Cognitive test workflow adjustment

Abstract

A method for adjusting a cognitive test routine as a function of detection of athletic impairment of a patient during execution of the cognitive test routine. In particular, the workflow for administering a neuropsychological assessment including one or more cognitive tests is adjusted according to the detection of athletic impairment.

Description

Technical Field

[0001] The present invention relates to the field of cognitive testing, and in particular to the configuration of a workflow for a battery of cognitive tests. Background Art

[0002] Many older adults may experience motor impairments, such as hand tremors, bradykinesia, or speech tremors. The prevalence of motor impairments is high in Parkinson's disease (PD), essential tremor (ET), and multiple sclerosis. Sometimes, motor impairments are a symptom of an underlying neurodegenerative disease (e.g., PD), but in other cases, motor impairments may be benign and may be caused by medications or other causes. Since patients with cognitive impairment and / or neurodegenerative diseases are the main target group for neuropsychological assessments (e.g., conducted on Philips' Intelligent Spatial Cognition (ISC) platform), a considerable number of subjects of neuropsychological assessments will have some form of motor impairment. Especially in the early stages, motor impairments may be subtle, so that patients fail to report motor impairments and healthcare providers are therefore unaware that the patient has a motor impairment.

[0003] The most common motor injury is tremor. This is an involuntary, rhythmic muscle contraction that causes shaking movements in one or more parts of the body. Tremors often affect the hands, but can also affect the arms, head, vocal cords, trunk, and legs. Tremors may be intermittent or constant. Tremors that affect the hands or arms can make it difficult to hold a pen and write or draw. Vocal tremor can cause speech to tremble and become more difficult to understand. Summary of the invention

[0004] Performance on cognitive tests that include a motor component may be impaired due to motor impairment, for example, because the patient may take longer to complete the test, or because the patient may make more errors during the test. For example, all tests performed on the ISC platform have a motor component and are therefore affected by motor impairment. Tests involving drawing or using a pen may be affected by hand tremors or bradykinesia. For example, these tests include the clock drawing test (CDT), the line connection test (TMT), and the Rey Osterrieth complex figure test (ROCFT). Tests involving speaking and that may be affected by language tremors include, for example, the Rey Auditory Verbal Learning Test (RAVLT) and the Controlled Oral Word Association Test (COWAT). Motor impairments can reduce performance on these tests due to motor impairments rather than cognitive impairments.

[0005] Thus, the inventors have recognized that systems and methods that can assist in automatically determining whether a particular result of a cognitive test is affected by a motor impairment would be beneficial.

[0006] The invention is defined by the claims.

[0007] One aspect of the present invention is a computer-implemented method comprising:

[0008] controlling the testing device to administer a neuropsychological assessment including a motor component to the subject, wherein the neuropsychological assessment includes: administering a plurality of cognitive tests according to an ordered schedule defined by a neuropsychological assessment workflow, and processing individual test results according to processing operations defined by the neuropsychological assessment workflow;

[0009] automatically detecting one or more motor impairments during performance of the neuropsychological assessment; and

[0010] generating a neuropsychological assessment workflow change recommendation in response to said detecting; and

[0011] An output is generated indicating the proposed workflow change.

[0012] It is an object according to embodiments of the present invention to provide a tool that can help differentiate motor impairment detection results as symptoms of neurodegenerative diseases (which are different from cognitive impairments). In the proposed method, this is based on detecting motor impairments during the performance of cognitive tests and adjusting the test workflow to compensate. This adjustment can be the way the cognitive test score is calculated and / or can be the way the test is performed.

[0013] The neuropsychological assessment may include a battery of cognitive tests, such as the Intelligent Spatial Cognition (ISC) battery.

[0014] In some embodiments, the method may further include implementing the workflow change suggestion by changing the workflow according to the suggestion.

[0015] In some embodiments, the processing operation includes combining scores obtained for a plurality of individual cognitive tests of the neuropsychological assessment into one or more cognitive domain scores according to a pre-associated cognitive domain category for each cognitive test.

[0016] In some embodiments, the cognitive domain scores include scores for one or more of: memory, working memory, processing speed, verbal processing, executive function, visuospatial processing.

[0017] The workflow adjustment recommendations may include recommendations for corrections to one or more cognitive domain scores.

[0018] This is to ensure that individual test scores as well as domain test scores are not unduly affected by athletic impairments.

[0019] In some embodiments, the processing operations include combining a plurality of cognitive domain scores into an overall cognitive score, and wherein the workflow adjustment recommendation includes a recommendation for correction to the overall cognitive score.

[0020] The workflow adjustments can be made at different points in time relative to the test being performed.

[0021] In some embodiments, in response to detecting a motor impairment during administration of a particular cognitive test, a user-perceivable output indicative of the workflow adjustment recommendation is generated during the neuropsychological assessment, and wherein the workflow change recommendation includes discontinuing a current cognitive test in which a motor impairment has been detected.

[0022] In some embodiments, in response to detecting a motor impairment during the neuropsychological assessment, a user-perceivable output indicative of the recommendation is generated during the assessment, and wherein the workflow change recommendation includes skipping a subsequent test scheduled according to the workflow that involves the same motor component as the cognitive test in which the motor impairment has been detected.

[0023] In some embodiments, in response to detecting a motor impairment during the neuropsychological assessment, an output indicative of the recommendation is generated after all tests scheduled according to the workflow have been administered, and wherein the workflow change recommendation includes adjusting a score combination portion of the processing operation to compensate for the scores obtained for any individual cognitive tests in which the motor impairment has been detected.

[0024] In some embodiments, scores for any test in which a motor impairment has been detected are removed from the score combination calculation.

[0025] In some embodiments, the workflow change suggestion may be generated only in response to a determined severity of the detected motion impairment exceeding a predefined threshold, and not generated otherwise.

[0026] In some embodiments, the workflow further comprises generating a clinical report indicating results of the neuropsychological assessment, and wherein the workflow change recommendation comprises adding an indication of any detected motor impairment to the report.

[0027] In some embodiments, detecting a motor impairment comprises detecting the presence of any one or more of: a motor tremor of a limb, a speech tremor, bradykinesia. These examples are certainly not exhaustive.

[0028] In some embodiments, the method further comprises determining a quantification of the severity of the detected motion impairment.

[0029] In some embodiments, administering at least one of the cognitive tests comprises acquiring a test data set comprising one or more data variables associated with performance of the test.

[0030] There are different methods to detect the one or more sports injuries.

[0031] One way is through spectral analysis.

[0032] In this regard, in some embodiments, detecting the one or more motion impairments comprises processing at least a subset of the test data set using spectral analysis to derive a spectral power spectrum, and wherein detecting at least one motion impairment is performed based on analyzing the spectral power spectrum. The spectral analysis may comprise Fourier analysis, and wherein the spectral power spectrum is a Fourier power spectrum. As another example, the spectral analysis may comprise wavelet analysis.

[0033] In some embodiments, the motion impairment to be detected comprises a motor tremor.The aforementioned one or more data variables may comprise a data variable as a function of time indicative of motion of a portion of the subject's body.

[0034] In some embodiments, the motor impairment to be detected includes a voice tremor.The aforementioned one or more data variables may include a data variable as a function of time indicative of speech acoustic data of the subject's spoken input.

[0035] In some embodiments, the motion impairment is determined based on detecting a deviation between a spectral power spectrum derived for the subject and a pre-stored normal spectral power spectrum for a healthy control group.

[0036] In some embodiments, the severity of the tremor is quantified based on the derived spectral power spectrum.

[0037] In some embodiments, wherein administering at least one of the cognitive tests comprises acquiring a test data set comprising a plurality of data variables associated with performance of the test, detecting the one or more motor impairments comprises applying a Bayesian model to the plurality of data variables collected for the test, wherein the Bayesian model is configured to output a probability score indicating a likelihood that a predetermined motor impairment is present. This provides multimodal detection of motor impairments.

[0038] In some embodiments, one or more Bayesian models are applied to generate corresponding probability scores for multiple different sports injuries, and wherein the multiple probability scores are combined into a combined probability score, and wherein the workflow change recommendation is generated only in response to the combined probability score exceeding a predefined threshold.

[0039] The combining may include forming a weighted combination.

[0040] Optionally, weighting factors can be estimated automatically based on past performance and the variability of current performance.

[0041] In some embodiments, detecting the motor impairment comprises distinctively identifying a type of motor impairment that has been detected, and wherein the workflow change suggestion is configured based on the type of cognitive impairment(s) detected.

[0042] In some embodiments, the method further comprises the step of determining a risk prediction for cognitive decline based on the neuropsychological assessment.

[0043] In some embodiments, administering at least one of the cognitive tests includes acquiring a test data set comprising one or more data variables associated with performance of the test, and the method further includes a calibration phase comprising acquiring a comparator data set comprising the same data variables collected during the at least one cognitive test, but the calibration phase does not include administering the at least one cognitive test.

[0044] In some embodiments, the comparator data set is analyzed to detect the presence of the same motor impairment as has been detected during administration of the cognitive test, and to quantify the severity of such motor impairment.

[0045] The concept of the present invention can also be embodied in the form of software.

[0046] Therefore, another aspect of the present invention is a computer program product comprising code means configured to, when run by a processor, cause the processor to perform a method according to any example or embodiment outlined in this document or according to any claim of this application.

[0047] The present invention can also be embodied in the form of hardware.

[0048] Accordingly, another aspect of the present invention is a system.

[0049] The system may include an input / output unit, and one or more processors adapted to:

[0050] controlling the testing device via control signals transmitted through the input / output to administer a neuropsychological assessment including a motor component to the subject, wherein the neuropsychological assessment includes administering a plurality of cognitive tests according to an ordered schedule defined by a neuropsychological assessment workflow, and processing individual test results according to processing operations defined by the neuropsychological assessment workflow;

[0051] automatically detecting one or more motor impairments during performance of said neuropsychological assessment;

[0052] generating a neuropsychological assessment workflow change recommendation in response to said detecting; and

[0053] An output is generated indicating the proposed workflow change.

[0054] The system may also include a memory storing a record of a standard neuropsychological workflow defining an ordered schedule of cognitive tests to be administered to a subject and defining processing operations for processing individual test results.

[0055] The system may further include the testing device for administering the neuropsychological assessment to a subject.

[0056] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] For a better understanding of the invention and to show more clearly how it may be put into practice, reference will now be made, by way of example only, to the accompanying drawings, in which:

[0058] Figure 1 The contribution of different cognitive tests to different cognitive domains is shown;

[0059] Figure 2 The steps of a method according to one or more embodiments are summarized;

[0060] Figure 3 is a block diagram of components of an example system according to one or more embodiments of the present invention;

[0061] Figure 4 schematically illustrates a process of analyzing a cognitive test data set to detect the presence and / or severity of a motor impairment according to a first example;

[0062] Figure 5 schematically illustrates a process of analyzing a cognitive test data set to detect the presence and / or severity of a motor impairment according to a second example; and

[0063] Figure 6 A schematic overview of the steps of an example multimodal method for detecting and analyzing multiple motor impairments and generating a composite impairment score based thereon is provided. DETAILED DESCRIPTION

[0064] The present invention will be described with reference to the accompanying drawings.

[0065] It should be understood that the detailed description and specific examples, although indicating exemplary embodiments of the device, system and method, are only for illustrative purposes and are not intended to limit the scope of the invention. These and other features, aspects and advantages of the device, system and method of the present invention will be better understood from the following description, the appended claims and the accompanying drawings. It should be understood that these drawings are merely schematic and are not drawn to scale. It should also be understood that the same reference numerals are used throughout the drawings to indicate the same or similar parts.

[0066] The present invention provides a method for adjusting a cognitive testing routine based on the detection of a motor impairment of a patient during the performance of the cognitive test. In particular, a workflow for administering a neuropsychological assessment comprising one or more cognitive tests is adjusted based on the detection of a motor impairment.

[0067] Digital neuropsychological assessment platforms include different cognitive tests. For example, the ISC platform provides a suite of neuropsychological tests that assess different cognitive functions. Typically, scores from individual neuropsychological tests are combined into cognitive domain scores based on pre-associated cognitive domain categories for each cognitive test. These categories of cognitive tests may include, for example, memory, verbal processing, executive function, and / or visuospatial processing. To illustrate, Figure 1 An overview of the cognitive tests available through the already mentioned Philips Intelligent Spatial Cognition (ISC) platform and the contribution of each test to each cognitive domain in a set of exemplary cognitive domains is shown. In this figure, the degree of gray indicates the size of the contribution to the score of the relevant cognitive domain, where the lightest gray indicates no contribution, and darker gray indicates a smaller or larger (proportional to the degree of gray) contribution of a particular test to a cognitive domain. Figure 1 This is an illustration from an ISC product brochure.

[0068] In some cases, the cognitive domain scores may then be merged or combined into an overall cognitive score.

[0069] The reader is also referred to the article “Digital cognitive assessment in clinical practice utilizing Philips IntelliSpace Cognition: early clinical and operational experience from the Dent Neurologic Institute” by Pinter, NK, Bottoms, A., Mann, C., Ajtai, B., Pasternak, J., Harlock, S., and Fritz, JV (n.d.), which provides a detailed discussion of the ISC test and its use in clinical practice.

[0070] It can be seen that in embodiments where such a composite score is calculated, performance on individual cognitive tests will therefore affect both the cognitive domain scores and the overall cognitive score.

[0071] Therefore, if the results for one or more cognitive tests are actually invalid for a particular patient due to the presence of a motor impairment (e.g., tremor) that affects the accuracy of the results, the scores on these cognitive tests will improperly affect the cognitive domain scores and the overall cognitive scores, and thus lead to an inaccurate assessment of cognitive function. In the case where the neuropsychological assessment is a digital assessment and is performed in a remote or unsupervised manner, special attention may be paid to this aspect. Therefore, it is important to understand whether motor impairment (e.g., tremor) affects cognitive test performance. According to an embodiment of the present invention, this information can then be used to take remedial measures by adjusting the workflow of the neuropsychological assessment (e.g., adjusting individual cognitive test scores, discounting scores, removing specific tests from domains and / or overall cognitive scores, stopping the execution of tests, or various other workflow changes) to adapt to motor impairment.

[0072] According to the application, the inventors propose a method that automatically detects motor impairments during the execution of a battery of digital cognitive tests (e.g., on the ISC platform already mentioned) and gives workflow suggestions. If a motor impairment is detected, a workflow change suggestion is given. Through the examples and discussions presented below, it will become clear that the workflow change suggestion can be understood as a change in the original or standard way of administering the test and processing individual test results and scores on a specific digital neuropsychological assessment (e.g., the ISC platform).

[0073] Figure 2 The steps of an example method according to one or more embodiments are summarized in block diagram form. These steps will be summarized before further explanation in the form of an example embodiment.

[0074] The method 10 includes controlling a testing device to administer 12 to a subject a neuropsychological assessment including a motor component.

[0075] The neuropsychological assessment includes administering a plurality of cognitive tests according to an ordered schedule defined by the neuropsychological assessment workflow. The neuropsychological assessment may also include processing individual cognitive test results according to processing operations defined by the neuropsychological assessment workflow.

[0076] The method 10 also includes automatically detecting 14 one or more motor impairments during the neuropsychological assessment. By way of example, the motor impairments that may be detected may include any one or more of: limb motor tremor, speech tremor, and bradykinesia.

[0077] The method 10 also includes generating 16 a neuropsychological assessment workflow change recommendation in response to the detecting.

[0078] The method 10 also includes generating 18 an output indicative of a workflow change suggestion.

[0079] The output 18 may be a user perceptible output, such that the end result of the method is merely a recommendation communicated to a user, such as a patient taking the test or a clinician supervising administration of the test.

[0080] The output may be a data output. The data output may be transmitted to one or more further locations or addresses.

[0081] The method may also include implementing the workflow change suggestion by changing the workflow according to the suggestion.

[0082] As mentioned above, the method can also be embodied in hardware form.

[0083] To illustrate, Figure 3 A schematic representation of an example system 30 is presented in accordance with one or more embodiments of the present invention.

[0084] The illustrated system includes a processing unit 32 including an input / output 34 and one or more processors 36 .

[0085] The one or more processors 36 are configured to perform the method according to any example or embodiment described in this document or according to any claim of this application. More specifically, the one or more processors 36 may be adapted to perform the following steps:

[0086] controlling the testing device 42 via control signals transmitted through the input / output 34 to administer to the subject a neuropsychological assessment that includes a motor component, and wherein the neuropsychological assessment includes administering a plurality of cognitive tests according to an ordered schedule defined by a neuropsychological assessment workflow, and processing individual test results according to processing operations defined by the neuropsychological assessment workflow;

[0087] automatically detecting one or more motor impairments during the performance of a neuropsychological assessment;

[0088] generating a neuropsychological assessment workflow change recommendation in response to said detecting; and

[0089] Generates output indicating proposed changes to the workflow.

[0090] One aspect of the present invention is a system that includes only the processing unit 32 (or in other words, includes only the input / output 34 and one or more processors 36).

[0091] Optionally, the system 30 can also include a memory 38 storing a record of a standard neuropsychological workflow that defines an ordered schedule of cognitive tests to be administered to a subject and defines processing operations for processing individual test results.

[0092] Optionally, the system may include a testing device 42 for administering a neuropsychological assessment to a subject. For example, this may be a computer-based (i.e., comprising a computing device) testing device. It may be a dedicated computing console or terminal. It may include a mobile computing device (e.g., a tablet, a smartphone, or a laptop). It may include a user interface. The user interface may include one or more sensory output devices for generating user-perceivable stimuli. The user interface may include a user input device. This may include one or more indicator devices (e.g., an electronic pen or pencil) and / or a touch screen and / or a gesture capture device (e.g., a handheld unit in which an inertial measurement unit is integrated or a camera / sensor-based gesture capture device).

[0093] In any embodiment, the neuropsychological assessment may be a digital neuropsychological assessment.A digital neuropsychological assessment includes a neuropsychological assessment (ie, a test administered electronically) that is administered by controlling a user interface device to generate prompt instructions and measure test response data.

[0094] As described above, in a typical neuropsychological assessment workflow, scores for individual neuropsychological tests are combined into cognitive domain (e.g., memory, executive function) scores. Cognitive domain scores can also be merged or combined into an overall cognitive score.

[0095] Thus, in some embodiments of the invention, the processing operation portion of the workflow includes combining scores obtained from multiple individual cognitive tests for a neuropsychological assessment into one or more cognitive domain scores according to the pre-associated cognitive domain category of each cognitive test. In a standard workflow, this combining operation may be defined by a combining function that specifies the weighting of each individual cognitive test score to be attached to the combination.

[0096] For example, cognitive domain scores may include any scores for any one or more of the following: memory, verbal processing, executive functioning, visuospatial processing.

[0097] In some embodiments, the workflow adjustment suggestion may include a suggestion for correction of one or more cognitive domain scores. Additionally or alternatively, it may include changing the weighting of one or more individual cognitive test scores that are attached to the combination forming the cognitive domain score. Additionally or alternatively, it may include excluding one or more cognitive test scores from the cognitive domain score calculation.

[0098] Furthermore, according to some embodiments, the processing operation portion of the workflow may include combining the plurality of cognitive domain scores into an overall cognitive score.Here, in some embodiments, the workflow adjustment suggestion includes a suggestion for correction of the overall cognitive score.

[0099] Different options exist regarding when in the workflow of a neuropsychological assessment a recommendation for a workflow change should be given and / or implemented.

[0100] In some embodiments, recommendations can be made during a neuropsychological assessment, such as to discontinue testing where motor impairment has been detected.

[0101] For example, in some embodiments, in response to detecting a motor impairment during administration of a particular cognitive test, a user-perceivable output is generated during a neuropsychological assessment indicating a workflow adjustment suggestion, and wherein the workflow change suggestion includes discontinuing a current cognitive test in which the motor impairment has been detected.

[0102] As another example, a workflow change recommendation may be made during an assessment, such as skipping subsequent tests involving the same motor component (ie, drawing, writing, or language) from a neuropsychological assessment.

[0103] For example, in some embodiments, in response to detecting a motor impairment during a neuropsychological assessment, a user-perceivable output indicative of a recommendation is generated during the assessment, and wherein the workflow change recommendation includes skipping a subsequent test scheduled according to the workflow that involves the same motor component as the cognitive test for which the motor impairment has been detected.

[0104] According to either of the above two examples, the user-perceivable output may be, for example, a visual output or an auditory output.

[0105] In some embodiments, workflow changes may be recommended following the assessment to generally remove tests for which motor impairment has been detected from the calculation of the relevant cognitive domain test scores and / or the overall cognitive score.

[0106] For example, in some embodiments, in response to detecting a motor impairment during a neuropsychological assessment, an output indicative of a recommendation is generated after all tests scheduled according to the workflow have been administered, and wherein the workflow change recommendation includes adjusting a score combination portion of the processing operation to compensate for the scores obtained for any individual cognitive tests for which a motor impairment has been detected. For example, the scores for any tests for which a motor impairment has been detected may be removed from the score combination calculation.

[0107] In any of the above cases, one option is to generate a workflow change suggestion only in response to a determined severity of the detected motion impairment exceeding a predefined threshold, and not otherwise. The option of how the severity can be determined is discussed below.

[0108] In some embodiments, the workflow may also include generating a clinical report indicating the results of the neuropsychological assessment.

[0109] In some embodiments, the workflow suggestion may include a warning or statement of possible tremor added to the clinical report.Such a workflow change can shorten the assessment time and ensure that the results of the overall cognitive assessment - or specific components - are reliable and valid.

[0110] Regarding the detection of motor impairment, different options are possible. Generally speaking, it is based on processing one or more data variables included in a test data set acquired as part of a cognitive test using a dedicated detection algorithm that detects the presence or absence of a motor impairment or the presence of at least one specific motor impairment and preferably determines a measure of its severity.

[0111] One advantageous approach is to use spectral analysis (eg, Fourier analysis or wavelet analysis) of the cognitive test data set.

[0112] In this method, one or more motor impairments are detected based on processing at least a subset of an acquired cognitive test data set using spectral analysis to derive a frequency domain power spectrum (e.g., a Fourier power spectrum), and wherein the detection of at least one motor impairment is performed based on the analysis of the spectral power spectrum. The cognitive test data set is a data set acquired as part of a normal administration of at least one of the cognitive tests, and it includes one or more data variables associated with performance of the test.

[0113] As a general principle, a given motion impairment can be detected and identified in this manner based on detecting deviations between a spectral power spectrum derived for a subject and a pre-stored normal spectral power spectrum for a healthy control group. Based on applying different detection criteria to any detected deviations and / or detecting or extracting the deviations in different ways, a dedicated comparison algorithm can be provided to detect each of one or more different motion impairments differently.

[0114] In some embodiments, the severity of the motion impairment can be quantified based on the derived spectral power spectrum.

[0115] To illustrate this concept, we will now refer to Figure 4 and Figure 5 It will be appreciated that all features of each of these specific examples are not essential to the concept of the present invention, and each feature is described to aid understanding and provide examples to illustrate the concept of the present invention.

[0116] Reference Figure 4 The first example described relates to a situation where the motor impairment to be detected includes motor tremor. The cognitive test being administered requires the user to perform a drawing task by moving a pencil pointer device on an input unit of the computing device. Figure 4 In the example shown, the input device is a touch screen display 52. ​​In the illustrated example, the computing device is a tablet computer.

[0117] A test data set is acquired, the test data set comprising data variables corresponding to a motion path of a pencil pointer device on the screen 52 .

[0118] Thus, this is an example of a more general case in which administering at least one cognitive test of a battery of cognitive tests comprises acquiring a test data set comprising data variables as a function of time indicative of movement of a part of a subject's body.

[0119] Then, according to this set of embodiments, a quantification of motor tremor is determined by automatically analyzing the tactile input provided by the pencil via the tablet. Figure 4Such an analysis is schematically illustrated in In this illustrated example, the test corresponds to the wired test (TMT) of the ISC platform described above.

[0120] The process involves comparing the Fourier spectrum derived from the captured tactile input (row b) with a baseline spectrum for a healthy population (row a).

[0121] The drawing trajectory captured by the input device 52 is processed to extract the x, y and z position variables 62b as a function of time (i.e., an x, y, z time series). This may involve a normalization process (e.g., via an instruction / test session or while the test subject is writing their name) using calibration data that has been acquired earlier in the workflow. Such normalization may be done in real time as the input data is acquired (i.e., as the user draws on the screen 52), or may be done in batch form at a later stage.

[0122] The x, y and z position variable data are processed to obtain time x, y and z first order time derivative data 64b. A Fourier transform is then applied to the x, y, z first order derivative signals 64b within a specified time window and a Fourier power spectrum 66a is derived therefrom. Alternatively, a Fourier transform may be applied directly to the position variable signal 62a.

[0123] Performing the analysis includes comparing with or referencing the Fourier spectrum 66a for healthy controls. To further aid understanding and interpretation, Figure 4 The intermediate steps of deriving position trajectory data 62a, first order derivative data 64a and Fourier spectrum 66a for a healthy control population are schematically illustrated. However, in practice, it is expected that these steps will be performed prior to the method of the present invention, wherein the Fourier spectrum is pre-stored, ready to be compared with or referenced to the newly captured data spectrum 66b. However, it is of course also possible to perform the Fourier analysis of the control data in real time.

[0124] Within the equivalent time window, the deviation of the Fourier power spectrum 66b derived from the test data set from the normal power spectrum 66a is determined. Thus, the frequency peaks corresponding to the tremor in the test spectrum 66b can be estimated. The tremor can then be quantified using the frequency band of each tremor (which can be detected), the amplitude of each tremor (e.g., detected based on the height of each peak), and the number of tremors (e.g., detected based on the number of peaks).

[0125] In the illustrated example, a clear peak can be detected in the test Fourier spectrum 66b. Quantification of this peak relative to the normal spectrum 66a (eg, spectral power or area under the curve) can be used to detect, for example, hand tremor.

[0126] Reference Figure 5 The second example described relates to the case where the motor impairment to be detected includes voice tremor. The detection and preferably quantification of voice tremor is determined by automatically analyzing the speech input provided via a microphone of the tablet computer 52, for example.

[0127] Thus, this is an example of a more general case where administering at least one cognitive test in a battery of cognitive tests includes acquiring a test data set comprising data variables as a function of time indicative of speech acoustic data of spoken input of a subject. For example, speech-based cognitive tests in the ISC platform include the Rey Auditory Verbal Learning Test (RAVLT) and the Controlled Oral Word Association Test (COWAT).

[0128] Row (a) corresponds to data for healthy controls. Row (b) corresponds to data acquired during cognitive testing for a test subject with voice tremor. Acoustic amplitude data 72b is acquired during the test and spectrally analyzed in the form of wavelet analysis within a defined time window to produce a power spectrum 74b in the frequency domain. A Fourier analysis may be performed instead. Peaks in the power spectrum can then be derived. Based on deviations from a normal power spectrum 74a derived from acoustic data 72a for a normal control population, motor impairments can be detected and quantified. For example, a number of detected deviations (e.g., the number of peaks present in the test data spectrum that are not present in the normal spectrum) can be used to quantify voice tremor. Additionally or alternatively, the area under the curve of abnormal peaks detected in the test spectrum 74b can be used to quantify the tremor and / or the height of such peaks.

[0129] If, rather than using a single data variable, at least one of a plurality of cognitive tests is administered (which includes acquiring a test data set that includes a plurality of data variables associated with performance on the test, and including the plurality of data variables of the test data set in the analysis), more accurate or more discriminating detection of one or more motor impairments is possible. This is referred to in the art as multimodal assessment.

[0130] For example, see above Figure 4 and Figure 5 In the examples discussed in, additionally, information regarding the speed and speed changes of pen movement, the angle of the pen on the tablet and changes in the pen angle, the number and speed of direction changes, and the pressure and pressure changes during drawing and / or writing can be monitored and acquired as part of the test data set. These data variables can then be used for multimodal detection of motor impairments affecting the hand and / or arm.

[0131] Multimodal assessment can be specific to a given cognitive test and can take into account the characteristics of a particular cognitive test. For example, for a trail-tracking test (TMT), velocity variations can be expected due to the visual distribution of the targets. Velocity decreases when approaching the target and increases when moving away from it. Velocity can also be affected by the distance between two consecutive targets. One solution is to define a perimeter around the target and exclude velocity variations within that perimeter from the analysis. These metrics—velocity, angle, change of direction, pressure—can be used together for multimodal detection of tremor.

[0132] Based on a plurality of measured test variables or parameters derived therefrom (e.g. from a spectral transform), it is possible to determine a motor impairment "score" P(T|C) representing the probability that a specific motor impairment (e.g., tremor) will occur given both prior knowledge (on the one hand) and measured parameters of the test variables or a spectral transform of the tremor during the task at hand (on the other hand).

[0133] One way to combine likelihood and prior knowledge is based on Bayesian probability theory.

[0134] For example, as a general principle, where administering at least one of the cognitive tests includes acquiring a test data set including a plurality of data variables, in some embodiments, detecting one or more motor impairments may include applying a Bayesian model to the plurality of data variables or parameters derived therefrom (e.g., parameters of a spectral transform). As mentioned above, the Bayesian model may be configured to output a probability score P(T|C) indicating the likelihood that a predetermined motor impairment is present.

[0135] Further details will now be set out by describing an example embodiment in which the motion impairment to be detected is tremor. However, it will be appreciated that the principles of the method can be more broadly applied to the detection of any motion impairment.

[0136] In the Bayesian model, the general inference is first decomposed into a "likelihood" probability distribution. This represents the probability of one or more cues (C) given the sensorimotor impairment data (M) (referred to as P(C|M)). Secondly, it is further decomposed into the so-called "prior", which represents the a priori expected probability P(M) related to the sensorimotor impairment data. The likelihood and prior are combined to produce the "posterior" (posterior) probability P(M|C), which represents the probability of the presence of a specific parameter in the sensorimotor impairment data (M) given the cue (C), and this is obtained from the Bayesian rule with the general form: P(T|C)∝P(C|T)P(T). If the score (probability) exceeds a predetermined threshold (e.g., a population average), a mark is generated in the log file. In addition, the (prior) condition that exceeds the threshold is also recorded in the log file.

[0137] The process of constructing and implementing a suitable Bayesian model is well known to those skilled in the art. For example, see the article in the Internet Encyclopedia (https: / / en.wikipedia.org / wiki / Bayes_theorem). Any textbook on Bayesian theory can serve as further reference.

[0138] In some embodiments, a composite multimodal tremor "score" can be determined to quantify the effects of a combination of multiple impairments (eg, hand tremor, vocal cord tremor).

[0139] In this regard, as a general principle, in some embodiments, one or more Bayesian models are applied to generate corresponding probability scores for a variety of different sports injuries, and wherein the multiple probability scores are combined into a combined probability score. In some embodiments, the quantitative results of such scores can be used to determine whether a workflow change is needed and what kind of workflow change is given. In some embodiments, a workflow change suggestion can be generated only in response to the combined probability score exceeding a predefined threshold.

[0140] For example, a patient may exhibit multiple tremors, each with a specific probability and prerequisites for the occurrence of the tremor. It would be valuable to combine the effects of multiple tremors, since multiple types of tremors can interfere with cognitive function in a cumulative manner. Multiple individual tremor scores can be combined using weighting factors. The weighting factors can be automatically estimated, for example, based on past presentations and variability of current presentations (e.g., variability of the above parameters from the spectral power spectrum). For example, a composite tremor score can be calculated from the scores for individual tremors as follows:

[0141] Weight 1 × Tremor 1 + Weight 2 × Tremor 2 = Composite Tremor

[0142] To summarize this concept, Figure 6 An example process flow for an exemplary Bayesian approach to motion impairment detection is outlined.

[0143] The process includes, in a first stage 82, determining a corresponding motion impairment probability score for each of a plurality of different motion impairments, for example in the manner discussed above.

[0144] The method also includes a multimodal processing operation in a second stage 84, wherein the plurality of individual probability scores obtained in stage 82 are combined with corresponding weights to form a composite motion impairment score (e.g., Figure 6 Composite tremor score in the example).

[0145] The method also includes a decision step in a third stage 86, wherein the algorithm evaluates the composite score with reference to one or more predefined criteria to determine whether a motor impairment exists. Figure 6 In the case shown, the composite score is compared to a predetermined threshold.

[0146] The method also includes generating a recommendation based on the decision step of stage 86 in stage 88. This can be a user-perceivable recommendation and can be communicated to the patient and / or the neurologist administering the neuropsychological assessment. Different recommendation items (e.g., Figure 6 (1), (2) and (3)) listed in the example are transmitted to different patients and neurologists, such as issuing a suggestion to the patient to stop completing the test, and also additionally issuing a suggestion to update the test result calculation, or transmitting an explanation to the neurologist.

[0147] In some embodiments, by using analysis according to the present invention, different types of motor impairments can be distinguished from each other, for example, tremors can be distinguished from bradykinesia, and head tremors can be distinguished from laryngeal (vocal cord) tremors. This will be valuable because some types of tremors are known to interfere more with cognitive function.

[0148] Thus, as a general principle, it can be said that in some embodiments, detecting a motor impairment comprises distinctively identifying a type of motor impairment that has been detected, and wherein workflow change suggestions are configured according to the type(s) of cognitive impairment detected.

[0149] The detection of motor impairment can be used to inform further analysis or detection. For example, the risk of future cognitive impairment or the progression of cognitive impairment can be predicted based at least in part on the detected motor impairment and severity or other parameters (e.g., type). Therefore, in some embodiments, the method may also include the following steps: determining a risk prediction for cognitive decline based on a neuropsychological assessment.

[0150] For example, studies have shown that tremor during the onset of Parkinson's disease (PD) is an important predictor of poor cognitive performance. Regarding different types of tremors, studies have shown that compared with healthy controls, PD patients with primary ataxia and tonic tremor (not PD patients with primary tremor) have reduced activity in the default mode network, and lower gray matter volume (Karunanayaka, PR, Lee, E.-Y., Lewis, MM, Sen, S., Eslinger, PJ, Yang, QX and Huang, X's "Default mode network differences between rigidity-and tremor-predominant Parkinson's disease" (Cortex, Vol. 81, pp. 239-250, 2017). Reduced activity in the default mode network is associated with cognitive impairment in PD (ibid). Different variables (including the size and type of tremor) and cognitive impairment (corrected for motor impairment) can be used, for example, to predict the risk of long-term (further) cognitive decline in neurodegenerative diseases.

[0151] In addition to quantifying the severity of motor impairment during administration of a battery of cognitive tests, a calibration phase can also be implemented in which the parameters of the analysis are set so that potential motor impairment can be subsequently quantified (e.g., based on detected data variables (e.g., frequency, amplitude, intensity, asymmetry)).

[0152] In particular, where administration of at least one of the cognitive tests includes acquiring a test data set (including one or more data variables associated with performance of the test), in some embodiments, the method may further include a calibration phase comprising acquiring a comparator data set including the same data variables collected during the at least one cognitive test, but the calibration phase does not include administration of the at least one cognitive test.

[0153] For example, a dedicated calibration data set of data variables can be acquired during rest (i.e., when no task is being performed). Typically, in e.g. PD, tremors are more severe at rest than when performing planned movements. Once the movement plan is initiated and executed, the severity of the tremor decreases. It is therefore of interest to compare the severity of the tremor at rest (before the test is administered) with the severity of the tremor during the performance of a battery of cognitive tests.

[0154] During analysis of the cognitive test data set to detect motor impairments, the comparator data set can be utilized to detect the presence and quantify the severity of the same motor impairments as those detected during administration of the cognitive test.

[0155] As already emphasized, while the examples in this document tend to focus on tremors, other types of motor impairments (e.g., bradykinesia or rigidity) can also be detected in a similar manner and used for workflow change recommendations in a similar manner to that described above. Bradykinesia (i.e., slow movement) and rigidity (i.e., stiffness of movement and resistance to movement) are common symptoms of neurodegenerative diseases (e.g., Parkinson's disease) and can also affect performance on cognitive tests in the absence of cognitive impairment. Similar to the above, detection of other types of motor impairments can also be used to generate workflow change recommendations in the following situations: (1) workflow change recommendations are generated during assessment such that: if motor impairment exceeds a certain threshold, the test in which motor impairment has been detected is discontinued; (2) workflow change recommendations are generated during assessment such that: if motor impairment exceeds a certain threshold, subsequent tests involving the same motor component (i.e., drawing, writing, or language) are skipped from the assessment, and / or (3) workflow change recommendations are generated after assessment such that: if motor impairment exceeds a certain threshold, the test in which motor impairment has been detected is removed from the overall cognitive model, and the test in which motor impairment has been detected is removed from individual test scores and domain test scores.

[0156] In another embodiment, disease-specific norm scores can be used to correct cognitive test scores according to the determined tremor level. In this way, it is still possible to longitudinally assess cognitive function in a group of patients with tremor without completely removing the test in which tremor has been detected from the calculation of the composite score.

[0157] The above-mentioned embodiment of the present invention adopts one or more processors. These one or more processors can be located in a single containing device, structure or unit, or can be distributed between multiple different devices, structures or units. Therefore, the processing device that is suitable for or configured to perform a specific step or task can correspond to the step or task performed by any one or more processing components of a plurality of processing components alone or in combination. The technician will understand how to implement such a distributed processing device. One or more processors can be implemented in a variety of ways with software and / or hardware to perform various functions required. The processor generally adopts one or more microprocessors, which can be programmed using software (e.g., microcode) to perform the required functions. The processor can be implemented as a combination of dedicated hardware for performing certain functions and one or more programmed microprocessors and related circuits for performing other functions.

[0158] Examples of circuits that may be used in various embodiments of the present disclosure include, but are not limited to, conventional microprocessors, application specific integrated circuits (ASICs), and field programmable gate arrays (FPGAs).

[0159] In various embodiments, the processor may be associated with one or more storage media, such as volatile and non-volatile computer memory, such as RAM, PROM, EPROM, and EEPROM. The storage media may be encoded with one or more programs that perform the desired functions when executed on one or more processors and / or controllers. The various storage media may be fixed within the processor or controller, or may be removable so that one or more programs stored thereon can be loaded into the processor.

[0160] Based on a study of the drawings, the disclosure, and the appended claims, those skilled in the art will be able to understand and implement variations of the disclosed embodiments in practicing the claimed invention. In the claims, the word "comprising" does not exclude other elements or steps, and the word "a" or "an" does not exclude a plurality.

[0161] A single processor or other unit may implement the functions of several items recited in the claims. The computer program may be stored / distributed on a suitable medium, such as an optical storage medium or solid-state medium supplied with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunications systems.

[0162] The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. If the term "suitable for" is used in the claims or the description, it should be noted that the term "suitable for" is intended to be equivalent to the term "configured to".

[0163] Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A computer-implemented method (10), comprising: controlling (12) a testing device to administer a neuropsychological assessment including a motor component to a subject, wherein the neuropsychological assessment includes: administering a plurality of cognitive tests according to an ordered schedule defined by a neuropsychological assessment workflow, and processing individual test results according to processing operations defined by the neuropsychological assessment workflow; automatically detecting (14) one or more motor impairments during performance of the neuropsychological assessment; and generating (16) a neuropsychological assessment workflow change recommendation in response to said detecting; An output indicative of the proposed workflow change is generated (18).

2. The method according to claim 1, further comprising: The workflow change suggestion is implemented by changing the workflow according to the suggestion.

3. The method according to claim 1 or 2, wherein: The processing operations include combining scores obtained for a plurality of individual cognitive tests of the neuropsychological assessment into one or more cognitive domain scores according to pre-associated cognitive domain categories for each cognitive test, and optionally wherein the cognitive domain scores include scores for one or more of: memory, verbal processing, executive function, visuospatial processing, and The workflow adjustment suggestions include suggestions for correction of one or more cognitive domain scores.

4. The method according to any one of claims 1 to 3, wherein: In response to detecting a motor impairment during administration of a particular cognitive test, generating a user-perceivable output during the neuropsychological assessment indicative of the workflow adjustment recommendation, and wherein the workflow change recommendation includes discontinuing a current cognitive test in which a motor impairment has been detected.

5. The method according to any one of claims 1 to 4, wherein: In response to detecting a motor impairment during the neuropsychological assessment, generating a user-perceivable output indicative of the recommendation during the assessment, and wherein the workflow change recommendation includes skipping a subsequent test scheduled according to the workflow that involves the same motor component as the cognitive test in which the motor impairment has been detected.

6. The method according to any one of claims 1 to 5, wherein: In response to detecting a motor impairment during the neuropsychological assessment, generating an output indicative of the recommendation after all tests scheduled according to the workflow have been administered, and wherein the workflow change recommendation includes: adjusting the score combination portion of the processing operation to compensate for the scores obtained for any individual cognitive test in which the motor impairment has been detected, and optionally, wherein the scores for any test in which the motor impairment has been detected are removed from the score combination calculation.

7. The method according to any one of claims 1 to 6, wherein: The method also includes determining a quantification of the severity of the detected motion impairment.

8. The method according to any one of claims 1 to 7, in, administering at least one of the cognitive tests comprises acquiring a test data set comprising one or more data variables associated with performance of the test, and Wherein detecting the one or more motion impairments comprises processing at least a subset of the test data set using spectral analysis to derive a spectral power spectrum, and wherein detecting at least one motion impairment is performed based on analyzing the spectral power spectrum.

9. The method according to claim 8, wherein: The motion impairment to be detected comprises a motor tremor, and wherein the one or more data variables comprise a data variable as a function of time indicative of movement of a portion of the subject's body.

10. The method according to claim 8 or 9, wherein: The motor impairment to be detected comprises a voice tremor, and wherein the one or more data variables comprise time-dependent data variables indicative of speech acoustic data of spoken input of the subject.

11. The method according to any one of claims 8 to 10, wherein: The severity of the tremor is quantified based on the derived spectral power spectrum.

12. A method according to any preceding claim, in, administering at least one of the cognitive tests comprises acquiring a test data set comprising a plurality of data variables associated with performance of the test, and Wherein detecting the one or more motor impairments comprises applying a Bayesian model to the plurality of data variables collected for the testing, wherein the Bayesian model is configured to output a probability score indicating a likelihood that a predetermined motor impairment is present.

13. The method according to claim 12, wherein: One or more Bayesian models are applied to generate respective probability scores for a plurality of different motion impairments, and wherein the plurality of probability scores are combined into a combined probability score, and wherein the workflow change recommendation is generated only in response to the combined probability score exceeding a predefined threshold.

14. A computer program product comprising code means configured to, when executed by a processor, cause the processor to perform the method according to any one of claims 1-13.

15. A system (30), comprising: Input / output (34); as well as One or more processors (36) adapted to: controlling a testing device (42) via control signals transmitted through the input / output to administer a neuropsychological assessment including a motor component to a subject, wherein the neuropsychological assessment includes administering a plurality of cognitive tests according to an ordered schedule defined by a neuropsychological assessment workflow and processing individual test results according to processing operations defined by the neuropsychological assessment workflow, automatically detecting one or more motor impairments during performance of said neuropsychological assessment, generating a neuropsychological assessment workflow change recommendation in response to the detecting, and An output is generated indicating the proposed workflow change.

Images