Human string-pulling with and without a string: movement, sensory control, and memory

Information-theory analysis of mouse string-pulling agrees with Fitts's Law: Increasing task difficulty engages multiple sensorimotor modalities in a dual oscillator behavior

Mouse string pulling, in which a mouse reels in a string with hand-over-hand movements, can provide insights into skilled motor behavior, neurological status, and cognitive function. The task involves two oscillatory movements connected by a string. The snout oscillates to track the pendulum movement of the string produced by hand-over-hand oscillations of pulling, and so the snout guides the hands to grasp the string. The present study examines the allocation of time required to pull strings of varying diameter. Movement is also described with end-point measures, string-pulling topography with 2D markerless pose estimates based on transfer learning with deep neural networks, and Mat-lab image-segmentation and heuristic algorithms for object tracking. With reduced string diameter, mice took longer to pull 60 cm long strings. They also made more pulling cycles, misses, and mouth engagements, and displayed changes in the amplitude and frequency of pull cycles. The time measures agree with Fitts's law in showing that increased task difficulty slows behavior and engages multiple compensatory sensorimotor modalities. The analysis reveals that time is a valuable resource in skilled motor behavior and information-theory can serve as a measure of its effective use.

Automated system for training and assessing reaching and grasping behaviors in rodents

BACKGROUND

Reaching, grasping, and pulling behaviors are studied across species to investigate motor control and problem solving. String pulling is a distinct reaching and grasping behavior that is rapidly learned, requires bimanual coordination, is ethologically grounded, and has been applied across species and disease conditions.

NEW METHOD

Here we describe the PANDA system (Pulling And Neural Data Analysis), a hardware and software system that integrates a continuous string loop connected to a rotary encoder, feeder, microcontroller, high-speed camera, and analysis software for the assessment and training of reaching, grasping, and pulling behaviors and synchronization with neural data.

RESULTS

We demonstrate this system in rats implanted with electrodes in motor cortex and hippocampus and show how it can be used to assess relationships between reaching, pulling, and grasping movements and single-unit and local-field activity. Furthermore, we found that automating the shaping procedure significantly improved performance over manual training, with rats pulling > 100 m during a 15-minute session.

COMPARISON WITH EXISTING METHODS

String-pulling is typically shaped by tying food reward to the string and visually scoring behavior. The system described here automates training, streamlines video assessment with deep learning, and automatically segments reaching movements into distinct reach/pull phases. No system, to our knowledge, exists for the automated shaping and assessment of this behavior.

CONCLUSIONS

This system will be of general use to researchers investigating motor control, motivation, sensorimotor integration, and motor disorders such as Parkinson's disease and stroke.

Algorithmic assessment of shoulder function using smartphone video capture and machine learning

Tears within the stabilizing muscles of the shoulder, known as the rotator cuff (RC), are the most common cause of shoulder pain-often presenting in older patients and requiring expensive advanced imaging for diagnosis. Despite the high prevalence of RC tears within the elderly population, there is no previously published work examining shoulder kinematics using markerless motion capture in the context of shoulder injury. Here we show that a simple string pulling behavior task, where subjects pull a string using hand-over-hand motions, provides a reliable readout of shoulder mobility across animals and humans. We find that both mice and humans with RC tears exhibit decreased movement amplitude, prolonged movement time, and quantitative changes in waveform shape during string pulling task performance. In rodents, we further note the degradation of low dimensional, temporally coordinated movements after injury. Furthermore, a logistic regression model built on our biomarker ensemble succeeds in classifying human patients as having a RC tear with > 90% accuracy. Our results demonstrate how a combined framework bridging animal models, motion capture, convolutional neural networks, and algorithmic assessment of movement quality enables future research into the development of smartphone-based, at-home diagnostic tests for shoulder injury.

Automated system for training and assessing string-pulling behaviors in rodents

String-pulling tasks have been used for centuries to study coordinated bimanual motor behavior and problem solving. String pulling is rapidly learned, ethologically grounded, and has been applied to many species and disease conditions. Typically, training of string-pulling behaviors is achieved through manual shaping and baiting. Furthermore, behavioral assessment of reaching, grasping, and pulling is often performed through labor intensive manual video scoring. No system, to our knowledge, currently exists for the automated shaping and assessment of string-pulling behaviors. Here we describe the PANDA system (Pulling And Neural Data Analysis), an inexpensive hardware and software system that utilizes a continuous string loop connected to a rotary encoder, feeder, microcontroller, high-speed camera, and analysis software for assessment and training of string-pulling behaviors and synchronization with neural recording data. We demonstrate this system in unimplanted rats and rats implanted with electrodes in motor cortex and hippocampus and show how the PANDA system can be used to assess relationships between paw movements and single-unit and local-field activity. We also found that automating the shaping procedure significantly improved overall performance, with rats regularly pulling >100 meters during a 15-minute session. In conclusion, the PANDA system will be of general use to researchers investigating motor control, motivation, and motor disorders such as Parkinson's disease, Huntington's disease, and stroke. It will also support the investigation of neural mechanisms involved in sensorimotor integration.

Impairments and Compensation in String-pulling After Middle Cerebral Artery Occlusion in the Rat

Stroke is a leading cause of long-term disability in humans, and it is frequently associated with impairments in the skilled use of the arms and hands. Many human upper limb impairments and compensatory changes have been successfully modeled in rodent studies of neocortical stroke, especially those that evaluate single limb use in tasks, such as reaching for food. Humans also use their hands for bilaterally coordinated movements, dependent upon interhemispheric cortical projections, which are also compromised by unilateral stroke. This study describes middle cerebral artery occlusion (MCAO) dependent changes in the bilaterally dependent hand use behavior of string-pulling in the rat. The task involves making hand-over-hand movements to pull down a string that contains a food reward attached to its end. MCAO rats missed the string more often with both hands than Sham rats. When the string was missed on the contralateral to MCAO body side, rats continued to cycle through subcomponents of string-pulling behavior as if the string were grasped in the hand. Rats also failed to make a grasping motion with the contralateral to MCAO hand when the string was missed and instead, demonstrated an open-handed raking-like motions. Nevertheless, with repeated attempts, rats performed components of string-pulling well enough to obtain a reward on the end of the string. Thus, string-pulling behavior is sensitive to bilateral impairments but is achieved with compensatory adjustments following MCAO. These aspects of MCAO string-pulling provide a foundation for studies that investigate the efficacy of therapeutic intervention which might enhance neuroplasticity and recovery. DATA AVAILABILITY: The datasets generated during the current study are available upon request.

A Tool for Low-Cost, Quantitative Assessment of Shoulder Function Using Machine Learning

Tears within the stabilizing muscles of the shoulder, known as the rotator cuff (RC), are the most common cause of shoulder pain-often presenting in older patients and requiring expensive, advanced imaging for diagnosis1-4. Despite the high prevalence of RC tears within the elderly population, there are no accessible and low-cost methods to assess shoulder function which can eschew the barrier of an in-person physical exam or imaging study. Here we show that a simple string pulling behavior task, where subjects pull a string using hand-over-hand motions, provides a reliable readout of shoulder health across animals and humans. We find that both mice and humans with RC tears exhibit decreased movement amplitude, prolonged movement time, and quantitative changes in waveform shape during string pulling task performance. In rodents, we further note the degradation of low dimensional, temporally coordinated movements after injury. Furthermore, a predictive model built on our biomarker ensemble succeeds in classifying human patients as having a RC tear with >90% accuracy. Our results demonstrate how a combined framework bridging task kinematics, machine learning, and algorithmic assessment of movement quality enables future development of smartphone-based, at-home diagnostic tests for shoulder injury.

Fine motor deficits exhibited in rat string-pulling behavior following exposure to sleep fragmentation and deep space radiation

Deep space flight missions will expose astronauts to multiple stressors, including sleep fragmentation and space radiation. There is debate over whether sleep disruptions are an issue in deep space. While these stressors independently impair sensorimotor function, the combined effects on performance are currently unknown. String-pulling behavior involves highly organized bimanual reach-to-grasp and withdraw movements. This behavior was examined under rested wakeful conditions and immediately following one session of sleep fragmentation in Sham and irradiated rats 3 months after exposure (10 cGy ⁴Helium or 5-ion simulated Galactic Cosmic Radiation). Sleep fragmentation disrupted several aspects of string-pulling behavior, such that rats' ability to grasp the string was reduced, reach endpoint concentration was more variable, and distance traveled by the nose increased in the Y-range compared to rested wakeful performance. Overall, irradiated rats missed the string more than Sham rats 3 months post-exposure. Irradiated rats also exhibited differential impairments at 3 months, with additional deficits unveiled after sleep fragmentation. ⁴Helium-exposed rats took longer to approach the string after sleep fragmentation. Further, rats exposed to ⁴Helium traveled shorter withdraw distances 3 months after irradiation, while this only emerged in the other irradiated group after sleep fragmentation. These findings identify sleep fragmentation as a risk for fine motor dysfunction in Sham and irradiated conditions, in addition to radiation exposure. There may be complex temporal alterations in performance that are stressor- and ion-dependent. Thus, it is critical to implement appropriate models of multi-flight stressors and performance assessments in preparation for future deep space flight missions.

Skilled movement and posture deficits in rat string-pulling behavior following low dose space radiation (28Si) exposure

Deep space flight missions beyond the Van Allen belt have the potential to expose astronauts to space radiation which may damage the central nervous system and impair function. The proposed mission to Mars will be the longest mission-to-date and identifying mission critical tasks that are sensitive to space radiation is important for developing and evaluating the efficacy of counter measures. Fine motor control has been assessed in humans, rats, and many other species using string-pulling behavior. For example, focal cortical damage has been previously shown to disrupt the topographic (i.e., path circuity) and kinematic (i.e., moment-to-moment speed) organization of rat string-pulling behavior count to compromise task accuracy. In the current study, rats were exposed to a ground-based model of simulated space radiation (5 cGy ²⁸Silicon), and string-pulling behavior was used to assess fine motor control. Irradiated rats initially took longer to pull an unweighted string into a cage, exhibited impaired accuracy in grasping the string, and displayed postural deficits. Once rats were switched to a weighted string, some deficits lessened but postural instability remained. These results demonstrate that a single exposure to a low dose of space radiation disrupts skilled hand movements and posture, suggestive of neural impairment. This work establishes a foundation for future studies to investigate the neural structures and circuits involved in fine motor control and to examine the effectiveness of counter measures to attenuate the effects of space radiation on fine motor control.

A Matlab-based toolbox for characterizing behavior of rodents engaged in string-pulling

String-pulling by rodents is a behavior in which animals make rhythmical body, head, and bilateral forearm as well as skilled hand movements to spontaneously reel in a string. Typical analysis includes kinematic assessment of hand movements done by manually annotating frames. Here, we describe a Matlab-based software that allows whole-body motion characterization using optical flow estimation, descriptive statistics, principal component, and independent component analyses as well as temporal measures of Fano factor, entropy, and Higuchi fractal dimension. Based on image-segmentation and heuristic algorithms for object tracking, the software also allows tracking of body, ears, nose, and forehands for estimation of kinematic parameters such as body length, body angle, head roll, head yaw, head pitch, and path and speed of hand movements. The utility of the task and software is demonstrated by characterizing postural and hand kinematic differences in string-pulling behavior of two strains of mice, C57BL/6 and Swiss Webster.