WO2023049254A1 - System and method for automatic correlation of motor behavior with neurophysiology - Google Patents

Abstract

A system and method that objectively, reproducibly, and quantitatively associating motor behavior with neurophysiology. The present invention includes a tracking system for capturing kinematic data and a recording electrode for capturing neural data resulting from passive and/or active movement of one or more body parts of a patient. The present invention is configured to correlate the kinematic data with the neural data. In some embodiments, the correlation identifies an implant location of an implant electrode for DBS from a plurality of testing locations. Some embodiments further include determining a statistical value for the correlation of the patient's recorded data relative to the correlation of the patient's data with previously recorded data in a reference library.

Description

SYSTEM AND METHOD FOR AUTOMATIC CORRELATION OF MOTOR BEHAVIOR WITH NEUROPHYSIOLOGY

CROSS-REFERENCE TO RELATED APPLICATIONS

This nonprovisional application is a continuation of and claims priority to provisional (or nonprovisional) application No. 63/247,083, entitled “SYSTEM AND METHOD FOR AUTOMATIC CORRELATION OF MOTOR BEHAVIOR WITH NEUROPHYSIOLOGY,” filed 9/22/2021 by the same inventors.

BACKGROUND OF THE INVENTION

1. Field of the Invention

This invention relates, generally, to neurophysiology. More specifically, it relates to the correlation of motor behavior with neurophysiology.

2. Brief Description of the Prior Art

Deep brain stimulation (DBS) is a neurosurgical procedure in which electrodes are implanted into a patient’s brain in an attempt to treat a particular neurological disorder. DBS is often used to treat movement disorders, such as Parkinson’s disease, Dystonia, and Essential tremor.

One of the critical challenges for this operation is ensuring that the electrode has been targeted to the best spot in the brain to treat the patient’s symptoms. To accomplish the technically challenging task of hitting a millimeters wide target deep in patients’ brains, surgeons record neural activity along a carefully planned trajectory, as each region has an electrophysiologic “fingerprint” that may be used as landmarks (see Fig. 1 B). Once the target nucleus is identified, patients are guided through passive and active movements by a neurologist specializing in movement disorders. This motor testing is performed in order to identify motor eloquent subregions of the target since implanting in these areas has been shown to significantly improve surgical outcomes.

Since DBS therapy was approved by the FDA in 1991 , nearly all aspects of the treatment have undergone improvements with the exception of functional mapping, which is illustrated in Figs. 1A-1 D. In fact, conventional mapping techniques would appear familiar to physicians performing ablative procedures at least as far back as the 1960’s. While this continuity speaks to the efficacy of motor mapping, the process is inherently flawed due to the reliance on human perception.

As part of conventional motor mapping during DBS implantation, clinicians insert electrode probe 10 into a patient brain 12 to record electrophysiological signals. The clinicians guide awake patients through sequences of movements (see Figs. 1 C-1 D) to determine whether changes in neural activity (i.e., electrophysiological signals recorded from the brain) correlate with the movement. Clinicians use their collective expertise to determine whether they observe changes in neural activity that corresponds to motor action. This inherently subjective process requires concurrent assessment of several information streams at once: patients’ movements, digital oscilloscope display of voltage data, and audio-converted multi-unit activity (MUA) of recorded neurons. During atypical cases, even the most practiced and perceptive teams may be unable to confidently assess motor responsiveness, resulting in prolonged procedure time and suboptimal outcomes secondary to off target placement. The central challenge neurosurgical teams face is the need to correlate multiple streams of ambiguous data under time pressure, aiming to achieve optimal therapeutic accuracy in the high-stakes environment of the operating room.

Accordingly, what is needed is a system and method to map motor movement more effectively, efficiently, and consistently. What is needed is also a system and method to quantitatively associate motor behavior with neural data. However, in view of the art considered as a whole at the time the present invention was made, it was not obvious to those of ordinary skill in the field of this invention how the shortcomings of the prior art could be overcome.

While certain aspects of conventional technologies have been discussed to facilitate disclosure of the invention, Applicants in no way disclaim these technical aspects, and it is contemplated that the claimed invention may encompass one or more of the conventional technical aspects discussed herein.

The present invention may address one or more of the problems and deficiencies of the prior art discussed above. However, it is contemplated that the invention may prove useful in addressing other problems and deficiencies in a number of technical areas. Therefore, the claimed invention should not necessarily be construed as limited to addressing any of the particular problems or deficiencies discussed herein.

In this specification, where a document, act or item of knowledge is referred to or discussed, this reference or discussion is not an admission that the document, act or item of knowledge or any combination thereof was at the priority date, publicly available, known to the public, part of common general knowledge, or otherwise constitutes prior art under the applicable statutory provisions; or is known to be relevant to an attempt to solve any problem with which this specification is concerned.

BRIEF SUMMARY OF THE INVENTION

The long-standing but heretofore unfulfilled need for a system and method that can effectively, efficiently, consistently, and quantitatively associate motor behavior with neural data is now met by a new, useful, and nonobvious invention.

The present invention includes a system and method for correlating neural data with kinematic data. The present invention further includes quantitatively determining an implant location for an implantation electrode within a brain of a patient. The present invention includes acquiring recorded data from at least a first testing location. The recorded data includes a first kinematic dataset and a first neural dataset. In some embodiments, the first kinematic dataset is obtained from tracking software and corresponds to movement of one or more body parts of the patient. In some embodiments, the first neural dataset is obtained from a recording electrode probe located at the first testing location during the movement of the one or more body parts of the patient.

Once the recorded data is acquired, a first correlation is calculated. The first correlation is a mathematical correlation of the first kinematic dataset with the first neural dataset. In addition, a reference library is digitally accessed, and a plurality of reference neural datasets are acquired from the reference library. A plurality of reference correlations is then determined. Each reference correlation is a mathematical correlation of the first kinematic dataset with one of the plurality of reference neural datasets.

The present invention then determines if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations. The first testing location is quantitatively identifiable as an implant location for the implantation electrode if the first correlation satisfies the correlation threshold relative to the plurality of reference correlations. Some embodiments further include implanting the implantation electrode at the implant location within the brain of the patient if the first correlation satisfies the correlation threshold relative to the plurality of reference correlations.

The method of the present invention further includes temporarily implanting the recording electrode probe within the brain of the patient and moving the recording electrode needle to the first testing location in a plurality of testing locations. The present invention also records patient kinematics for the one or more body parts, which are tracked using one or more video capturing devices in conjunction with the tracking system. The kinematic dataset can include measurements of Euclidean distance, cosine similarity, velocity, or acceleration of one or more tracking markers.

The step of determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations may include determining if the first correlation is greater than a predetermined number of the plurality of reference correlations. In addition, the correlation threshold may be a statistical significance relative to the plurality of reference correlations.

The present invention may further determine a statistical value of the first correlation relative to the correlation of the plurality of reference correlations and identify the first testing location as a viable implant location for the implantation electrode if the statistical value satisfies a predetermined threshold.

The correlation threshold may be a quantitative value indicating that the first correlation is greater than 80% of the plurality of reference correlations. In some embodiments, the correlation threshold is a quantitative value indicating that the first correlation is greater than 95% of the plurality of reference correlations.

Further, the acquired plurality of reference neural datasets from the reference library may include neural datasets from a same target brain area in which the first neural dataset was recorded. The step of acquiring the plurality of reference neural datasets from the reference library further includes randomly selecting the neural datasets from the reference library. Some embodiments include cutting a length of each of the reference neural datasets to a same length as the first kinematic dataset.

The present invention may further include acquiring a second kinematic dataset corresponding to movement of one or more body parts of the patient and a second neural dataset obtained from the recording electrode probe located at a second testing location during the movement of the one or more body parts of the patient. The present invention then determines a second correlation for the second kinematic dataset with the second neural dataset and a second plurality of reference correlations. Each reference correlation is a mathematical correlation of the second kinematic dataset with one of the plurality of reference neural datasets. If the second correlation satisfies the correlation threshold relative to the plurality of reference correlations, the second testing location is quantitatively identifiable as the implant location for the implantation electrode.

The reference library includes a plurality of reference kinematic datasets from the reference library. The reference kinematic datasets are acquired and each of the plurality of reference correlations is a mathematical correlation of the first neural dataset with one of the plurality of reference kinematic datasets. The present invention then determines if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations, and if so, the first testing location is quantitatively identified as the implant location for the implantation electrode.

Similarly, a second plurality of reference correlations can be determined, wherein each is a mathematical correlation of the second neural dataset with one of the plurality of reference kinematic datasets. If the second correlation satisfies the correlation threshold relative to the plurality of reference correlations, the second testing location is quantitatively identified as the implant location for the implantation electrode.

The present invention may further include a system for automatically correlating motor behavior with neurophysiology data by executing the steps substantially described above and herein.

These and other important objects, advantages, and features of the invention will become clear as this disclosure proceeds.

The invention accordingly comprises the features of construction, combination of elements, and arrangement of parts that will be exemplified in the disclosure set forth hereinafter and the scope of the invention will be indicated in the claims. BRIEF DESCRIPTION OF THE DRAWINGS

For a fuller understanding of the invention, reference should be made to the following detailed description, taken in connection with the accompanying drawings, in which:

Figs. 1 A is an exemplary illustration of a probe passing through the thalamus, zona inserta, subthalamic nucleus, and substantia nigra areas of a brain.

Fig. 1 B is an exemplary display of electrode readings from a probe passing through the thalamus, zona inserta, subthalamic nucleus, and substantia nigra areas of a brain.

Fig. 1 C is an exemplary display of electrode readings from a probe in the subthalamic nucleus areas of a brain in response to passive movements of a patient’s limbs.

Fig. 1 D is an exemplary display of electrode readings from a probe in the subthalamic nucleus areas of a brain in response to active movements of a patient’s limbs.

Figs. 2A is a schematic of an embodiment of the system of the present invention.

Fig. 2B is an illustration of an exemplary set up in an operating room.

Fig. 2C is an illustrative diagram of an embodiment of the system of the present invention during use.

Fig. 3 is an illustration of markerless trackers on a patient’s hand from a dorsal view.

Fig. 4 is an illustration of markerless trackers on a patient’s hand from a ventral view.

Fig. 5 is a flowchart of an embodiment of the present invention.

Fig. 6 is a flowchart for the correlative analysis between neural and kinematic data.

Fig. 7 is an exemplary display of kinematic data and neural data and overlays as might be seen on a GUI during operation.

Fig. 8 is a diagram of the process of the correlative analysis between neural and kinematic data.

Figs. 9A includes a compilation of graphs of data indicating that the selected kinematic data is strongly correlated to the recorded MUA data in comparison to a library of recorded MUA data. The upper left graph corresponds to Euclidian distance for a 1 .67 second recording of hand clenches. The middle left graph includes time-locked MUA recordings for the above kinematic extraction. The bottom left graph is an overlay of the kinematic and MUA data showing a high degree of similarity by qualitative assessment. The large graph to the right includes a distribution of dynamic time warping (DTW) distance when comparing the extracted kinematic data in the upper left graph to a library of MUA recordings. For DTW measures, percentiles approaching 0 represent statistical significance. Fig. 9B is a compilation of graphs for additional active movement examples and how well the recorded kinematic data corresponds to the recorded MUA data in comparison to the library of MUA data.

Fig. 9C is a compilation of graphs for additional active movement examples and how well the recorded kinematic data corresponds to the recorded MUA data in comparison to the library of MUA data.

Fig. 9D includes comparisons of DTW or cross-correlation (xcorr) measures for active movement segments clinically determined as highly correlated in the intraoperative setting when using the segment of interest’s kinematic fata to compare to a library of MUA recordings (x-axis as described in the large graph in Fig. 9A) or segment of interest’s MUA data to compare to a library of kinematic recordings (y-axis). Xcorr percentiles approaching 100 represent statistical significance as DTW measures the distance between two signals and xcorr is a measure of signal overlap. Paired dot plots demonstrate relationship between DTW and xcorr measures when both are calculated using MUA libraries. Each data point represents a unique segment of movement at least 1 .5 seconds long. Generally, DTW measures outperformed xcorr and comparisons against the MUA library were more robust.

Fig. 9E includes similar data as Fig. 9D but with active movements clinically determined as not correlated.

Fig. 9F includes similar data as Fig. 9C but corresponds to passive movements.

Fig. 9G includes similar data as Fig. 9D but corresponds to passive movements.

Fig. 9H includes similar data as Fig. 9E but corresponds to passive movements.

DETAILED DESCRIPTION OF THE INVENTION

In the following detailed description of the preferred embodiments, reference is made to the accompanying drawings, which form a part thereof, and within which are shown by way of illustration specific embodiments by which the invention may be practiced. It is to be understood that other embodiments may be utilized, and structural changes may be made without departing from the scope of the invention.

As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” include plural referents unless the content clearly dictates otherwise. As used in this specification and the appended claims, the term “or” is generally employed in its sense including “and/or” unless the context clearly dictates otherwise.

In the following description, for the purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of embodiments of the present technology. It will be apparent, however, to one skilled in the art that embodiments of the present technology may be practiced without some of these specific details. The techniques introduced here can be embodied as special-purpose hardware (e.g., circuitry), as programmable circuitry appropriately programmed with software and/or firmware, or as a combination of specialpurpose and programmable circuitry. Hence, embodiments may include a machine-readable medium having stored thereon instructions which may be used to program a computer (or other electronic devices) to perform a process. The machine-readable medium may include, but is not limited to, floppy diskettes, optical disks, compacts disc read-only memories (CD-ROMs), magneto-optical disks, ROMs, random access memories (RAMs), erasable programmable read-only memories (EPROMs), electrically erasable programmable read-only memories (EEPROMs), magnetic or optical cards, flash memory, or other type of media/machine- readable medium suitable for storing electronic instructions.

The phrases “in some embodiments,” “according to some embodiments,” “in the embodiments shown,” “in other embodiments,” and the like generally mean the particular feature, structure, or characteristic following the phrase is included in at least one implementation. In addition, such phrases do not necessarily refer to the same embodiments or different embodiments.

As used herein, the term “tracking system” refers to a system configured to track the movement and location of one or more navigation markers. Some embodiments include sensors that track the position and movement of the markers relative to a coordinate origin point and the associated axes (e.g., X, Y, and Z axes) of a standard coordinate system. The tracking system may be any known to a person of ordinary skill in the art, including but not limited to optical tracking systems, RF tracking systems, ultrasonic tracking systems, electromagnetic tracking systems, and inertial tracking systems. An exemplary tracking system is NDl’s Polaris Vega tracking systems.

The tracking systems are designed to track specific markers, which can be physical and/or digital markers. The digital markers are tracked using software adapted to create and register a digital marker on an object and assess the alignment, orientation, movement, and/or position relative to a defined coordinate system.

The physical markers can be active or passive markers designed to allow a sensor to track the alignment, orientation, movement, and/or position relative to a defined coordinate system. The active markers are designed to actively emit some form of energy (e.g., EM waves or radiation) that can be tracked by a particular sensor in a tracking system. The passive markers are configured to react to or reflect energy thereby allowing a predetermined sensor to track said markers. Any markers and sensors known to a person of ordinary skill in the art may be used such that a tracking system can determine the alignment, orientation, movement, and/or position relative to a defined coordinate system.

As used herein, the term “recording electrode” means any device insertable into a patient’s brain that includes one or more electrodes capable of recording the electrical activity of the individual neurons from the surrounding brain structure. As used herein, the term “implantation electrode” means any device insertable into a patient’s brain that includes one or more electrodes capable of transmitting electrical energy to the surrounding brain structure. The implantation electrode may be implanted for long term therapeutic benefit. In some embodiments, the recording electrode is also the implantation electrode.

As previously noted, conventional approaches to correlating motor behavior with neurophysiology involves clinicians subjectively determining whether movements are associated with neural activity by comparing visual information of the patient moving, auditory information of the neural activity converted to tones, and additional information sources including but not limited to communication between clinical staff. Even under ideal circumstances, this is an error prone process that requires multiple, highly trained personnel and consumes a significant amount of intraoperative time as well as resources. Ultimately, the conventional approach is heavily reliant on subjective, qualitative analyses.

The present invention is specifically designed to overcome the inherent problems of the current technology, which are associated with subjectively and qualitatively attempting to identify associations between motor behavior and neural data. The present invention includes a system and method that is configured to objectively, reproducibly, and quantitatively associate motor behavior with neural data. The present invention overcomes the problem by objectively quantifying movement using a tracking system and mathematically correlating that movement to electrophysiological data to determine which regions of the brain experience neural activity in response to active and/or passive movements of one or more body parts of a patient.

The experimental testing results of the present invention demonstrate that the identification of mathematical correlations between extracted kinematics and neural data (e.g., multi-unit activity or MUA, referring to an electrical signal aggregating the firing of one or more neurons) relative to a plurality of reference correlations is better than the conventional approach. In addition, the present invention is superior at tracking active movements which are harder for clinicians to assess given lack of control over patient initiated movement onset yet generally yield stronger neuronal responses compared to passive movements (see Fig. 1 C and 1 D). Thus, the present invention overcomes the challenges functional neurosurgical teams face in the operating room and improves motor assessment and the association of motor movement with neural data. Such associations are particularly important in DBS clinical outcomes and other procedures in which electrodes are used to treat disorders.

The system of the present invention is a compact and mobile form factor conducive to assisting surgical procedures. In some embodiments, as illustrated in Figs. 2, the system is composed of multiple video cameras 102 (at least two) connected to a central computing unit (“CPU”) 104. CPU 104 may be any device having a memory and a processor for executing a program performing at least the steps described herein. Preferably, CPU 104 will include a user interface with a visual display (see Fig. 2B), thereby allowing a user to visually interact with CPU 104 and perceive the data captured by the system. In some embodiments, the present invention is a computer executable method or is a method embodied in software for executing the steps described herein. Further explanation of the hardware and software can be found in the Hardware and software infrastructure examples section below.

CPU 104 is configured to trigger video cameras 102 for time-locked acquisition of kinematic data and/or initiate tracking system 105 for time-locked acquisition of kinematic data. CPU 104 stores the resulting video/data on a local or remote data store and analyzes the data. The video recordings may be post-hoc concatenated into an existing library 106 of similar data to be used for neural network training and testing.

T racking system 105 is used in conjunction with video cameras 102 to track body parts. In some embodiments, video cameras 102 capture video and the tracking software is used to track body parts on the recorded video files post hoc as exemplified in Fig. 8 as steps K1 -K3. Using a machine learning-driven algorithm, acquired video recordings/files are processed using existing open-source tools (e.g., Tensorflow). This processing step allows for automated detection and tracking of specific body parts as exemplified by step K2. For DBS implantation, the system detects and tracks one or more individual finger joints, knuckles, palms, wrists, forearms, elbows, and shoulders. However, the system may detect and track other body parts.

Specific body parts may be tracked from multiple perspectives, which enables the quantification of movement kinematics with high fidelity. As such, the one or more cameras 102 are arranged in different locations as exemplified in Figs. 2B-C. Cameras 102 may be synced so that the frame capture rates are the same from the various cameras 102. The captured videos and the corresponding kinematic data are also synced with the captured neural data to allow CPU 104 to identify correlations between movement and neural data as shown in Fig. 7. An exemplary correlation process is provided in Fig. 8 and will be discussed in greater detail in subsequent sections.

Body parts are identified and tracked using tracking systems 105 known to a person of ordinary skill in the art. For example, a dorsal view of the hand may be tracked including one or more palm landmark points and various landmark points on each digit of the hand including but not limited to, the fingertip landmark point, distal interphalangeal landmark point, proximal interphalangeal landmark point, interphalangeal landmark point, and metacarpophalangeal landmark point as shown in Fig. 3. Likewise, the ventral view of a hand may be tracked including one or more palm landmark points and various landmark points on each digit of the hand including but not limited to, the fingertip landmark point, distal interphalangeal landmark point, proximal interphalangeal landmark point, interphalangeal landmark point, and metacarpophalangeal landmark point as shown in Fig. 4.

The system may use marker-less tracking as exemplified in Figs. 3-4. The marker-less tracking may include computer vision-based technology which uses neural networks to optically identify and track objects over time. However, physical trackers may additionally or alternatively be used with the tracking technology to objectively quantify movement of one or more of the patient’s body parts.

The system of the present invention further includes a recording electrode probe, such as probe 10 exemplified in Fig. 1A. The recording electrode probe may include one or more electrodes along a body of the probe, each with a known location relative to a portion of the recording electrode probe intended to remain external from the patient’s body (e.g., the handle of the probe).

The recording electrode probe is in communication with CPU 104 and is configured to record electrical activity of the individual neurons from the surrounding brain structure. Thus, the recorded neural data can be transferred to CPU 104 for processing. The recording electrode probe may be in communication with CPU 104 through a wire or the communication may be wireless.

Tracking system 105 may also be configured to identify the relative location, orientation, and movement of the recording electrode, and in turn the location of the one or more recording electrodes in the recording electrode probe relative to a predetermined coordinate system based on the patient’s brain. Tracking system 105 may be the same tracking system configured to track the patient’s motor behavior or may be a secondary tracking system tasked with tracking the recording and implantation electrodes. The recording electrode may include a physical marker secured to the handle to allow the tracking system to track the relative location, orientation, and movement recording electrode. Tracking system 105 may be configured to track the recording electrode using a digital tracker. Regardless of tracking system 105, the present invention tracks at least the depth and entry angle of the electrode needle. This information is associated with the captured neural datasets. As a result, the neural datasets can be analyzed post-hoc, and the exact corresponding location of the recording electrode remains known.

Some embodiments of the present invention also include an implantation electrode. The implantation electrode includes one or more electrodes connected to a power source for delivery electricity to the patient’s surrounding tissue. The implantation electrode may be any implantation electrode known to a person of ordinary skill in the art. In some embodiments, the implantation is the recording electrode probe.

The system of the present invention further includes reference library 106 accessible by CPU 104 as depicted in Fig. 2A. Reference library 106 stores a plurality of datasets which are used to create reference correlations for determining the statistical significance of the correlation of the neural datasets and the kinematic datasets recorded at a particular testing location. Reference library 106 may include a plurality of datasets recorded from the same patient and/or other patients. Reference library 106 may include a plurality of reference neural datasets, a plurality of reference kinematic datasets, and/or a combination of reference neural datasets and reference kinematic datasets. As will be explained below, the recorded datasets can be correlated to the reference datasets to quantitatively determine whether the recorded kinematic datasets of a testing location are associated with the recorded neural datasets for the same testing location. As a result, the present invention can quantitatively determine whether the recorded motor behavior is associated with the recorded neural data. In turn, the present invention can more accurately determine optimal implant location of an implantation electrode.

The novel method of the present invention includes tracking motor behavior of one or more body parts and mathematically correlating the recorded movement with the recorded neural data as exemplified in Figs. 5-8. The method may include identifying a precise implantation electrode/lead implant location. As such, the method further includes testing implant locations by performing or instructing a patient to perform a series of passive and/or active motor movement tests. The body parts are tracked using one or more tracking systems 105 and the corresponding kinematic datasets are correlated with recorded neural datasets captured by recording electrode 10 temporarily implanted in the brain. The tests are performed at a plurality of different testing locations in the brain to determine which locations are best correlated with certain motor movements.

As provided in Figs. 1 and 5, identifying a precise electrode implant location includes creating an opening in the patient’s skull to allow the surgeon to temporarily insert recording electrode 10 into the patient’s brain at step 202. In some instances, one or more medical images of the patient’s brain are captured before surgery to provide the surgeon with a rough idea of the testing area within the patient’s brain. Depending on the surgery, the testing area is typically an identifiable section of the brain. For DBS to treat Parkinson’s disease, the testing area is most commonly the subthalamic nucleus (STN). Representative findings illustrated in Figs. 9 are from such surgical cases.

Some embodiments include identifying a plurality of predetermined testing locations within the testing area based on the preoperative images of the patient’s brain and/or the expertise and experience of the surgeon. Each of the testing locations are generally spaced from each other by a distance of 1 mm to 0.4 mm. The difference in testing locations may be equivalent to the diameter of the recording electrode probe or the size of the electrode in the recording electrode probe.

Prior to inserting the recording electrode probe into the patient’s brain, video cameras 102 are instructed to begin recording. Depending on the system, tracking system(s) 105 are also initiated or are employed after cameras 102 capture the patient’s movement during the motor tests. The next step in the method of the present invention includes inserting the recording electrode probe into the patient’s brain to a first testing location in the testing area at step 204. The recording electrode needle may be any recording electrode probe known in the art that includes one or more electrodes capable of recording the electrical activity of the surrounding brain structure. Prior to initiating testing, the neural data recording system is initiated. The neural data recording system is in communication with the recording electrode probe and is configured to receive and store the captured neural datasets.

The surgeon can then conduct and/or instruct the patient to perform a series of passive and/or active motor movements while the system records neural data and kinematic data at step 206. The system records patient kinematics for one or more body parts using video capturing devices 102 in conjunction with tracking system 105 and recording electrode 10 captures the responsive electrical signals in the brain structure at the first testing location within the patient’s brain. The surgeon and/or patient perform one or more motor tests at each of the plurality of testing locations and the system captures the corresponding data at step 208.

During the data acquisition steps 206 and 208, the present invention can display the recordings of the data on a GUI as exemplified in Fig. 7. The display may include the kinematic data recordings and neural data recordings shown independently and/or overlayed. In addition, the user can select a region of interest by truncating the recorded data to the appropriate recording time that corresponds to the region of interest.

Referring back to Figs. 5-6, upon completion and/or during the testing of each of the plurality of testing locations (/'.e., “regions of interest”), CPU 104 begins the process of mathematically correlating the kinematic datasets from the motor movement tracking data with the recorded neural datasets (/'.e., “MUA data”) for each testing location at step 210. The testing location with the greatest mathematical correlation can be identified as the implant location and the surgeon will be notified of said location. The surgeon can then implant an implantation electrode into the patient’s brain at the implant location.

In some embodiments, the correlation process is initiated in response to a surgeon deciding that enough data has been captured. In some embodiments, the correlation process is initiated in response to the system determining that sufficient data has been captured.

During correlation 210 of the recorded kinematic data with the recorded neural data, CPU 104 calculates kinematic measures of one or more markers captured by the tracking system (see steps K2 and K3 in Fig. 8). Such measures include but are not limited to Euclidean distance, cosine similarity, velocity, and acceleration of the markers. The findings illustrated in Fig. 8 quantify kinematics using Euclidean distance. More specifically, the movement of the various trackers or landmark points can be averaged as exemplified in step K3 in Fig. 8. This information is then mathematically correlated with neural datasets recorded from the corresponding testing location as exemplified in the DTW analysis in Fig. 8 and step 211 in Fig. 6. Mathematical correlations are established based on changes in neural activity and kinematic variables, providing both a general indication of motor responsiveness as well as more detailed information like which portions of a given movement induces the greatest neurophysiologic changes.

To recap, the relationship between time series kinematic datasets and neural datasets is quantified using a moving window. The kinematic datasets and neural datasets can be overlaid, and the mathematical correlation of the data can be assessed using known methods including but not limited to dynamic time warping, cross correlation, and regression models. The method of the present invention may further include comparing the correlation of kinematic datasets and neural datasets for each testing location to determine which location exhibits the greatest correlation and said location can be identified as the preferred implant location.

However, neural datasets are often captured in a sinusoidal pattern. Therefore, coincidental correlations may occur without there being any causal correlation between the motor movement and the neural datasets. Thus, the present invention includes steps 212-214 corresponding to accessing and using reference datasets from one or more reference libraries 106 to aid in the robustness of the correlations/comparisons and prevent false positive findings.

As previously explained, the one or more reference libraries 106 include kinematic and/or neural datasets which are correlated with the recorded data for one or more testing locations. In some embodiments, of the present invention, CPU 104 randomly selects a predetermined number of reference datasets from reference library 106 for correlation.

CPU 104 further identifies the length of the datasets recorded at each testing location and cuts the length of the reference datasets to match the length of the recorded datasets. This ensures that the data can be appropriately correlated. The present invention may also randomly select the start point of the reference data and then cut the data to the required length. If cut data is too short, then the system automatically appends an inverted copy of the cut data to the initial segment, and cuts it to the correct length.

It should be noted that reference library 106 stores reference data for the same nucleus or same testing region in the brain in which the current testing location is found. For example, DBS testing and implantation often occurs in the STN for Parkinson’s disease. So, when conducting DBS testing on a patient within the STN, the present invention will rely on a reference library having reference data that was also accumulated from the STN. As further examples, DBS testing and implantation could occur in the GPi (globus pallidus internal segment) for treating dystonia, and in the VIM (ventral intermediate nucleus of the thalamus) for essential tremor. Thus, when determining the implant location of an implant electrode for DBS, the data used from the reference library will be reference data previously recorded from the same target brain areas for the same patient and/or a collection of other patients. CPU 104 calculates correlations between the patient’s kinematic datasets and previously acquired neural datasets (“reference neural data”) contained in reference library 106. Referring to Fig. 5, the method includes determining whether a specific sequence of the patient’s recorded kinematic data for a testing location is sufficiently related to the recorded time-locked neural data for the same testing location by comparing the correlation of the recorded kinematic data with a plurality of reference neural data stored in the reference library at step 21 . At step 214, the present invention determines if the correlation of the specific sequence of the patient’s recorded kinematic data and neural data for a testing location is statistically significant (e.g., p < 0.05, i.e., 5^th percentile (pct)) when compared to the null distribution (null distribution is created in step 213) generated by iteratively correlating the patient’s kinematic data and the reference neural data from the reference library. If the statistical value exceeds a predetermined threshold, then the correlation is causal rather than coincidental (see Fig. 9). In other words, the present invention can quantitatively determine that the patient’s kinematic data for a testing location is associated with the patient’s recorded neural data from the testing location.

Some embodiments use the patient’s recorded neural data and the previously acquired kinematic data (“reference kinematic data”) contained in a reference library. In such embodiments, the method includes determining whether a specific sequence of the patient’s recorded neural data for a testing location is sufficiently related to the time-locked kinematic data for the same testing location by comparing that correlation to the correlation of the same neural data with a plurality of reference kinematic data stored in reference library 105.

After calculating the correlation of the patient’s kinematic data with the neural data in the reference library or the patient’s neural data with the kinematic data in the reference library, a statistical significance for the relatedness of the patient’s recorded neural and kinematic data for the one or more testing locations can be determined based on the relatedness of the patient’s recorded data with the reference library data at step 214. Some embodiments identify the statistical significance for the relatedness of the patient’s neural and kinematic data for each testing location. Any testing locations exceeding a threshold for statistical significance are identified as potential implantation locations for the implant electrode. In some embodiments, the testing location having the greatest statistical significance for relatedness is identified as the implant location.

Some embodiments include testing the plurality of testing locations until one of the testing locations captures data having a statistical significance for the relatedness of the current patient’s neural and kinematic data that satisfies a predetermined threshold. If the statistical significance satisfies a predetermined threshold for a particular testing location, then that testing location is identified as the implant location for the implantation electrode needle.

In some embodiments, the threshold of relatedness is a greater mathematical correlation than 80% of the correlations of the patient’s recorded data with the reference data in the reference library. In some embodiments, the threshold of relatedness is a greater mathematical correlation than 95% of the correlations of the patient’s recorded data with the reference data in the reference library. In some embodiments, the threshold is a statistical significance that is less than or equal to a standard p value of 0.05 (i.e., less than or equal to the 5^th percentile (pct) when comparing to iteratively constructed null distribution; see Figs. 9). In some embodiments, the threshold is a statistical significance that is less than or equal to a standard p value of 0.2 (i.e., less than or equal to the 20^th percentile (pct) when comparing to iteratively constructed null distribution).

Some embodiments of the present invention acquire kinematic data and neural data at different frequencies. Thus, some embodiments resample the acquired kinematic data and/or the neural data to the same frequency prior to initiating the correlation process. For example, kinematic data at 60Hz and neural data at 44kHz can each be resampled to 1200Hz. A frequency of 1200Hz adds minimal noise when upsampling kinematic data and preserves the most relevant information contained within the electrophysiologic signal when recording at most 3 neurons per lead with firing rates <125Hz.

For processing the electrophysiology data, some embodiments of the system apply filters for analyzing the neural data (see e.g., step N1 in Fig. 8). For example, the time series voltage data can be zero-lag filtered with a 2nd order butterworth bandpass filter between 300 and 3000 Hz normalized to the Nyquist frequency (22kHz) (step N2 in Fig. 8 exemplifies the normalizing of the MUA data). Some embodiments also calculate the standard deviation of background noise (SDbn) by taking the median of the absolute value of the band passed signal divided by 0.6745 to establish a threshold for spike detection which would not be inflated by anticipated high firing rate. In some embodiments, beginning with a baseline multiplier of 4.5 (4.5*SDbn) applied to the threshold, average firing rate is estimated within a possible range of 2.5*SDbn to 6.5*SDbn until an average firing rate of >80Hz and <400Hz is achieved or the limits of the range met.

In some embodiments, firing rate thresholds are determined based on expected nucleus firing rates (e.g., 80-125Hz for STN neurons) and expected number of neurons per recording (e.g., 1 -3 neurons). This dynamic but constrained threshold is employed to account for varying levels of signal to noise (spike to background) found between recordings taken at different depths and across subjects. To reconstruct the peak positions of detected spikes when downsampling from 44kHz to 1200Hz for comparison with kinematic data, timestamps are divided by 44/1.2 and rounded to the nearest integer values. In cases where this process results in multiple instances of a downsampled timestamp value, only one representation is retained. Next, instantaneous firing rate are calculated for each sample point (1200Hz) as the reciprocal of the interval between two spikes surrounding a sampled time point.

Some embodiments focus on a single marker for a single body part. Alternatively, for consistency in analyzing kinematic data, some embodiments average the values for multiple related body parts (e.g., all five fingertips) to represent general motion (see e.g., K3 in Fig. 8). Averaging across all five fingertips has the additional benefit of reducing the impact of noisy or artifactual kinematic data from individual points. In some embodiments, prior to comparative analyses, signals are normalized (-1 to 1 ) and smoothed to aid in visual assessment of similarity/dissimilarity.

To quantify the relationship between kinematic and neural time series data, an embodiment of the present invention uses dynamic time warping (DTW) analyses (see Fig. 9B). For DBS analysis, an upper limit of 200ms is used for the lag between signals based on documented delays of ~100ms between corticomotor neuronal firing and joint movement (Chen & Hallett, 1999; Cheney & Fetz, 1980; Lamarre, Spidalieri, & Lund, 1981 ; Van Acker, Luchies, & Cheney, 2016; Wannier, Maier, & Hepp-Reymond, 1991 ). The additional buffer is intended to account for potential PD related effects on motor processing, which would presumably increase the duration and variability of the lag (Burciu & Vaillancourt, 2018; McGregor & Nelson, 2019; Underwood & Parr-Brownlie, 2021 ).

Experimentation:

Intraoperative electrophysiological recordings were collected from 7 subjects recruited at the University of Colorado Anschutz Medical Campus through the Neurology Movement Disorders Center in patients undergoing DBS surgery for treatment of PD. The STN was targeted for all patients.

Kinematic Data Setup, Calibration and Acquisition

The procedure for DBS implantation surgery targeting STN for treatment of PD has been previously described in detail.¹¹¹ Briefly, following standard imaging-based stereotactic planning for trajectory and surgical target, intraoperative microelectrode recordings (MER) were used to locate STN. All collected and analyzed electrophysiological data were acquired using the NeuroOmega system (Alpha Omega Engineering, Nazareth Israel). Raw and spiking signals were sampled at 44 kHz by a 16-bit A/D converter (using ±1.25V input range, i.e., ~2 pV amplitude resolution) and band-passed from 0.7 to 9000 Hz using a hardware four-pole Butterworth filter. MER began at 25 millimeters above the ventral border of STN and advanced in steps of 100-1000 micrometers.

After the patient was positioned on the operating table, two FLIR cameras (USB 3.0 Blackfly) mounted on monopods (Avella A324D Aluminum 67 Inch Video Monopod) were positioned to capture motor testing. One camera was positioned at the foot of the bed with the other positioned at the side of the bed contralateral to motor testing. Cameras were connected to a laptop which controlled image data collection via a FLIR Spinnaker Python SDK and usergenerated Python GUI (CU Anschutz IDEA Core). To allow 3D analysis after data capture, brief recordings of a 6x8 checkerboard were collected to permit triangulation between the cameras. During post-processing, using DeepLabCut, corner detection of the checkerboard grid was used to assess calibration quality. A mean reprojection pixel error (MRPE) metric threshold (< 1 pixel) was sought; among the ten cases in this novel investigation, the mean error was 0.0244 pixels, indicative of acceptable accuracy and reliability.

The cameras were used to acquire kinematic data collected during the routine MER assessment of STN activity. Per standard procedure, once STN was encountered, motor testing via passive and active upper and lower limb movements was conducted every 0.3 to 0.5 mm. Video data were collected for 10-90 seconds during kinematic testing (9 average number of videos per subject, 35 average duration per video). Clinical notes describing MUA responsiveness to movement were extracted from electronic medical records and characterized as either present or absent. All subsequent analyses were computed offline.

Kinematic Data Extraction

All kinematic data extraction and neural network training and testing was conducted with DeepLabCut and run in Linux. Briefly, patient specific models were constructed with an automatic, k-means clustering extraction method to isolate a subset of captured frames. Frames were manually labeled (700-912 frames per case, p=764, 3.22% of total; range = 14,400-50,400 per case, p=29,900) to identify 21 visual anatomical landmarks of the hand (both ventral and dorsal views), including the base of the palm, center of the palm, metacarpophalangeal joints, proximal interphalangeal joints, distal interphalangeal joints, and tips of all digits.

Several parameters were adjusted prior to network training. The TrainingFraction parameter, which defined a fraction of the network data used for training and testing was set at 0.95, indicative of 95% reserved for iterative training. i¹' The pretrained ResNet-50 was employed for all trainings, which contains 50 iteratively-trained layers in object identification and probability density mapping for accurate tracking, demonstrating efficacy given a very small RMSE (3.09 ± 0.04) and accuracy with datasets as small as 200 frames, i¹' The p-cutoff variable was raised from 0.01 to 0.06, as it serves to establish the likelihood threshold of distinguishing background from points of interest since the labeled points of interest were small. Lastly, pos_dist_thresh was lowered from 17 to 15 (case 1 ) and 13 (cases 2-7) given small fields of interest akin to p-cutoff.

These models were trained until a plateau was reached in the network performance as measured by Huber loss to yield the smallest root mean square error, in which training iterations ranged from 195,000-268,100 (p=224,250). Networks achieved proficient performance as defined by >95% accuracy of labeled test frames. In instances where accuracy remained below the >95% threshold despite best efforts, sessions were not considered for further evaluation. Under such standards, among all cases, video samples deemed viable increased from 28.57% to 100% given adjustment of aforementioned parameters and iteration count.

Kinematic Data Analysis Following extraction of marker-less tracking data from video files, data were exported to Matlab wherein kinematic measures including Euclidean distance, cosine similarity, velocity, and acceleration of each labeled point were calculated. For consistency and to simplify our analyses, the study utilized Euclidean distance as the primary kinematic variable for all results. In order to compare kinematic data acquired at 60Hz with MUA acquired at 44kHz, each were resampled to 1200Hz. This frequency was chosen as it added minimal noise when upsampling kinematic data and preserved the most relevant information contained within the electrophysiologic signal since the expectation was to record at most 3 neurons per lead with firing rates <125Hz, thereby staying comfortably within 1200Hz. The Matlab function resample was used to account for anti-aliasing. With respect to kinematic data, artifacts from temporary loss of tracked points were detected using a 4.5*SD threshold. Within windows identified as artifactual, kinematic data was locally interpolated to maintain underlying structure of movements.

Electrophysiology Data Analysis

For processing the electrophysiology data, the system applied standard filters for analyzing multi-unit activity (MUAjJ⁵' In brief, time series voltage data was zero-lag filtered (Matlab function filtfilt) with a 2nd order butterworth bandpass filter between 300 and 3000 Hz normalized to the Nyquist frequency (22kHz). To establish a threshold for spike detection which would not be inflated by anticipated high firing rate, an estimate for the standard deviation of background noise (SDbn) was calculated by taking the median of the absolute value of the band passed signal divided by 0.6745.^[5’⁷¹ Beginning with a baseline multiplier of 4.5 (/'.e., 4.5*SDbn) applied to the threshold, average firing rate was estimated within a possible range of 2.5* SDbn to 6.5* SDbn until an average firing rate of >80Hz and <400Hz was achieved or the limits of the range met.

Firing rate thresholds were determined based on expected STN firing rates (80-125Hz) and expected number of neurons per recording (1 -3 neurons). This dynamic but constrained threshold was employed to account for varying levels of signal to noise (spike to background) found between recordings taken at different depths and across subjects. To reconstruct the peak positions of detected spikes when downsampling from 44kHz to 1200Hz for comparison with kinematic data, timestamps were divided by 44/1.2 and rounded to the nearest integer values. In cases where this process resulted in multiple instances of a downsampled timestamp value, only one representation was retained. Next, instantaneous firing rate was calculating for each sample point (1200Hz) as the reciprocal of the interval between two spikes surrounding a sampled time point.

Comparative measures

Video data were passively collected during the routine kinesthetic testing and not intentionally time-locked to the start movements. Thus, investigators selected frames which contained motor testing post-hoc for further analysis. Examples of movements utilized during intraoperative motor testing include “passive” movements during which the attending movement disorder neurologist manipulated the patient’s limbs, or “active” movements characterized by initiation and continuation of movements by patients themselves. Frame numbers were then converted to the appropriate resampled start and end points for 1200Hz sampling frequency. For consistency in analyzing kinematic data, an averaged value for all five fingertips was used to represent general motion. Averaging across all five fingertips had the additional benefit of reducing the impact of noisy or artifactual kinematic data from individual points. Prior to comparative analyses, signals were normalized (-1 to 1 ) and smoothed to aid in visual assessment of similarity/dissimilarity.

To quantify the relationship between kinematic and MUA time series data, dynamic time warping (DTW) analyses (Matlab functions dtw) (see Fig. 9A) was used. An upper limit of 200ms was used for the lag between signals based on documented delays of ~100ms between corticomotor neuronal firing and joint movement.!⁷'’ I®]’ Pl- h2], [13] The additional buffer is intended to account for potential PD related effects on motor processing, which would presumably increase the duration and variability of the lag.!®!’ ot I¹¹1 Given previous findings concerning STN firing facilitating movement initiation as well as motor braking, it is likely that both phasic and anti-phasic relationships would be revealed between kinematics and MUA. As DTW analyses are less amenable to parsing anti-phasic relationships than phasic, signals using cross correlation (Matlab, xcorr) were first compared and if a negative r value results, one of the two signals was inverted before DTW analysis. In general, findings between xcorr and DTW analyses were mutually supportive (see Figs. 9D, 9E, 9G, and 9H). For DTW, the Euclidean distance between relative points along each curve was calculated resulting in values closer to 0 representing more similar signals.

The conclusions were further validated by applying bootstrapping methods to our comparisons between the kinematic and electrophysiological data. Specifically, the kinematic data of interest was compared with a reference library of MUA activity, which was composed of 1 ,000 randomly selected sections of electrophysiologic data taken from other instances of motor testing across all patients. The window was randomly selected from each MUA recording to ensure a systematic error would not compromise the statistical measures. To prevent attenuated electrophysiologic data series in instances where the randomly selected start of the reference period occurred at the end of a recording, electrophysiologic data was replicated, inverted, and appended to the end of the time series data originally replicated. This ensured consistent lengths for randomly selected electrophysiologic data without the need to interpolate.

To further test the findings, the same bootstrapping method was employed but with comparing the selected electrophysiologic recordings to a library of kinematic data (Fig. 9D). The comparison of selected kinematic data to electrophysiologic library was preferentially used because electrophysiologic signals were found to be less variable and were not susceptible to quiescent periods as kinematic data were (i.e., a patient’s hand may be out of frame or be at rest for significant portions of a recording, while MUA is consistently active), thereby serving as a more consistent reference. Furthermore, compiling a substantial electrophysiologic library is more feasible than a large kinematic library given the widespread use of MER by neurosurgical clinical teams.

Ultimately, the experimentation proves that present invention can quantify a qualitatively complex procedure prone to human error in a fast and reproducible manner. It should also be noted that specific subtypes of patients will have a specific electrophysiological characteristics that respond differently to different types of treatments. As a result, the present invention can be used to determine a subclass of individuals to better determine the best treatment for an individual falling within one of these subclasses.

Hardware and software infrastructure examples

The present invention may be embodied on various computing systems and/or platforms that perform actions responsive to software-based instructions. The following provides an antecedent basis for the information technology that may be utilized to enable the invention.

The computer readable medium described in the claims below may be a computer readable signal medium or a computer readable storage medium. A computer readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of the computer readable storage medium would include the following: an electrical connection having one or more wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In the context of this document, a computer readable storage medium may be any non-transitory, tangible medium that can contain, or store a program for use by or in connection with an instruction execution system, apparatus, or device.

A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wire-line, optical fiber cable, radio frequency, etc., or any suitable combination of the foregoing. Computer program code for carrying out operations for aspects of the present invention may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C#, C++, Visual Basic, Matlab, Python, Julia or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages.

Aspects of the present invention may be described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general-purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer readable medium that can direct a computer, other programmable data processing apparatus, or other devices to function in a particular manner, such that the instructions stored in the computer readable medium produce an article of manufacture including instructions which implement the function/act specified in the flowchart and/or block diagram block or blocks.

The computer program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other devices to cause a series of operational steps to be performed on the computer, other programmable apparatus or other devices to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide processes for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks.

References

[1] Mathis A, Mamidanna P, Cury KM, Abe T, Murthy VN, Mathis MW, et al. DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nat Neurosci. 2018;21 (9):1281 -9.

[2] Mathis A, Schneider S, Lauer J, Mathis MW. A Primer on Motion Capture with Deep Learning: Principles, Pitfalls, and Perspectives. Neuron. 2020;108(1 ):44-65.

[3] Masse NY, Jarosiewicz B, Simeral JD, Bacher D, Stavisky SD, Cash SS, et al. Reprint of "Non-causal spike filtering improves decoding of movement intention for intracortical BCIs". J Neurosci Methods. 2015;244:94-103. [4] Quian Quiroga R. What is the real shape of extracellular spikes? J Neurosci Methods. 2009;177(1 ):194-8.

[5] Quiroga RQ, Nadasdy Z, Ben-Shaul Y. Unsupervised spike detection and sorting with wavelets and superparamagnetic clustering. Neural Comput. 2004;16(8):1661 -87.

[6] Burciu, R. G., & Vaillancourt, D. E. (2018). Imaging of Motor Cortex Physiology in Parkinson's Disease. Mov Disord, 33(11 ), 1688-1699. doi:10.1002/mds.102

[7] Chen, R., & Hallett, M. (1999). The time course of changes in motor cortex excitability associated with voluntary movement. Can J Neurol Sci, 26(3), 163-169. doi: 10.1017/S0317167100000196

[8] Cheney, P. D., & Fetz, E. E. (1980). Functional classes of primate corticomotoneuronal cells and their relation to active force. J Neurophysiol, 44(4), 773-791. doi:10.1 152/jn.1980.44.4.773

[9] Lamarre, Y., Spidalieri, G., & Lund, J. P. (1981 ). Patterns of muscular and motor cortical activity during a simple arm movement in the monkey. Can J Physiol Pharmacol, 59(7), 748-756. doi:10.1139/y81 -1 11

[10] McGregor, M. M., & Nelson, A. B. (2019). Circuit Mechanisms of Parkinson's Disease. Neuron, 101(6), 1042-1056. doi:10.1016/j.neuron.2019.03.004

[11 ] Underwood, C. F., & Parr-Brownlie, L. C. (2021 ). Primary motor cortex in Parkinson's disease: Functional changes and opportunities for neurostimulation. Neurobiol Dis, 147, 105159. doi: 10.1016/j.nbd.2020.105159

[12] Van Acker, G. M., 3rd, Luchies, C. W., & Cheney, P. D. (2016). Timing of Cortico- Muscle Transmission During Active Movement. Cereb Cortex, 26(8), 3335-3344. doi:10.1093/cercor/bhv151

[13] Wannier, T. M., Maier, M. A., & Hepp-Reymond, M. C. (1991 ). Contrasting properties of monkey somatosensory and motor cortex neurons activated during the control of force in precision grip. J Neurophysiol, 65(3), 572-589. doi:10.1 152/jn.1991 .65.3.572

Where a definition or use of a term in a reference is inconsistent or contrary to the definition of that term provided herein, the definition of that term provided herein applies and the definition of that term in the reference does not apply.

The advantages set forth above, and those made apparent from the foregoing description, are efficiently attained. Since certain changes may be made in the above construction without departing from the scope of the invention, it is intended that all matters contained in the foregoing description or shown in the accompanying drawings shall be interpreted as illustrative and not in a limiting sense. It is also to be understood that the following claims are intended to cover all of the generic and specific features of the invention herein described, and all statements of the scope of the invention that, as a matter of language, might be said to fall therebetween.

Claims

24 What is claimed is:

1 . A method for quantitatively determining an implant location for an implantation electrode within a brain of a patient, comprising: acquiring recorded data from a first testing location, the recorded data including: a first kinematic dataset obtained from a tracking system, wherein the first kinematic dataset corresponds to movement of one or more body parts of the patient; a first neural dataset, wherein the first neural dataset obtained from a recording electrode probe located at the first testing location during the movement of the one or more body parts of the patient; calculating a first correlation, wherein the first correlation is a mathematical correlation of the first kinematic dataset with the first neural dataset; digitally accessing a reference library and acquiring a plurality of reference neural datasets from the reference library; determining a plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the first kinematic dataset with one of the plurality of reference neural datasets; determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations; whereby the first testing location is quantitatively identifiable as an implant location for the implantation electrode if the first correlation satisfies the correlation threshold relative to the plurality of reference correlations.

2. The method of claim 1 , further including: acquiring a second kinematic dataset obtained from tracking system, wherein the second kinematic dataset corresponds to movement of one or more body parts of the patient; acquiring a second neural dataset, wherein the second neural dataset was obtained from the recording electrode probe located at a second testing location during the movement of the one or more body parts of the patient; determining a second correlation for second kinematic dataset with the second neural dataset; determining a second plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the second kinematic dataset with one of the plurality of reference neural datasets; determining if the second correlation satisfies the correlation threshold relative to the second plurality of reference correlations; whereby the second testing location is quantitatively identifiable as the implant location for the implantation electrode if the second correlation satisfies the correlation threshold relative to the second plurality of reference correlations.

3. The method of claim 1 , wherein determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations includes determining if the first correlation is greater than a predetermined number of the plurality of reference correlations.

4. The method of claim 1 , wherein the correlation threshold is a statistical significance relative to the plurality of reference correlations.

5. The method of claim 1 , wherein the correlation threshold is a quantitative value indicating that the first correlation is greater than 80% of the plurality of reference correlations.

6. The method of claim 1 , wherein the correlation threshold is a quantitative value indicating that the first correlation is greater than 95% of the plurality of reference correlations.

7. The method of claim 1 , wherein acquiring the plurality of reference neural datasets from the reference library further includes randomly selecting the neural datasets from the reference library.

8. The method of claim 1 , further including cutting a length of each of the reference neural datasets to a same length as the first kinematic dataset.

9. The method of claim 1 , wherein the acquired plurality of reference neural datasets from the reference library includes neural datasets from a same target brain area in which the first neural dataset was recorded.

10. The method of claim 1 , further including: temporarily implanting the recording electrode probe within the brain of the patient and moving the recording electrode needle to the first testing location in a plurality of testing locations; recording patient kinematics for the one or more body parts, wherein the patient kinematics are tracked using one or more video capturing devices in conjunction with a tracking system; and recording neural data through the recording electrode probe.

11 . The method of claim 1 , further including implanting the implantation electrode at the implant location within the brain of the patient.

12. The method of claim 1 , wherein the kinematic dataset includes measurements of Euclidean distance, cosine similarity, velocity, or acceleration of one or more tracking markers.

13. A method for quantitatively determining an implant location for an implantation electrode within a brain of a patient, comprising: acquiring recorded data from a first testing location, the recorded data including: a first kinematic dataset obtained from tracking system, wherein the first kinematic dataset corresponds to movement of one or more body parts of the patient; a first neural dataset, wherein the first neural dataset was obtained from a recording electrode probe located at the first testing location during the movement of the one or more body parts of the patient; calculating a first correlation, wherein the first correlation is a mathematical correlation of the first kinematic dataset with the first neural dataset; digitally accessing a reference library and acquiring a plurality of reference kinematic datasets from the reference library; determining a plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the first neural dataset with one of the plurality of reference kinematic datasets; determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations; whereby the first testing location is quantitatively identifiable as an implant location for the implantation electrode if the first correlation satisfies the correlation threshold relative to the plurality of reference correlations.

14. The method of claim 13, further including: acquiring a second kinematic dataset obtained from tracking system, wherein the second kinematic dataset corresponds to movement of one or more body parts of the patient; acquiring a second neural dataset, wherein the second neural dataset was obtained from the recording electrode probe located at a second testing location during the movement of the one or more body parts of the patient; determining a second correlation for second kinematic dataset with the second neural dataset; 27 determining a second plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the second neural dataset with one of the plurality of reference kinematic datasets; determining if the second correlation satisfies the correlation threshold relative to the plurality of reference correlations; whereby the second testing location is quantitatively identifiable as the implant location for the implantation electrode if the second correlation satisfies the correlation threshold relative to the plurality of reference correlations.

15. The method of claim 13, wherein determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations includes determining if the first correlation is greater than a predetermined number of the plurality of reference correlations.

16. The method of claim 13, wherein the correlation threshold is a statistical significance relative to the plurality of reference correlations.

17. The method of claim 13, wherein the correlation threshold is a quantitative value indicating that the first correlation is greater than 80% of the plurality of reference correlations.

18. The method of claim 13, wherein the correlation threshold is a quantitative value indicating that the first correlation is greater than 95% of the plurality of reference correlations.

19. The method of claim 13, wherein acquiring the plurality of reference kinematic datasets from the reference library further includes randomly selecting the kinematic datasets from the reference library.

20. The method of claim 13, further including cutting a length of each of the reference kinematic datasets to a same length as the first neural dataset.

21 . The method of claim 13, wherein the acquired plurality of reference kinematic datasets from the reference library includes kinematic datasets from a same target brain area in which the first kinematic dataset was recorded.

22. The method of claim 13, further including: temporarily implanting the recording electrode probe within the brain of the patient and moving the recording electrode needle to the first testing location in a plurality of testing locations; recording patient kinematics for the one or more body parts, wherein the patient kinematics are tracked using one or more video capturing devices in conjunction with the tracking system; and 28 recording neural data through the recording electrode probe.

23. The method of claim 13, further including implanting the implantation electrode at the implant location within the brain of the patient.

24. The method of claim 13, wherein the kinematic dataset includes measurements of Euclidean distance, cosine similarity, velocity, or acceleration of one or more tracking markers.

25. A method for quantitatively determining an implant location for an implantation electrode within a brain of a patient, comprising: temporarily implanting a recording electrode probe within the brain of the patient to a first testing location in a plurality of testing locations; in response to a first set of movements of one or more body parts of the patient: recording a first kinematic dataset obtained from tracking system, wherein the first kinematic dataset corresponds to movement of the one or more body parts of the patient; recording a first neural dataset, wherein the first neural dataset is obtained during the movement of the one or more body parts of the patient; calculating a first correlation, wherein the first correlation is a mathematical correlation of the first kinematic dataset with the first neural dataset; digitally accessing a reference library and acquiring a plurality of reference neural datasets from the reference library; determining a plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the first kinematic dataset with one of the plurality of reference neural datasets; determining a statistical value of the first correlation relative to the correlation of the plurality of reference correlations; and identifying the first testing location as a viable implant location for the implantation electrode if the statistical value satisfies a predetermined threshold.

26. The method of claim 25, further including implanting the implantation electrode at the first testing location if the statistical value satisfies the predetermined threshold.

27. A method for quantitatively determining an implant location for an implantation electrode within a brain of a patient, comprising: temporarily implanting a recording electrode probe within the brain of the patient to a first testing location in a plurality of testing locations; 29 in response to a first set of movements of one or more body parts of the patient: recording a first kinematic dataset obtained from tracking system, wherein the first kinematic dataset corresponds to movement of the one or more body parts of the patient; recording a first neural dataset, wherein the first neural dataset is obtained during the movement of the one or more body parts of the patient; calculating a first correlation, wherein the first correlation is a mathematical correlation of the first kinematic dataset with the first neural dataset; digitally accessing a reference library and acquiring a plurality of reference kinematic datasets from the reference library; determining a plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the first neural dataset with one of the plurality of reference kinematic datasets; determining a statistical value of the first correlation relative to the correlation of the plurality of reference correlations; and identifying the first testing location as a viable implant location for the implantation electrode if the statistical value satisfies a predetermined threshold.

28. The method of claim 27, further including implanting the implantation electrode at the first testing location if the statistical value satisfies the predetermined threshold.

29. A method for correlating neural data from a testing location with kinematic data, comprising: acquiring recorded data from a first testing location, the recorded data including: a first kinematic dataset obtained from a tracking system, wherein the first kinematic dataset corresponds to movement of one or more body parts of the patient; a first neural dataset, wherein the first neural dataset obtained from a recording electrode probe located at the first testing location during the movement of the one or more body parts of the patient; calculating a first correlation, wherein the first correlation is a mathematical correlation of the first kinematic dataset with the first neural dataset; digitally accessing a reference library and acquiring a plurality of reference neural datasets from the reference library; 30 determining a plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the first kinematic dataset with one of the plurality of reference neural datasets; and determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations.

30. The method of claim 29, further including: acquiring a second kinematic dataset obtained from tracking system, wherein the second kinematic dataset corresponds to movement of one or more body parts of the patient; acquiring a second neural dataset, wherein the second neural dataset was obtained from the recording electrode probe located at a second testing location during the movement of the one or more body parts of the patient; determining a second correlation for second kinematic dataset with the second neural dataset; determining a second plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the second kinematic dataset with one of the plurality of reference neural datasets; and determining if the second correlation satisfies the correlation threshold relative to the second plurality of reference correlations.

31 . The method of claim 29, wherein determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations includes determining if the first correlation is greater than a predetermined number of the plurality of reference correlations.

32. The method of claim 29, wherein the correlation threshold is a statistical significance relative to the plurality of reference correlations.

33. The method of claim 29, wherein the correlation threshold is a quantitative value indicating that the first correlation is greater than 80% of the plurality of reference correlations.

34. The method of claim 29, wherein the correlation threshold is a quantitative value indicating that the first correlation is greater than 95% of the plurality of reference correlations.

35. The method of claim 29, wherein acquiring the plurality of reference neural datasets from the reference library further includes randomly selecting the neural datasets from the reference library. 31

36. The method of claim 29, further including cutting a length of each of the reference neural datasets to a same length as the first kinematic dataset.

37. The method of claim 29, wherein the acquired plurality of reference neural datasets from the reference library includes neural datasets from a same target brain area in which the first neural dataset was recorded.

38. The method of claim 29, further including: temporarily implanting the recording electrode probe within the brain of the patient and moving the recording electrode needle to the first testing location in a plurality of testing locations; recording patient kinematics for the one or more body parts, wherein the patient kinematics are tracked using one or more video capturing devices in conjunction with a tracking system; and recording neural data through the recording electrode probe.

39. The method of claim 29, further including implanting the implantation electrode at the implant location within the brain of the patient.

40. The method of claim 29, wherein the kinematic dataset includes measurements of Euclidean distance, cosine similarity, velocity, or acceleration of one or more tracking markers.

41 . A system for correlating neural data from a testing location with kinematic data, comprising: a tracking system configured to obtain a first kinematic dataset, wherein the first kinematic dataset corresponds to movement of one or more body parts of a patient; a recording electrode probe configured to record neural data from a first testing location during the movement of the one or more body parts of the patient; a computer system configured to calculate a first correlation, wherein the first correlation is a mathematical correlation of the first kinematic dataset with the first neural dataset; a reference library in communication with the computer system, the reference library containing a plurality of reference neural datasets; the computer system configured to execute the following steps: determine a plurality of reference correlations, wherein each reference correlation is a mathematical correlation of the first kinematic dataset with one of the plurality of reference neural datasets; and 32 determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations.

42. The system of claim 29, wherein determining if the first correlation satisfies a correlation threshold relative to the plurality of reference correlations includes determining if the first correlation is greater than a predetermined number of the plurality of reference correlations.

43. The system of claim 29, wherein the correlation threshold is a statistical significance relative to the plurality of reference correlations.

44. The system of claim 29, wherein the correlation threshold is a quantitative value indicating that the first correlation is greater than 80% of the plurality of reference correlations.

45. The system of claim 29, wherein the correlation threshold is a quantitative value indicating that the first correlation is greater than 95% of the plurality of reference correlations.

46. The system of claim 29, wherein acquiring the plurality of reference neural datasets from the reference library further includes randomly selecting the neural datasets from the reference library.

47. The system of claim 29, further including cutting a length of each of the reference neural datasets to a same length as the first kinematic dataset.

48. The system of claim 29, wherein the acquired plurality of reference neural datasets from the reference library includes neural datasets from a same target brain area in which the first neural dataset was recorded.