Data-driven analyses of motor impairments in animal models of neurological disorders.

Hardeep Ryait,Ian Q Whishaw,Matthew Harvey,Jamshid Faraji,Gerlinde A S Metz,Andrea Gomez-Palacio Schjetnan,Artur Luczak,Jon Doan,Edgar Bermudez-Contreras,Behroo Mirza Agha,Aaron Gruber,Majid Mohajerani,James Ashe

doi:10.1371/journal.pbio.3000516

Abstract

Behavior provides important insights into neuronal processes. For example, analysis of reaching movements can give a reliable indication of the degree of impairment in neurological disorders such as stroke, Parkinson disease, or Huntington disease. The analysis of such movement abnormalities is notoriously difficult and requires a trained evaluator. Here, we show that a deep neural network is able to score behavioral impairments with expert accuracy in rodent models of stroke. The same network was also trained to successfully score movements in a variety of other behavioral tasks. The neural network also uncovered novel movement alterations related to stroke, which had higher predictive power of stroke volume than the movement components defined by human experts. Moreover, when the regression network was trained only on categorical information (control = 0; stroke = 1), it generated predictions with intermediate values between 0 and 1 that matched the human expert scores of stroke severity. The network thus offers a new data-driven approach to automatically derive ratings of motor impairments. Altogether, this network can provide a reliable neurological assessment and can assist the design of behavioral indices to diagnose and monitor neurological disorders.

Highlights

Classification and quantification of behavior is central to understanding normal brain function and changes associated with neurological conditions [1,2]
The first consisted of a convolutional network (ConvNet), Inception-V3 [36] (Methods)
The function of the ConvNet was to convert each video frame (300 × 300 pixels) to a set of 2,048 features to reduce the dimensionality of the data

Summary

Introduction

Classification and quantification of behavior is central to understanding normal brain function and changes associated with neurological conditions [1,2]. Investigations of neurological disorders are aided by preclinical animal analogues that include laboratory rodents such as rats and mice. Whereas hand use is important to most human activities, rodents use their hands for building nests, digging, walking, running, climbing, pulling strings, grooming, caring for young, and for feeding—essentially, for much of their behavior. A number of laboratory tests have been developed to assess skilled hand use in rodents, including having an animal reach into a tube or through a window to retrieve a food pellet or having an animal operate a manipulandum or pull on a string to obtain food [3,4,5,6,7,8,9,10,11].

Methods

Results

Conclusion