy. As described herein, based on the provided diary inputs, a decision function can determine the probability of migraine mitigation treatment response. The patient and clinician can use the decision function output as a factor when deciding whether to continue the particular medication or to switch to another. The decision function can also be useful during clinical trials that are testing the efficacy of different treatments for diseases or disorders associated with head pain, assisting with “go” vs. “no-go” decisions and shortening the amount of time to complete the trials.

The methods and process disclosed herein provide early determination of the likelihood that a disease or disorders associated with head pain (e.g., PTH) will persist. For example, the determination can be used by subjects and their clinicians when determining if PTH mitigation treatment should be initiated early after onset of PTH. For example, subjects can provide data within an electronic diary that includes a decision function. When the decision function output suggests that the probability of PTH persistence is high, interventions to reduce the likelihood of PTH persistence would be initiated. When the decision function output suggests that the probability of PTH persistence is low, patient exposure to PTH mitigation interventions could be avoided. As described herein, the data for determination of PTH persistence are collected using a PTH electronic diary that executes the decision function. Subjects would start providing data in the diary as soon as possible after the onset of PTH. The decision function output from the diary would be shared with the patient and their clinician, who would use the information when making early treatment decisions regarding the use of PTH mitigation treatments, the need for close clinical follow-up (e.g., the frequency of clinical office visits), and the need for patient referrals to clinicians with expertise in PTH.

FIG. 5 shows an example method 100 of treating a disease or disorder associated with head pain. In some embodiments, the method 100 includes receiving, at step 102 one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time. In some examples, the one or more inputs provided by a subject into a diary. For example, a diary can be deployed on a computing device and configures to include one or more questions for the subject to respond to in a given predetermined amount of time. The questions and the predetermined amount time can be determined based on the disease or disorder associated with head pain (e.g., PTH or migraine).

The method 100 can further comprise extracting, at step 104, one or more features from the one or more inputs. The one or more features can be attributes of the inputs. For example, the one or more features can be a severity of a headache. In some examples, an extracted feature can be the presence or absence of head pain.

The method 100 can further comprise analyzing, at step 106, the one or more features by determining at least the slope, intercept, and frequency of the one or more features. For example, an analysis of the one or more features described herein without limitation include a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope. The method 100 can further comprise recommending, at step 108, a treatment for the subject based on the analysis of the one or more features. For example, the diary can deploy a decision function. The decision function can recommend a treatment based on the analyzed extracted features. In some embodiments, the recommended treatment can be to start a medication, stop a medication, or change a medication. In some embodiments, via the diary, the output of the decision function can be automatically shared with the medical provider and/or the subject during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider upon completion of the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time.

The method 100 can further comprise providing, at step 110, one or more treatments to the subject based on the recommendation. For example, a subject can receive a treatment based on the recommendation by the diary. In some embodiments, a health care provider can receive the recommendation and provide the treatment based on the recommendation.

FIG. 6 shows an example method 200 of treating a disease or disorder associated with head pain. In some embodiments, the method 200 includes receiving, at step 202 one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time. In some examples, the one or more inputs provided by a subject into a diary. For example, a diary can be deployed on a computing device and configures to include one or more questions for the subject to respond to in a given predetermined amount of time. The questions and the predetermined amount time can be determined based on the disease or disorder associated with head pain (e.g., PTH or migraine).

The method 200 can further comprise extracting, at step 204, one or more features from the one or more inputs. The one or more features can be attributes of the inputs. For example, the one or more features can be a severity of a headache. In some examples, an extracted feature can be the presence or absence of head pain.

The method 200 can further comprise analyzing, at step 206, the one or more features by determining at least the slope, intercept, and frequency of the one or more features. For example, an analysis of the one or more features described herein without limitation include a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope.

The method 200 can further comprise determining, at step 208, a therapeutic outcome based on the analysis of the one or more features. For example, the diary can deploy a decision function. The decision function can determine a therapeutic outcome including, a determination of whether or not the subject will recover after an amount of time (e.g., 1 week, 2weeks, 3, weeks, 4 weeks, 2 months, 3 months, etc.), a determination of the subject’s response to an initial treatment, a determination of a subject’s susceptibility to a persistence to a disease or disorder associated with head pain. In some embodiments, via the diary, the output of the decision function can be automatically shared with the medical provider and/or the subject during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider upon completion of the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time.

The method 200 can further comprise providing, at step 210, one or more treatments to the subject based on the determination of the therapeutic outcome. For example, a subject can receive a treatment based on the determining by the diary. In some embodiments, a health care provider can receive the determination and provide the treatment based on the determination.

FIG. 7 shows an example method 300 of treating a disease or disorder associated with head pain. In some embodiments, the method 300 includes receiving, at step 302 one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time. In some examples, the one or more inputs provided by a subject into a diary. For example, a diary can be deployed on a computing device and configures to include one or more questions for the subject to respond to in a given predetermined amount of time. The questions and the predetermined amount time can be determined based on the disease or disorder associated with head pain (e.g., PTH or migraine). The method 300 can further comprise extracting, at step 304, one or more features from the one or more inputs. The one or more features can be attributes of the inputs. For example, the one or more features can be a severity of a headache. In some examples, an extracted feature can be the presence or absence of head pain.

The method 300 can further comprise analyzing, at step 306, the one or more features by determining at least the slope, intercept, and frequency of the one or more features. For example, an analysis of the one or more features described herein without limitation include a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope.

The method 300 can further comprise determining, at step 308, a susceptibility of the subject to the disease or disorder associated with head pain based on the analysis of the one or more features. For example, the diary can deploy a decision function. The decision function can determine a susceptibility of the subject to develop symptoms associate with the disease or disorder associated with head pain, a susceptibility of the subject’s response to an initial treatment, a determination of a subject’s susceptibility to a persistence to a disease or disorder associated with head pain. In some embodiments, via the diary, the output of the decision function can be automatically shared with the medical provider and/or the subject during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider upon completion of the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time.

The method 300 can further comprise providing, at step 310, one or more treatments to the subject based on the determination of the susceptibility. For example, a subject can receive a treatment based on the determination of the susceptibility by the diary. In some embodiments, a health care provider can receive the determination of the susceptibility and provide the treatment based on the determination.

FIG. 8 schematically depicts an example non-transitory, computer-readable medium 420 storing software comprising instructions executable by one or more computers 410 which, upon such execution, cause the one or more computers to perform operations comprising receive, at 422 one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time. In some examples, the one or more inputs provided by a subject into a diary that is configured with the non-transitory, computer-readable medium 420 and deployed on a computing device (e.g., computer 410). In some examples, the diary is not deployed on a computing device but is instead deployed on a server or cloud device. In some examples, a diary can be deployed on a computing device and configures to include one or more questions for the subject to respond to in a given predetermined amount of time. The questions and the predetermined amount time can be determined based on the disease or disorder associated with head pain (e.g., PTH or migraine). The instructions stored in the non-transitory, computer-readable medium 420 can further extract, at 424, one or more features from the one or more inputs. The one or more features can be attributes of the inputs. For example, the one or more features can be a severity of a headache. In some examples, an extracted feature can be the presence or absence of head pain. The instructions stored in the non-transitory, computer-readable medium 420 can further analyze, at 426, the one or more features by determining at least the slope, intercept, and frequency of the one or more features. For example, an analysis of the one or more features described herein without limitation include a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope. The instructions stored in the non-transitory, computer-readable medium 420 can further generate a recommendation, at 428, for a treatment for the subject based on the analysis of the one or more features, wherein the subject is treated based on the generated recommendation. For example, the diary can deploy a decision function. The decision function can recommend a treatment based on the analyzed extracted features. In some embodiments, the recommended treatment can be to start a medication, stop a medication, or change a medication. In some embodiments, via instructions executed by the computer, the output of the decision function can be automatically shared with the medical provider and/or the subject during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider upon completion of the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time. In this instance, one or more treatments can be provided to the subject based on the recommendation. For example, a subject can receive a treatment based on the recommendation by the diary. In some embodiments, a health care provider can receive the recommendation and provide the treatment based on the recommendation.

FIG. 9 schematically depicts an example non-transitory, computer-readable medium 520 storing software comprising instructions executable by one or more computers 510 which, upon such execution, cause the one or more computers to perform operations comprising receive, at 522 one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time. In some examples, the one or more inputs provided by a subject into a diary that is configured with the non-transitory, computer-readable medium 520 and deployed on a computing device (e.g., computer 510). In some examples, the diary is not deployed on w computing device but is instead deployed on w server or cloud device. In some examples, a diary can be deployed on a computing device and configures to include one or more questions for the subject to respond to in a given predetermined amount of time. The questions and the predetermined amount time can be determined based on the disease or disorder associated with head pain (e.g., PTH or migraine). The instructions stored in the non-transitory, computer-readable medium 520 can further extract, at 524, one or more features from the one or more inputs. The one or more features can be attributes of the inputs. For example, the one or more features can be a severity of a headache. In some examples, an extracted feature can be the presence or absence of head pain. The instructions stored in the non-transitory, computer-readable medium 520 can further analyze, at 526, the one or more features by determining at least the slope, intercept, and frequency of the one or more features. For example, an analysis of the one or more features described herein without limitation include a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope. The instructions stored in the non-transitory, computer-readable medium 520 can further generate a determination, at 528, of a treatment for the subject based on the analysis of the one or more features, wherein the subject is treated based on the generated determination. For example, the diary can deploy a decision function. The decision function can determine a therapeutic outcome including, a determination of whether or not the subject will recover after an amount of time (e g., 1 week, 2 weeks, 3, weeks, 4 weeks, 2 months, 3 months, etc ), a determination of the subject’s response to an initial treatment, a determination of a subject’s susceptibility to a persistence to a disease or disorder associated with head pain. In some embodiments, via instructions executed by the computer, the output of the decision function can be automatically shared with the medical provider and/or the subject during intervals (e.g., daily, weekly, etc.) within the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider upon completion of the predetermined amount of time. In some embodiments, a subject can be prompted to share the output of the decision function with the medical provider during intervals (e g., daily, weekly, etc.) within the predetermined amount of time. In this instance, one or more treatments can be provided to the subject based on the generated determination. For example, a subject can receive a treatment based on the generated determination by the decision function. In some embodiments, a health care provider can receive the recommendation and provide the treatment based on the generated determination.

The invention will be further described in the following examples, which do not limit the scope of the invention described in the claims.

EXAMPLES

Example 1. Migraine treatment response: outcome determined for 9-12 weeks.

Individuals with migraine frequently must try multiple mitigation treatments, one after the next, before finding one that is effective for them, meaning it substantially reduces their headache frequency. Typically, each new treatment is trialed for a minimum of 3 months before a determination of its efficacy can be made. The process of cycling through multiple ineffective treatments before finding one that works can therefore take years. This invention can help reduce the time required to find an effective migraine mitigation treatment by determining if a treatment will result in a significant reduction in headache frequency after a much shorter time. The following example shows the process for determining after one month if an individual is likely to experience a significant reduction in headache frequency 9-12 weeks after starting treatment. Migraine treatment change: outcome determined for one time-point: 9-12 weeks. See FIG. 2.

Diary

The following items are collected via an electronic headache diary.

• Did you have a headache today? (1 : yes, 0: no): referred to as variable headache

• How disabling was the headache/migraine? (1 : Able to function normally, 2 Function somewhat reduced, 3 Function severely reduced, 4: Not able to function at all (e.g., bedbound)): referred to as variable disabling

• In the past 24 hours, overall, how difficult was it to do your usual activities? (1 : Not difficult, 2: A little difficult, 3 : Moderately difficult, 4: Very difficult, 5: Extremely difficult): referred to as variable difficulty

Feature Extraction

Three types of features are extracted from the diary data.

1) Headache frequency over a specified time-period. Example: headache frequency over the first 7 days = # “yes” responses/7

2) Slope of the variables (e.g., inputs). The slope of a line fitted to the question responses (e.g., inputs) over time. This gives information on the trend.

3) Intercept of the variables (e.g., inputs). The intercept of a line fitted to the question responses (e.g., inputs) over time.

The slope and intercept are as if you plotted the response value for each diary response day and drew a line through the points and found the slope of that line.

Features used to determine headache reduction: headache slope, headache intercept, headache frequency in the first 7 days, and headache frequency in the first 14 days, disabling intercept, difficult intercept.

Example 2: Accurate Determination of Post-Traumatic Headache Persistence Using One

Week of Headache Diary Data

Introduction

PTH is a common symptom following mTBI. Currently, there is no accurate way of determining which patients will recover quickly from PTH and which patients will continue to have PTH months or years after mTBI. This inability to determine headache persistence is frustrating to patients and a treatment dilemma to clinicians. Daily headache diary data over 7 days was used to determine headache persistence at 3 months.

Methods

Individuals with acute PTH due to mTBI completed a daily headache e-diary beginning 1-59 days post mTBI. Diary information about daily headache presence or absence, headache intensity (mild, moderate, severe), and medication use for headache (yes/no) were utilized. Accuracy for determining headache persistence was assessed from the first 7 days of headache diary data. Headache frequency over the first 7 days was calculated, and additional features related to trends over time were generated from the headache diary data over the same period. Missing diary data were imputed using the carry-forward method.

Results

PTH persistence at three months was determined with 78% leave-one-out cross validation accuracy using the headache frequency and trend features related to headache incidence, headache intensity, and medication use generated from the first 7 days of diary data. This determination model achieved 73% accuracy on an unseen sample of patients with PTH.

Example 3: Model exploration

Introduction

Two models were evaluated in this example. Both use the first 28 days of diary data to determine a 50% decrease in headache frequency in weeks 9-12. The first model uses headache frequency over the first 7 and 14 days along with features from the “headache” (presence), “disability” and “difficult” diary questions to achieve 88% repeated cross validation accuracy on the validation set (not used to train). The ROC AUC on the validation data was 0.84. The kappa is 0.69 which is rated as moderate to good. The sensitivity is 0.94 and specificity is 0.71. Performance of the second model is similar except the cross-validation ROC AUC was 0.80. This model used the first 28 days of diary data, the headache frequency over the first 7 days and features from the “headache” (presence), “intensity” and “duration” diary questions. The migraine treatment response models were developed using data from two migraine study cohorts:

1^st cohort: 720 individuals with migraine. After cleaning and feature extraction a subset of samples were used to train the decision function. The model was tested on a holdout test set from that study. The same process was used to generate a model using a second migraine study (2^a,i cohort). Each of the models was tested on the data from the other study. Additionally, each model was tested on data from the ongoing prospective validation study (n=15).

The PTH models were developed using data from the ongoing NIH and DOD studies (n=43) described at the start of the application. They are tested on new data as it is collected.

Cleaning:

1. Open the diary files.

2. Remove data collected outside of relevant time period or collected in error.

3. Remove rows where headache y/n and intensity, duration and disabling are all NA.

4. For diary dates with multiple entries on the same date, choose the one with the fewest NAs. This will tend to be the data with headache vs not, more complete.

5. If a headache duration or intensity was reported but was reported that they did not have a headache, change the headache status to yes.

6. If they reported taking an abortive on a day they did not have a headache change it to no abortive

7. Create a full calendar of diary entries from first to last diary entry with blanks where there is no real diary entry. a. Find the earliest (valid) diary entry per person. b. Create data frame with empty.

8. Fill in missing data a. When there was no headache reported, i. change “abortive” to no b. for every date when there is no entry, fill in with previous date and mark as “missing” ->imputed_expanded.csv c. for rows with some preexisting data but other missing data i. fill in intensity, disabling, duration and difficult with values with 0 if there was no headache that day or values from the previous headache day if there was a headache that day. If there was no previous headache day then fill in with average over all headache days. ii. fill in “move_body”, “move_head”, “bed”, “physical” and “bend” with values from previous day with same headache status.

Save the output of this step as input to train the model.

Tabulate number of diary entries available in first 7, 14, 28 days in original data. This will be used to select valid data to train and test model on.

Calculate response variable for use in training or evaluating a model.

1. Assign each diary day to a 28-day month beginning with the first diary day with valid data. a. Days (56-84] -> weeks 9-12

2. Imputed_ha<- 28/diary_entries * diary_headaches

3. Baseline ha28 <- Baseline.Days.Headache../28.' (imputed from Baseline. Days. Mod. Sev.Headache../30d.')

4. Wk_9.12_ha_improved<- baseline_ha28 - imputed ha

5. Wk_9.12.ha_pct <- wk_9.12_ha_improve/baseline_ha28

6. Wk_9.12.ha_50pct<- wk_9.12.ha_pct >=0.5

Table 2: Model Parameters used to generate algorithm. Time_periods determines the length(s) of the time periods whose data will be used to generate the slope and intercept features. The other parameters in this table are used to select which patient’s data should be used in the modeling.

Select Data

Data has valid response value (0, 1)

Data has the required number of diary entries in the first 7, 14, 28 days and in the response period. For example, Min 7 diary is the required number of diary entries in the first 7 days. Min_14_diary is the required number of diary entries in the first 14 days. etc.

Split into train, validate and test sets

Stratified split on response variable and number of diary entries

Train: 80%, validate: 10%, test: 10%

Table 3: Potential variables.

Pred rfe\ main function to take the cleaned features, extract analyzed features and fit the models while selecting the best set of analyzed features.

1. For each time period in time_periods and each diary variable chosen fit a linear model per person and extract the slope and intercept as features. 2. Find the headache frequency over the first 7 and 14 days per person as features.

3. Fit the model on the features created. Use reverse feature elimination with 10-fold repeated cross validation with 5 repeats to identify best features and fit model. Upsample and use¹ subsets of 1 :25 variables for fitting. These steps can be accomplished with existing tools such as R Caret. a. RfeControl: a control object which allows you to set the values to be used in rfe. b. Train control. c. Rfe: backward feature selection algorithm.

The functions a, b, and c are functions from the tool (e.g., Classification and Regression Training package (Caret)).

Table 4: Model control parameters Table 5 is a summary of performance metric results from the development of the migraine treatment response determination model.

Models 1, 2 and 3 used a 14 day diary period. Model 5 used 14- and 28-day diary periods together and models 4,6 and 7 used 28 day diary period.

Model #4: period = 28 days, 7 day &14 day headache frequency. Potential variables (*included in final model): headache*, abortive today, 5 intensity, disability*, difficult*.

Model #7: period = 28 days, 7-day headache frequency. Potential variables: headache*, intensity*, duration*

FIGS. 4A-4C: Box plots showing the distribution of headache frequency in the sample data set over time. The sample data set shown here was taken from a research study and tracks the response of migraineurs to new treatment regimens. This data set was used to train and test the initial migraine improvement models. The bar of each box indicates the median of the group at that timepoint. The decreasing trend indicates the median headache frequency of the population was decreasing over time.

FIG. 4A shows imputed headache frequency of the data set used to create the migraine improvement model by week.

FIG. 4B shows the imputed headache frequency of the data set used to create the migraine improvement model by month.

FIG. 4C Shows the imputed moderate or severe headache frequency of the data set used to create the migraine improvement model by month.

Example 4: Algorithm development for PTH and Migraine

Introduction

A headache diary (e.g., diary) is used to collect daily records (e.g., inputs) of a person’s (e.g., a subject’s) experience with headache. The diary data is the input to determine if the subject’s headache frequency is likely to decrease or not. Three applications were evaluated. 1) determining if a person with PTH is likely to experience a significant decrease in headache frequency during weeks 9-12 following the start of the headache diary compared to the headache frequency reported at the start of the diary period, adjusted for headache frequency pre-TBI. 2) determining if a person with PTH is likely to experience a significant decrease in headache frequency during the sixth month following the start of the headache diary compared to the headache frequency at the start of the diary period adjusted for pre- TBI headache frequency. 3) determining if a person with migraine is likely to experience a significant decrease in headache frequency in the 9-12 weeks following migraine mitigation treatment changes compared to the patient’s monthly headache frequency prior to the treatment change.

Method Application for PTH To determine if a person with PTH is likely to experience a significant decrease in headache frequency 9-12 weeks after starting the headache diary, the person completes a headache diary for a predetermined amount of time. The questions in the headache diary are configured to the diary and include questions on the presence or absence of headache, headache features and symptoms, and medication use.

At the end of the predetermined amount of time, the method assesses the individual’s diary data for missing or incorrect data and data imputation methods are employed when applicable.

Custom feature values are then derived (e.g., extracted features) from the individual’s imputed diary data (e.g., inputs). These features were developed during algorithm development and concern summary values and trends in the diary data.

The custom features are input into the PTH determination algorithm. The algorithm was developed and tested using historical data. It was trained and tested on separate data sets, a rigorous and reproducible method for determination modeling.

For determining PTH outcomes at weeks 9-12, diary data from a 7-day period was collected. The algorithm takes in the custom features and returns a determination of whether the subject is likely to experience a significant decrease in headache frequency during weeks 9-12 following the time they begin diary entry. The determination output is a percentage (likelihood) or a binary value (likely/unlikely) that reflects the likelihood of PTH improvement.

To determine if a PTH patient is likely to experience a significant decrease in headache frequency six months after the start of the headache diary the same process occurs with several differences 1) the time period over which diary data can be collected is 28 days; 2) the diary data used for determination is different; and 3) the algorithm is specific to determination of significant PTH reduction at 6 months following the start of the diary entry rather than determination of reduction at 9-12 weeks.

Method Application for Migraine

Similarly, there is a different algorithm for determining if an individual with migraine is likely to experience a significant decrease in headache frequency 9-12 weeks following a change in migraine mitigation treatment. Twenty-eight days of diary data is collected. The questions asked in the diary are configured to that outcome. The data used to develop and test the migraine related algorithm were different than the data used to develop and test the PTH algorithms.

Creation of the PTH and Migraine Outcome Determination Algorithms

Development and testing of the determination algorithms utilized data from studies. Missing data were imputed. From the imputed dataset, multiple trend features and summary features were extracted over a time-period or periods.

The extracted feature dataset was used to train classification models which output the probability of the specified objective occurring. Each model is a linear combination of the extracted features that maximizes the sum of probabilities of observing the responses. Each final algorithm is a combination of data collection, and features, to make the determination model for that algorithm’s specific objective.

OTHER EMBODIMENTS

It is to be understood that while the invention has been described in conjunction with the detailed description thereof, the foregoing description is intended to illustrate and not limit the scope of the invention, which is defined by the scope of the appended claims. Other aspects, advantages, and modifications are within the scope of the following claims.

Claims

WHAT IS CLAIMED IS:

1. A method of treating migraine, comprising:

(a) receiving one or more inputs from a subject having migraines, wherein the one or more inputs are received over a predetermined amount of time;

(b) extracting one or more features from the one or more inputs;

(c) analyzing the one or more features by determining at least a slope, an intercept, and a frequency of the one or more features;

(d) recommending a treatment plan for the subject based on the analysis of the one or one or more features; and

(e) providing one or more treatments to the subject based on the recommendation.

2. A method of determining a therapeutic outcome for a disease or disorder associated with head pain, comprising:

(a) receiving one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time;

(b) extracting one or more features from the one or more inputs;

(c) analyzing the one or more features by determining at least a slope, an intercept, and a frequency of the one or more features;

(d) determining a therapeutic outcome based on the analysis of the one or more features; and

(e) providing one or more treatments to the subject based on the determination of the therapeutic outcome.

3. The method of 2, wherein, prior to (a), the method comprises administering an initial therapy for the treatment of the disease or disorder associated with head pain.

4. The method of 3, wherein (e) comprises providing one or more treatments that are different from the initial therapy.

5. The method of 3, wherein (e) comprises providing one or more treatments that comprise an altered dose or regimen of the initial therapy.

6. The method of any one of claims 2-5, wherein the initial therapy comprises a drug compound.

7. A method of determining a susceptibility of a subject to a disease or disorder associated with head pain, comprising:

(a) receiving one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time;

(b) extracting one or more features from the one or more inputs;

(c) analyzing the one or more features by determining at least the slope, intercept, and frequency of the one or more features;

(d) determining a susceptibility of the subject to the disease or disorder associated with head pain based on the analysis of the one or more features; and

(e) providing one or more treatments to the subject based on the determined susceptibility.

8. The method of any one of claims 1-7 wherein the one or more inputs comprise answers to one or more predetermined questions.

9. The method of claim 8, wherein the one or more predetermined questions comprise questions directed to a presence or absence of a headache, a severity of a headache, a degree of debilitation caused by a headache, a presence or absence of a medication taken for a headache, a peak of severity of a headache, and combinations thereof.

10. The method of any one of claims 1-9, wherein the one or more inputs are received by a diary that is associated with the subject.

11. The method of claim 10, wherein the diary comprises a computer based diary.

12. The method of claim 11, wherein the computer based diary comprises an application on a mobile device.

13. The method of any one of claims 1-12, wherein the one or more treatments comprises a drug compound.

14. The method of claim 13, wherein the drug compound is one or more of antihypertensive, an anticonvulsant, an anti-calcitonin gene-related peptide (CGRP) monoclonal antibody, anti-CGRP small molecule antagonists, an antidepressant, oa botulinum toxin, nitric oxide, orexin, glutamate, NSAIDS an antihistamine, memantine, or monoclonal antibodies. .

15. The method of any one of claims 1-14, wherein the one or more treatments comprises an enrollment to a study.

16. The method of any one of claims 1-15, wherein the treatment is vitamins, supplements, use of neuromodulation devices, lifestyle change, or acupuncture or a combination thereof.

17. The method of any one of claims 1-16, wherein the predetermined amount of time comprises about 1 day to about 30 days.

18. The method of any one of claims 1-17, wherein the predetermined amount of time comprises about 7 days.

19. The method of any one of claims 1-17, wherein the predetermined amount of time comprises about 28 days.

20. The method of any one of claims 1-19, wherein the one or more features comprise a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope or a combination thereof.

21. The method of any one of claims 1-20, wherein the one or more inputs are received by a computing device.

22. The method of any one of claims 21, wherein the computing device extracts the one or more features of the one or more inputs.

23. The method of claim 21 or claim 22, wherein the computing device comprises a processor and a memory device storing instructions that when executed by the processor analyze the one or more features by determining at least the slope, intercept, and frequency of the one or more features.

24. The method of any one of claims 1-23 wherein the analysis is performed by executing a regression algorithm.

25. The method of any one of claims 2-24, wherein the disease or disorder associated with head pain comprises migraine.

26. The method of any one of claims 2-24, wherein the subject is a human that has been identified as having migraine headaches.

27. The method of any one of claims 2-24, wherein the disease or disorder associated with head pain comprises a traumatic brain injury.

28. The method of any one of claims 2-24, wherein the subject is a human that has been identified as having post-traumatic headache.

29. A non-transitory, computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:

(a) receive one or more inputs from a subject having migraine, wherein the one or more inputs are received over a predetermined amount of time;

(b) extract one or more features from the one or more inputs;

(c) analyze the one or more features by determining at least the slope, intercept, and frequency of the one or more features; and

(d) generate a recommendation for a treatment plan for the subject based on the analysis of the one or one or more features, wherein the subject is treated based on the generated recommendation.

30. A non-transitory, computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:

(a) receive one or more inputs from a subject having a disease or disorder associated with head pain, wherein the one or more inputs are received over a predetermined amount of time;

(b) extract one or more features from the one or more inputs;

(c) analyze the one or more features by determining at least the slope, intercept, and frequency of the one or more features; and

(d) generate a determination of a therapeutic outcome based on the analysis of the one or more features, wherein the subject is treated based on the generated determination.

31. The computer-readable medium of 30, wherein, prior to (a), the instructions further comprise receiving information about an initial therapy for the treatment of the disease or disorder associated with head pain.

32. The computer-readable medium of 31, wherein (d) comprises providing one or more treatments that are different from the initial therapy.

33. The computer-readable medium of 32, wherein (d) comprises providing one or more treatments that comprise an altered dose or regimen of the initial therapy.

34. The computer-readable medium of claim 31, wherein the initial therapy comprises drug compound.

35. The computer-readable medium of any one of claims 30-34 wherein the instructions upon such execution, cause the one or more computers to perform operations comprising:

(e) determine a susceptibility of the subject to the disease or disorder associated with head pain based on the determination of the therapeutic outcome; and

(f) generate a recommendation of one or more treatments to the subject based on the determined susceptibility.

36. The computer-readable medium of any one of claims 29-35 wherein the one or more inputs comprise answers to one or more predetermined questions.

37. The computer-readable medium of claim 36, wherein the one or more predetermined questions comprise questions directed to a presence or absence of a headache, a severity of a headache, a degree of debilitation caused by a headache, a presence or absence of a medication taken for a headache, a peak of severity of a headache, and combinations thereof.

38. The computer- readable medium of any one of claims 29-37, wherein the one or more inputs are received by a diary that is associated with the subject.

39. The computer-readable medium of claim 38, wherein the diary comprises a computer based diary.

40. The computer-readable medium of claim 39, wherein the computer based diary comprises an application on a mobile device.

41. The computer- readable medium of any one of claims 29-40, wherein the subject is treated with one or more drug compounds.

42. The computer-readable medium of claim 41, wherein the one or more drug compounds comprise one or more of antihypertensive, an anticonvulsant, an anti-calcitonin gene-related peptide (CGRP) monoclonal antibody, anti-CGRP small molecule antagonists, an antidepressant, a botulinum toxin, nitric oxide, orexin, glutamate, NSA1DS an antihistamine, memantine, or monoclonal antibodies.

43. The computer-readable medium of any one of claims 29-42, wherein the subject is treated by being enrolled to a study.

44. The computer-readable medium of any one of claims 29-43, wherein the subject is treated with vitamins, supplements, use of neuromodulation devices, lifestyle change, or acupuncture or a combination thereof.

45. The computer- readable medium of any one of claims 29-44, wherein the predetermined amount of time comprises about 1 day to about 30 days.

46. The computer-readable medium of any one of claims 29-45, wherein the predetermined amount of time comprises about 7 days.

47. The computer-readable medium of any one of claims 295-45, wherein the predetermined amount of time comprises about 28 days.

48. The computer- readable medium of any one of claims 29-47, wherein the analyzing of the one or more features include a headache frequency, a headache intercept, a headache slope, a difficult frequency, a difficult intercept, a difficult slope, a disabling frequency, a disabling intercept, a disabling slope, an intensity frequency, an intensity intercept, an intensity slope, a medication frequency, a medication intercept, and a medication slope or a combination thereof.

49. The computer-readable medium of any one of claims 30-48, wherein the disease or disorder associated with head pain comprises a migraine.

50. The computer-readable medium of any one of claims 30-48, wherein the subject is a human that has been identified as having migraine headaches.

51. The computer-readable medium of any one of claims 30-48, wherein the disease or disorder associated with head pain comprises a traumatic brain injury.

52. The computer-readable medium of any one of claims 30-48, wherein the subject is a human that has been identified as having post-traumatic headache.