This disclosure relates to systems and methods for determining human performance. More specifically, this disclosure relates to systems and methods for determining whether a cancer patient will need unplanned medical care during cancer therapy.
Biomechanical characterization of human performance is known. Using biomechanical characterization of human performance to inform decisions about oncological therapy in an effort to reduce or avoid a need for unplanned medical care (e.g., caused by deterioration of a cancer patient) is also known. However, typical biomechanical characterization of human performance for oncological or other reasons often comprises either a qualitative assessment by medical personnel, or an invasive biomechanical characterization test. These require significant experimental setup that includes numerous sensors. In addition, qualitative assessments are difficult to standardize due to their intrinsically subjective nature. Invasive tests provide reliable information but are not feasible for large scale applications.
One aspect of the disclosure relates to a system configured to determine whether a cancer patient will need unplanned medical care during cancer therapy. The system comprises one or more sensors, one or more processors, and/or other components. The one or more sensors may be configured to generate output signals conveying spatial position information related to spatial positions of one or more anatomical sites on the cancer patient while the cancer patient performs a prescribed movement. The one or more anatomical sites may comprise an anatomical site that corresponds to a center of mass of the cancer patient, and/or other anatomical sites indicative of mobility of a cancer patient—e.g., a spine base, a knee, a hip, etc.
The one or more processors may be configured to determine one or more kinematic parameters indicative of the movement of the cancer patient during the prescribed movement based on the spatial position information. The one or more kinematic parameters may comprise an acceleration and/or other kinematic parameters of the anatomical site that corresponds to the center of mass of the cancer patient and/or other anatomical sites indicative of mobility. The one or more processors may be configured to determine whether the cancer patient will need unplanned medical care during cancer therapy based on the acceleration and/or other kinematic parameters of the anatomical site that corresponds to the center of mass and/or other anatomical sites indicative of mobility of the cancer patient.
Another aspect of the disclosure relates to a system configured to determine whether a cancer patient will need unplanned medical care during cancer therapy. The system comprises one or more sensors, one or more processors, and/or other components. The one or more sensors may be configured to generate output signals conveying physical activity information related to physical activity performed by the cancer patient. The one or more processors may be configured to determine one or more physical activity parameters indicative of the physical activity of the cancer patient based on the physical activity information. The one or more physical activity parameters may comprise metabolic equivalence (METs). The one or more processors may be configured to determine whether the cancer patient will need unplanned medical care during cancer therapy based on the metabolic equivalence of the cancer patient.
Still another aspect of the disclosure relates to a method for determining whether a cancer patient will need unplanned medical care during cancer therapy with a determination system. The system may comprise one or more sensors, one or more processors, and/or other components. The method comprises generating, with the one or more sensors, output signals conveying spatial position information related to spatial positions of one or more anatomical sites on the cancer patient while the cancer patient performs a prescribed movement. The one or more anatomical sites may comprise an anatomical site that corresponds to a center of mass of the cancer patient and/or other anatomical sites indicative of mobility of the cancer patient. The method may comprise determining, with the one or more processors, one or more kinematic parameters indicative of the movement of the cancer patient during the prescribed movement based on the spatial position information. The one or more kinematic parameters may comprise an acceleration of the anatomical site that corresponds to the center of mass of the cancer patient and/or other kinematic parameters indicative of mobility of the cancer patient. The method may comprise determining, with the one or more processors, whether the cancer patient will need unplanned medical care during cancer therapy based on the acceleration of the anatomical site that corresponds to the center of mass of the cancer patient and/or other kinematic parameters indicative of the mobility of the cancer patient.
Yet another aspect of the disclosure relates to a method for determining whether a cancer patient will need unplanned medical care during cancer therapy with a determination system. The system comprises one or more sensors, one or more processors, and/or other components. The method comprises generating, with the one or more sensors, output signals conveying physical activity information related to physical activity performed by the cancer patient. The method comprises determining, with the one or more processors, one or more physical activity parameters indicative of the physical activity of the cancer patient based on the physical activity information. The one or more physical activity parameters may comprise metabolic equivalence (METs). The method may comprise determining, with the one or more processors, whether the cancer patient will need unplanned medical care during cancer therapy based on the metabolic equivalence of the cancer patient.
It should be noted that, in some embodiments, the patient need not be a cancer patient, and the unplanned medical care may be sought during any future period of time. In some embodiments, the systems and methods described herein may be applied to one or more other cell proliferative disorders, and/or other disorders all together.
These and other objects, features, and characteristics of the system and/or method disclosed herein, as well as the methods of operation and functions of the related elements of structure and the combination of parts and economies of manufacture, will become more apparent upon consideration of the following description and the appended claims with reference to the accompanying drawings, all of which form a part of this specification, wherein like reference numerals designate corresponding parts in the various figures. It is to be expressly understood, however, that the drawings are for the purpose of illustration and description only and are not intended as a definition of the limits of the invention. As used in the specification and in the claims, the singular form of “a”, “an”, and “the” include plural referents unless the context clearly dictates otherwise.
Observing the way a patient moves provides a clinician with valuable information about frailty. This is important for patients undergoing difficult treatments such as chemotherapy. A comprehensive geriatric (e.g., frailty) assessment can predict complications and side effects from cancer treatment. However, clinicians' assessments are often qualitative, subjective, and lack agreement among clinicians. Available tools and metrics such as the Eastern Cooperative Oncology Group (ECOG) performance status, body mass index (BMI) measurements, Mini Mental State Exam (MMSE) results, and the Charlson Comorbidity Index (CCI), are often part of a comprehensive geriatric assessment, but few clinicians perform a complete assessment because such assessments are time consuming.
Laboratory based invasive methods have been developed to biomechanically quantify elements of human performance. Many of these methods comprise conducting gait analysis using an accelerometer, a gyroscope, and other types of wearable sensors and motion capture systems to detect and differentiate conditions in patients with osteoarthritis, neuromuscular disorders, and cerebral palsy. However, these methods are associated with high cost, lengthy time required to perform tests, and general difficulty in interpreting results.
Although these tools and metrics are known, and continue to be used because of their practicality, standardization of patient stratification, and speed of assessment; inter- and intra-observer variability, gender discrepancies, sources of subjectivity in physician assigned performance assessments, and a lack of standard conversions between different evaluation scales continue to exist. As such, there is a need for a system and method for more objective classification of a patient's physical function that may be used to guide decisions about oncological therapy in an effort to reduce or avoid a need for unplanned medical care.
Advantageously, system 100 is a non-invasive motion-capture based performance assessment system which can (i) determine kinematic parameters that characterize a cancer patient's biomechanical performance and/or physical activity parameters that characterize a level of physical activity of the cancer patient, and (ii) determine whether a cancer patient will need unplanned medical care during cancer therapy based on the kinematic and/or physical activity parameters. In some embodiments, system 100 comprises one or more of a body position sensor 102; a physical activity sensor 104; computing platform 114 comprising a processor 106, a user interface 116 and electronic storage 118; external resources 120; and/or other components.
Body position sensor 102 may be configured to generate one or more output signals conveying spatial position information and/or other information. The spatial position information and/or other information may be a time series of information that conveys spatial position information about the body and/or body parts of a cancer patient over time. In some embodiments, the spatial position information may comprise visual information representing the body and/or individual body parts of the cancer patient, and/or other information. The visual information representing the cancer patient may include one or more of still images, video images, and/or other information. For example, body position sensor 102 may be configured such that the spatial position information includes body position signals conveying information associated with the position of one or more body parts of the cancer patient relative to each other and/or other reference locations. In some embodiments, the visual information may be and/or include a wire-frame representation of the cancer patient and/or other visual information. According to some embodiments, body position sensor 102 may include an infrared stereoscopic sensor configured to facilitate determination of user body positions, such as for example the KinectlM available from Microsoft™ of Redmond, Wash., and/or other sensors.
Body position sensor 102 may be configured such that the spatial information comprises information associated with one or more body positions and/or other physical characteristics of the cancer patient. The spatial position information in the output signals may be generated responsive to a prescribed movement performed by the cancer patient and/or at other times. A given body position may describe, for example, a spatial position, orientation, posture, and/or other positions of the cancer patient and/or of one or more body parts of the cancer patient. A given physical characteristic may include, for example, a size, a length, a weight, a shape, and/or other characteristics of the cancer patient, and/or of one or more body parts of the cancer patient. The output signals conveying the spatial position information may include measurement information related to the physical size, shape, weight, and/or other physical characteristics of the cancer patient, movement of the body and/or one or more body parts of the cancer patient, and/or other information. The one or more body parts of the cancer patient may include a portion of the first user's body (e.g., one or more of a head, neck, torso, foot, hand, head, arm, leg, and/or other body parts).
The spatial position information may be related to spatial positions of one or more anatomical sites on the cancer patient. The one or more anatomical sites may be and/or correspond to the body parts described above, for example. The one or more anatomical sites may comprise an anatomical site (e.g., a body part) that is indicative of a patient's mobility, corresponds to a center of mass of the cancer patient, and/or include other anatomical sites. In some embodiments, locations that are indicative of a patient's mobility and/or correspond to the center of mass may be a location at a base of a spine of the cancer patient, a location near a hip or hips, a location near a knee, and/or other locations.
By way of a non-limiting example,
The spatial position information (e.g., from body position sensor 102 shown in
By way of a non-limiting example,
Returning to
In some embodiments, as described above, body position sensor 102 and/or physical activity sensor 104 may be stand-alone devices, separate from one or more other components of system 100, and communicate with one or more other components of system 100 (e.g., computing platform 114) as a peripheral device. In some embodiments, body position sensor 102 and/or physical activity sensor 104 may be integrated with computing platform 114 as a single device (e.g., as a camera that is part of computing platform 114, as an activity tracking sensor built into computing platform 114, etc.). In some embodiments, body position sensor 102, physical activity sensor 104, and/or computing platform 114 may be associated with the cancer patient and/or may be carried by the cancer patient. For example, body position sensor 102 and/or physical activity sensor 104 may be included in a Smartphone associated with the cancer patient. As such, information related to physical activity of the cancer patient may be obtained throughout the day as the cancer patient goes about his daily business and/or participates in specific activities.
Although body position sensor 102 and physical activity sensor 104 are depicted in
Computing platform 114 may include one or more processors 106, a user interface 116, electronic storage 118, and/or other components. Processor 106 may be configured to execute computer program components. The computer program components may be configured to enable an expert or user associated with a given computing platform 114 to interface with system 100 and/or external resources 120, and/or provide other functionality attributed herein to computing platform 114. By way of non-limiting example, computing platform 114 may include one or more of a desktop computer, a laptop computer, a handheld computer, a tablet computing platform, a Smartphone, a gaming console, and/or other computing platforms.
Processor 106 is configured to provide information-processing capabilities in computing platform 114 (and/or system 100 as a whole). As such, processor 106 may comprise one or more of a digital processor, an analog processor, a digital circuit designed to process information, an analog circuit designed to process information, a state machine, and/or other mechanisms for electronically processing information. Although processor 106 is shown in
As shown in
It should be appreciated that although components 108, 110, 112, and 113 are illustrated in
Communication component 108 may be configured to facilitate bi-directional communication between computing platform 114 and one or more other components of system 100. In some embodiments, the bi-directional communication may facilitate control over one or more of the other components of system 100, facilitate the transfer of information between components of system 100, and/or facilitate other operations. For example, communication component 108 may facilitate control over body position sensor 102 and/or physical activity sensor 104 by a user (e.g., the cancer patient, a doctor, a nurse, a caregiver, etc.). The control may be based on entries and/or selections made by the user via user interface 116, for example, and/or based on other information. As another example, communication component 108 may facilitate uploading and/or downloading data to or from body position sensor 102, physical activity sensor 104, external resources 120, and/or other components of system 10.
Continuing with this example, communication component 108 may be configured to receive the spatial information and/or the physical activity information in the output signals from body position sensor 102 and/or physical activity sensor 104. The output signals may be received directly and/or indirectly from body position sensor 102 and/or physical activity sensor 104. For example, body position sensor 102 may be built into computing platform 114, and the output signals from body position sensor 102 may be transmitted directly to communication component 108. As another example, physical activity sensor 104 may be a separate wrist worn device. The output signals from the wrist worn device may be wirelessly transmitted to communication component 108.
In some embodiments, communication component 108 may be configured to cause display (e.g., on user interface 116) of the spatial information, the physical activity information, a determination, and/or other information. In some embodiments, communication component 108 may be configured to cause display (e.g., on user interface 116) of a graphical control interface to facilitate user control of body position sensor 102, physical activity sensor 104, and/or other components of system 100.
Pre-processing component 110 is configured to pre-process the spatial information, the physical activity information, and/or other information received by communication component 108. In some embodiments, pre-processing comprises filtering, converting, normalizing, adjusting, and/or other pre-processing operations performed on the spatial information, the physical activity information, and/or other information in the output signals from body position sensor 102, physical activity sensor 104, and/or other components of system 100. In some embodiments, pre-processing component 110 may be configured to automatically segment (and/or facilitate manually segmenting) the spatial information to trim irrelevant data at the beginning and end of a prescribed movement while a patient is stationary. Preprocessing component 110 may be configured to pre-process the spatial information to compensate for irregularities in the spatial information caused by the positioning of body position sensor 102 relative to a given cancer patient, features of an environment or location where the prescribed movement occurs, and/or other factors. In some embodiments, pre-processing component 110 may be configured such that pre-processing includes coordinate transformation for three-dimensional data coordinates included in the spatial information. For example, the spatial information received by communication component 108 may be distorted such that a level plane such as a clinic floor appears sloped in the spatial information, for example. In this example, the angle of distortion, 8, may range between about 5 and about 20°. Pre-processing component 110 may be configured to resolve this distortion by performing an automated element rotation about an x-axis of the spatial information. As other examples, in some embodiments, pre-processing may include filters to remove other background humans from the images prior to analysis during the CTT exam; and, for a wrist worn sensor (e.g., as described herein), pre-processing may include adjustments for weight, gender, race, time, diet, and location prior to calculation of metabolic equivalents.
Parameter component 112 may be configured to determine one or more kinematic parameters, physical activity parameters, and/or other parameters. Parameter component 112 may be configured to determine the one or more kinematic and/or physical activity parameters based on the information in the output signals from body position sensor 102 and/or physical activity sensor 104, the pre-processing performed by pre-processing component 110, and/or other information. In some embodiments, the one or more determined kinematic and/or physical activity parameters may be features extracted from the spatial position or physical activity information, and/or other parameters. In some embodiments, the determined kinematic and/or physical activity parameters may comprise less bytes of data than the spatial position information and/or the physical activity information conveyed by the one or more output signals.
In some embodiments, parameter component 112 may be configured to determine one or more kinematic parameters indicative of the movement of the cancer patient during the prescribed movement based on the spatial position information and/or other information. The one or more kinematic parameters may comprise one or more positions of a given anatomical site (e.g., 1-20 shown in
In some embodiments, determining the one or more kinematic parameters indicative of the movement of the cancer patient during the prescribed movement based on the spatial position information comprises determining anatomical site position vectors for the one or more anatomical sites. The anatomical site position vectors may comprise three-dimensional time series generated for given positions of the one or more anatomical sites at time points (e.g., 402, 404, 406, 408 shown in
By way of a non-limiting example, a position vector
{right arrow over (η)}(t)=xi(t), yi(t), zi(t)
for an anatomical site i may be used to calculate the anatomical site's velocity magnitude,
v
i(t)=∥{right arrow over (η)}(t)∥
and acceleration magnitude,
a
i(t)=∥{right arrow over (ri)}(t)∥
using the mean-value theorem. In the absence of distribution of mass information, specific kinetic energy,
ke
i(t)=½vi2(t)
and specific potential energy
pe
i(t)=gΔzi=g(zi(t)−zi(t=1))
quantities may be used to describe the energy signature of each anatomical site. Parameter component 112 may be configured such that the sagittal angle, θs(t), is defined as the angle formed between the vector originating at the spine base and pointing in the direction of motion, and the vector connecting the anatomical sites for the spine base (e.g., 1 in
In some embodiments, parameter component 112 may be configured to determine one or more physical activity parameters indicative of the physical activity of the cancer patient based on the physical activity information and/or other information. In some embodiments, the one or more physical activity parameters may comprise an amount of time a cancer patient engages in physical activity, a level (e.g., low or high, above or below a predetermined threshold level, etc.) of the physical activity, an amount of energy expended during the physical activity, an amount of calories burned during the physical activity, metabolic equivalence (METs) associated with the physical activity, and/or other parameters. In some embodiments, parameter component 112 may be configured to aggregate (e.g., sum, average, etc.), normalize, and/or perform other operations for the one or more physical activity parameters for a given evaluation period (e.g., per hour, per day, per week, for the time between doctor visits, etc.). In some embodiments, parameter component 112 may be configured to aggregate a given physical activity parameter for the evaluation period only for instances of physical activity that breach a predetermined threshold level during the evaluation period.
For example, in some embodiments, parameter component 112 may be configured to determine total (e.g., a summation of) METs associated with physical activity performed by the cancer patient during the evaluation period. In some embodiments, a total number of METs may be an indication of any and all physical activity by a cancer patient during an evaluation period. METs provide an indication of an amount of energy consumed while sitting at rest relative to an amount of energy consumed while performing a physical activity. In some embodiments, METs may be calculated based on a determination of mechanical work completed. One MET, for example, is equal to 1.1622 watts/kg, where a watt of work is equal to the energy required to move an object at constant velocity of one meter/second against a force of one Newton. Acceleration against force may be determined by integration of a directional force vector from a three-axis accelerometer sensor (e.g., as described herein) and correcting for the weight of the wearer, for example.
In some embodiments, parameter component 112 may be configured such that only METs associated with high levels of physical activity (e.g., physical activity that breaches a predetermined threshold level) may be included in the total. In some embodiments, parameter component 112 may be configured to determine total daily, weekly, or monthly active hours above a threshold of, for example, 1.5 METs (light), 3METs (moderate), or 6 METs (vigorous) physical activity. In some embodiments, parameter component 112 may determine a fraction of daytime hours spent in non-sedentary activity. Total distance travelled and steps taken may be alternative measures of activity, for example.
The physical activity parameters determined by parameter component 112, aggregation operations, threshold levels, and/or other characteristics of parameter component 112 may be determined at manufacture of system 100, determined and/or adjusted by a user via user interface 116, and/or determined in other ways.
Determination component 113 may be configured to determine whether a cancer patient will need unplanned medical care. In some embodiments, the determination of whether the cancer patient will need unplanned medical care during cancer therapy is indicative of a future reaction of the cancer patient to chemotherapy and/or radiation during cancer therapy. In some embodiments, the determining may be based on the acceleration (in any direction) of the anatomical site that corresponds to the center of mass of the cancer patient (e.g., the spine base) and/or other information. In some embodiments, determination component 113 may be configured to determine whether the cancer patient will need unplanned medical care during cancer therapy based on relative accelerations (and/or any other motion parameters) of anatomical sites. For example, determination component 113 may be configured to determine whether the cancer patient will need unplanned medical care based on a comparison of a first acceleration of a first anatomical site to one or more second accelerations of one or more second anatomical sites. In some embodiments, determination component 113 may be configured to determine whether a cancer patient will need unplanned medical care based on acceleration of an anatomical site relative to a reference site (e.g., an exam table, a patient bed, a computer, and/or other reference sites).
In some embodiments, the determining may be based on the metabolic equivalence determined for the cancer patient, and/or other information.
In some embodiments, determining whether the cancer patient will need unplanned medical care during cancer therapy may comprise determining whether the cancer patient will need unplanned medical care during a future period of time that corresponds to one or more cancer therapy treatments received by the cancer patient. In some embodiments, the future period of time is about two months and/or other periods of time. This example is not intended to be limiting.
In some embodiments, determination component 113 may be configured such that determining whether the cancer patient will need unplanned medical care comprises comparing the acceleration of the center of mass of the cancer patient to an acceleration threshold, comparing the METs for the cancer patient to a METs threshold, and/or comparing other parameters to other thresholds, and determining the cancer patient will need unplanned medical care during cancer therapy responsive to a breach of one or more of the thresholds. By way of a non-limiting example, in some embodiments, the spine base acceleration threshold may be about one meter per second squared (1 m/s2), and the METs threshold may be about zero waking hours above 1.5METs (these are merely examples). Determination component 113 may be configured such that if the acceleration of the spine base is in breach of (e.g., below in this example) the spine base acceleration threshold, and/or if the METs are in breach of (e.g., below in this example) the METs threshold, the cancer patient is determined to need unplanned medical care. These examples are not intended to be limiting. The thresholds may be any thresholds on any parameters that are indicative of whether the cancer patient will need unplanned medical care during cancer therapy. In some embodiments, the thresholds may be determined at manufacture of system 100, determined and/or adjusted based on entries and/or selections made by a user via user interface 116, learned by determination component 113 (e.g., as described below), and/or determined in other ways.
In some embodiments, determination component 113 may be configured such that determining whether the cancer patient will need unplanned medical care comprises comparing a spine base acceleration (and/or other parameter) time series (e.g., determined as described above) and/or a physical activity (e.g., as indicated by METs) over time dataset to a corresponding baseline and/or reference dataset. In some embodiments, determination component 113 may be configured to determine a distance between the spine base acceleration time series and/or the physical activity over time dataset and the corresponding baseline and/or reference dataset. For example, the time series for a given feature (e.g., the acceleration of the spine base) may be compared to a baseline and/or reference dataset using Euclidean metric dynamic time warping (DTW), which assigns a distance of zero for completely identical series and larger distances for more dissimilar series.
By way of a non-limiting example,
Returning to
In some embodiments, determination component 113 is configured to categorize the cancer patient as either likely to likely to need unplanned medical care or unlikely to need unplanned medical care during cancer therapy. In some embodiments, determination component 113 is configured to determine a likelihood (e.g., a numerical value on a continuous scale, a high-medium-low indication, a color representation of the likelihood, etc.) the cancer patient will need unplanned medical care, and categorize the cancer patient into two or more groups based on the likelihood. Determination component 113 may be configured such that the likelihood is inversely correlated to the acceleration of the spine base, the METs, and/or other parameters. For example, higher acceleration of a cancer patient's spine base indicates lower likelihood the cancer patient will need unplanned medical care. Similarly, the higher the number of METs for the cancer patient, the lower the likelihood the cancer patient will need unplanned medical care. In some embodiments, the categorization boundaries, the likelihood determination method, and/or other information may be determined at manufacture of system 100, determined and/or adjusted based on entries and/or selections made by a user via user interface 116, learned by determination component 113 (e.g., as described below), and/or determined in other ways.
In some embodiments, determination component 113 may be configured such that determining whether the cancer patient will need unplanned medical care and/or categorizing the cancer patient as either likely or unlikely to need unplanned medical care may include predicting ECOG scores. In some embodiments, the ECOG scores may be predicted based on the acceleration of the spine base of the cancer patient, the METs associated with the cancer patient, and/or other information, and the determination of whether or not the cancer patient will need unplanned medical care may be based on the ECOG scores.
In some embodiments, determination component 113 may be and/or include a trained prediction model. The trained prediction model may be an empirical model and/or other trained prediction models. The trained prediction model may perform some or all of the operations of determination component 113 described herein. The trained prediction model may predict outputs (e.g., whether or not the cancer patient will need unplanned medical care, ECOG scores, etc.) based on correlations between various inputs (e.g., the spatial information, the physical activity information, etc.).
As an example, the trained prediction model may be a machine learning model. In some embodiments, the machine learning model may be and/or include mathematical equations, algorithms, plots, charts, networks (e.g., neural networks), and/or other tools and machine learning model components. For example, the machine learning model may be and/or include one or more neural networks having an input layer, an output layer, and one or more intermediate or hidden layers. In some embodiments, the one or more neural networks may be and/or include deep neural networks (e.g., neural networks that have one or more intermediate or hidden layers between the input and output layers).
As an example, the one or more neural networks may be based on a large collection of neural units (or artificial neurons). The one or more neural networks may loosely mimic the manner in which a biological brain works (e.g., via large clusters of biological neurons connected by axons). Each neural unit of a neural network may be connected with many other neural units of the neural network. Such connections can be enforcing or inhibitory in their effect on the activation state of connected neural units. In some embodiments, each individual neural unit may have a summation function that combines the values of all its inputs together. In some embodiments, each connection (or the neural unit itself) may have a threshold function such that a signal must surpass the threshold before it is allowed to propagate to other neural units. These neural network systems may be self-learning and trained, rather than explicitly programmed, and can perform significantly better in certain areas of problem solving, as compared to traditional computer programs. In some embodiments, the one or more neural networks may include multiple layers (e.g., where a signal path traverses from front layers to back layers). In some embodiments, back propagation techniques may be utilized by the neural networks, where forward stimulation is used to reset weights on the “front” neural units. In some embodiments, stimulation and inhibition for the one or more neural networks may be more free flowing, with connections interacting in a more chaotic and complex fashion. In some embodiments, the intermediate layers of the one or more neural networks include one or more convolutional layers, one or more recurrent layers, and/or other layers.
The machine learning model may be trained (i.e., whose parameters are determined) using a set of training data. The training data may include a set of training samples. The training samples may include spatial information and/or physical activity information, for example, for prior cancer patients, and an indication of whether the prior cancer patients needed unplanned medical care. Each training sample may be a pair comprising an input object (typically a vector, which may be called a feature vector, which may be representative of the spatial and/or physical activity information) and a desired output value (also called the supervisory signal)—for example indicating whether unplanned medical care was needed. A training algorithm analyzes the training data and adjusts the behavior of the machine learning model by adjusting the parameters of the machine learning model based on the training data. For example, given a set of N training samples of the form {(x1, y1), (x2, y2), . . . , (xN, yN)} such that xi is the feature vector of the i-th example and yi is its supervisory signal, a training algorithm seeks a machine learning model g: X→Y, where X is the input space and Y is the output space. A feature vector is an n-dimensional vector of numerical features that represent some object (e.g., the spatial information and/or the physical activity information for a cancer patient as described above). The vector space associated with these vectors is often called the feature space. During training, the machine learning model may learn various parameters such as the spine base acceleration threshold, the METs threshold, the time series distance determination threshold, the categorization boundaries and/or other thresholds as described above. After training, the machine learning model may be used for making predictions using new samples. For example, the trained machine learning model may be configured to predict ECOG scores, whether or not a cancer patient will need unplanned medical care, and/or other information based on corresponding input spatial information and/or physical activity information for the cancer patient.
In some embodiments, determination component 113 may be configured to facilitate adjustment of the cancer therapy and/or other therapies. The adjustment may be based on the determination of whether the patient will need unplanned medical care and/or other information. In some embodiments, facilitating may comprise determining and displaying recommended changes, determining one or more additional parameters from the information in the output signals from the one or more sensors, and/or other operations. For example, based on the determination of whether the patient will need unplanned medical care, in treating a patient with a PD-L1 high expressing lung cancer, an oncologist may choose to treat a patient with a high risk with checkpoint inhibitor therapy alone, rather than a combination of chemotherapy with checkpoint inhibitor therapy. Similarly, a patient with an oral cavity squamous cell carcinoma undergoing combined chemo-radiation may be treated with a lower intensity weekly low-dose cisplatin regimen rather than a higher intensity regimen of high dose cisplatin given at 3 week intervals. Alternatively, physicians may decide to dose reduce chemotherapy to 80% (for example) of the usual standard dose prior to administration of the 1st cycle in anticipation of poor tolerability.
Body position sensor 102, physical activity sensor 104, and processor 106 may be configured to generate, determine, communicate, analyze, present, and/or perform any other operations related to the determinations, the spatial information, the physical activity information and/or any other information in real-time, near real-time, and/or at a later time. For example, the spatial information and/or physical activity information may be stored (e.g., in electronic storage 118) for later analysis (e.g., determination of a prediction). In some embodiments, the stored information may be compared to other previously determined information (e.g., threshold values, etc.), and/or other information.
As shown in
It is to be understood that other communication techniques, either hard-wired or wireless, are also contemplated by the present disclosure as user interface 116. For example, the present disclosure contemplates that user interface 116 may be integrated with a removable storage interface provided by computing platform 114. In this example, information may be loaded into computing platform 114 from removable storage (e.g., a smart card, a flash drive, a removable disk) that enables the user to customize the implementation of computing platform 114. Other exemplary input devices and techniques adapted for use with computing platform 114 as user interface 116 include, but are not limited to, an RS-232 port, RF link, an IR link, modem (telephone, cable or other). In short, any technique for communicating information with computing platform 114 and/or system 100 is contemplated by the present disclosure as user interface 116.
Electronic storage 118 may include electronic storage media that electronically stores information. The electronic storage media of electronic storage 118 may include one or both of system storage that is provided integrally (i.e., substantially non-removable) with computing platform 114 and/or removable storage that is removably connectable to computing platform 114 via, for example, a port (e.g., a USB port, a firewire port) or a drive (e.g., a disk drive). Electronic storage 118 may include one or more of optically readable storage media (e.g., optical disks), magnetically readable storage media (e.g., magnetic tape, magnetic hard drive, floppy drive), electrical charge-based storage media (e.g., EEPROM, RAM), solid-state storage media (e.g., flash drive), and/or other electronically readable storage media. Electronic storage 118 may include one or more virtual storage resources (e.g., cloud storage, a virtual private network, and/or other virtual storage resources). Electronic storage 118 may store software algorithms, information determined by processor 106, information received from external resources 120, information entered and/or selected via user interface 116, and/or other information that enables system 100 to function as described herein.
External resources 120 include sources of information such as databases, websites, etc.; external entities participating with system 100 (e.g., systems or networks that store data associated with the cancer patient), one or more servers outside of system 100, a network (e.g., the internet), electronic storage, equipment related to Wi-Fi™ technology, equipment related to Bluetooth® technology, data entry devices, or other resources. In some embodiments, some or all of the functionality attributed herein to external resources 120 may be provided by resources included in system 100. External resources 120 may be configured to communicate with computing platform 114, physical activity sensor 104, body position sensor 102, and/or other components of system 100 via wired and/or wireless connections, via a network (e.g., a local area network and/or the internet), via cellular technology, via Wi-Fi technology, and/or via other resources.
Body position sensor 102, physical activity sensor 104, computing platform 114, and/or external resources 120 may be operatively linked via one or more electronic communication links. For example, such electronic communication links may be established, at least in part, via wires, via local network using Wi-Fi, Bluetooth, and/or other technologies, via a network such as the Internet and/or a cellular network, and/or via other networks. It will be appreciated that this is not intended to be limiting, and that the scope of this disclosure includes embodiments in which body position sensor 102, physical activity sensor 104, computing platform 114, and/or external resources 120 may be operatively linked via some other communication media, or with linkages not shown in
In some embodiments, method 600 may be implemented in one or more processing devices (e.g., a digital processor, an analog processor, a digital circuit designed to process information, an analog circuit designed to process information, a state machine, and/or other mechanisms for electronically processing information). The one or more processing devices may include one or more devices executing some or all of the operations of method 600 in response to instructions stored electronically on an electronic storage medium. The one or more processing devices may include one or more devices configured through hardware, firmware, and/or software to be specifically designed for execution of one or more of the operations of method 600.
At an operation 602, output signals may be generated. In some embodiments, the output signals may convey spatial position information related to spatial positions of one or more anatomical sites on the cancer patient while the cancer patient performs a prescribed movement. The spatial position information may comprise visual information representing the body of the cancer patient and/or other information. The one or more anatomical sites may comprise an anatomical site that corresponds to a center of mass of the cancer patient. In some embodiments, the one or more anatomical sites may comprise anatomical sites indicative of mobility and/or the center of mass of a cancer patient, and/or other anatomical sites. In some embodiments, a location that corresponds to the center of mass and/or that is indicative of mobility may be a location at a base of a spine of the cancer patient, a location at or near the hips of a cancer patient, locations and/or near the knees of a cancer patient, and/or other locations. The prescribed movement may comprise movement associated with a chair to table (CTT) exam and/or other movement, for example.
In some embodiments, the output signals may convey physical activity information related to physical activity performed by the cancer patient. In these embodiments, the one or more sensors may comprise a wrist worn motion sensor and/or other sensors, for example. In some embodiments, operation 602 may be performed by one or more sensors similar to or the same as body position sensor 102 and/or physical activity sensor 104 (shown in
At an operation 604, kinematic and/or physical activity parameters may be determined. In some embodiments, the one or more determined kinematic and/or physical activity parameters may be features extracted from the spatial position or physical activity information, and/or other parameters. In some embodiments, the determined kinematic and/or physical activity parameters may comprise less bytes of data than the spatial position information and/or the physical activity information conveyed by the one or more output signals. In some embodiments, operation 604 may include determining one or more kinematic parameters indicative of the movement of the cancer patient during the prescribed movement based on the spatial position information and/or other information. The one or more kinematic parameters may comprise velocities, accelerations, and/or other kinematic parameters. For example, the one or more kinematic parameters may comprise an acceleration of an anatomical site that corresponds to the center of mass of the cancer patient, a velocity and/or acceleration of an anatomical site indicative of mobility of the cancer patient, and/or other parameters. In some embodiments, determining the one or more kinematic parameters indicative of the movement of the cancer patient during the prescribed movement based on the spatial position information comprises determining anatomical site position vectors for the one or more anatomical sites. The anatomical site position vectors may comprise three-dimensional time series generated for given positions of the one or more anatomical sites at given time points during the prescribed movement. This may also include determining accelerations for the one or more anatomical sites based on the anatomical site position vectors using a mean-value theorem. The acceleration of an anatomical site that corresponds to the center of mass (for example) of the cancer patient may be determined using the mean-value theorem based on anatomical site position vectors for the anatomical site that corresponds to the center of mass of the cancer patient, for example.
In some embodiments, operation 604 may include determining one or more physical activity parameters indicative of the physical activity of the cancer patient based on the physical activity information and/or other information. In these embodiments, the one or more physical activity parameters may comprise metabolic equivalence (METs) and/or other parameters. In some embodiments, operation 604 may be performed by one or more processors configured to execute a computer program component similar to or the same as parameter component 112 (shown in
Operation 606 may include determining whether a patient will need unplanned medical care. In some embodiments, the determining may be based on an acceleration of an anatomical site that corresponds to the center of mass of the cancer patient, velocities and/or accelerations of anatomical sites indicative of mobility, and/or other information. In some embodiments, the determining may be based on the metabolic equivalence determined for the cancer patient, and/or other information.
In some embodiments, the determination of whether the cancer patient will need unplanned medical care during cancer therapy is indicative of a future reaction of the cancer patient to chemotherapy and/or radiation during cancer therapy. In some embodiments, determining whether the cancer patient will need unplanned medical care during cancer therapy comprises determining whether the cancer patient will need unplanned medical care during a future period of time that corresponds to one or more cancer therapy treatments received by the cancer patient. In some embodiments, the future period of time is about two months and/or other periods of time. In some embodiments, operation 606 comprises categorizing the cancer patient as either likely to likely to need unplanned medical care or unlikely to need unplanned medical care during cancer therapy. In some embodiments, operation 606 comprises determining a likelihood the cancer patient will need unplanned medical care, and categorizing the cancer patient into two or more groups based on the likelihood. In some embodiments, operation 606 may be performed by one or more processors configured to execute a computer program component similar to or the same as determination component 113 (shown in
At an operation 608, therapy may be adjusted. The adjusted therapy may be the cancer therapy and/or other therapies. The adjusting may be based on the determination of whether the patient will need unplanned medical care and/or other information. In some embodiments, adjusting may include facilitating adjustment of the cancer therapy based on the determination of whether the cancer patient will need unplanned medical care during cancer therapy. In some embodiments, facilitating may comprise determining and displaying recommended changes, determining one or more additional parameters from the information in the output signals from the one or more sensors, and/or other operations. In some embodiments, operation 608 may be performed by one or more processors configured to execute a computer program component similar to or the same as determination component 113 (shown in
Although the present technology has been described in detail for the purpose of illustration based on what is currently considered to be the most practical and preferred embodiments, it is to be understood that such detail is solely for that purpose and that the technology is not limited to the disclosed embodiments, but, on the contrary, is intended to cover modifications and equivalent arrangements that are within the spirit and scope of the appended claims. For example, it is to be understood that the present technology contemplates that, to the extent possible, one or more features of any embodiment can be combined with one or more features of any other embodiment.
This patent application is a national phase filing of, and claims the benefit of, International Patent Application No. PCT/US2019/067950, filed on Dec. 20, 2019, entitled “SYSTEM AND METHOD FOR DETERMINING HUMAN PERFORMANCE”, naming Peter Kuhn and Jorge Nieva as inventors, and designated by attorney docket no. 043871-0508992, which claims the benefit of Provisional Patent Application No. 62/783,921 filed on Dec. 21, 2018, entitled “SYSTEM AND METHOD FOR DETERMINING HUMAN PERFORMANCE”, naming Peter Kuhn and Jorge Nieva as inventors, and designated by attorney docket no. 043871-0501304. The entire content of the foregoing patent application is incorporated herein by reference, including all text, tables and drawings.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/US2019/067950 | 12/20/2019 | WO | 00 |
Number | Date | Country | |
---|---|---|---|
62783921 | Dec 2018 | US |