Limits...
A dedicated visual pathway for prey detection in larval zebrafish.

Semmelhack JL, Donovan JC, Thiele TR, Kuehn E, Laurell E, Baier H - Elife (2014)

Bottom Line: Two-photon calcium imaging revealed a small visual area, AF7, that was activated specifically by the optimal prey stimulus.We identified neurons with arbors in AF7 and found that they projected to multiple sensory and premotor areas: the optic tectum, the nucleus of the medial longitudinal fasciculus (nMLF) and the hindbrain.These findings indicate that computations in the retina give rise to a visual stream which transforms sensory information into a directed prey capture response.

View Article: PubMed Central - PubMed

Affiliation: Department Genes-Circuits-Behavior, Max Planck Institute of Neurobiology, Martinsried, Germany.

ABSTRACT
Zebrafish larvae show characteristic prey capture behavior in response to small moving objects. The neural mechanism used to recognize objects as prey remains largely unknown. We devised a machine learning behavior classification system to quantify hunting kinematics in semi-restrained animals exposed to a range of virtual stimuli. Two-photon calcium imaging revealed a small visual area, AF7, that was activated specifically by the optimal prey stimulus. This pretectal region is innervated by two types of retinal ganglion cells, which also send collaterals to the optic tectum. Laser ablation of AF7 markedly reduced prey capture behavior. We identified neurons with arbors in AF7 and found that they projected to multiple sensory and premotor areas: the optic tectum, the nucleus of the medial longitudinal fasciculus (nMLF) and the hindbrain. These findings indicate that computations in the retina give rise to a visual stream which transforms sensory information into a directed prey capture response.

Show MeSH

Related in: MedlinePlus

Prey capture and spontaneous behavior can be classified using five parameters.(A) Each parameter used in the virtual prey SVM is plotted vs the other four. (B) Accuracy of the virtual prey SVM as more parameters are used to classify the bouts as prey capture or spontaneous swims. The dotted line represents the number of parameters used in the final version. The parameters plotted were: 1. Maximum tail curvature, 2. Number of peaks in tail angle, 3. Mean tip angle, 4. Maximum tail angle, 5. Mean tip position. (C). Accuracy of the paramecium SVM plotted vs number of parameters. The parameters shown here were 1. Mean tip angle, 2. Maximum tail curvature, 3. Number of peaks in tail angle, 4. Mean tip position, 5. Mean number of frames between peaks, 6. Maximum tail angle.DOI:http://dx.doi.org/10.7554/eLife.04878.004
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4281881&req=5

fig1s1: Prey capture and spontaneous behavior can be classified using five parameters.(A) Each parameter used in the virtual prey SVM is plotted vs the other four. (B) Accuracy of the virtual prey SVM as more parameters are used to classify the bouts as prey capture or spontaneous swims. The dotted line represents the number of parameters used in the final version. The parameters plotted were: 1. Maximum tail curvature, 2. Number of peaks in tail angle, 3. Mean tip angle, 4. Maximum tail angle, 5. Mean tip position. (C). Accuracy of the paramecium SVM plotted vs number of parameters. The parameters shown here were 1. Mean tip angle, 2. Maximum tail curvature, 3. Number of peaks in tail angle, 4. Mean tip position, 5. Mean number of frames between peaks, 6. Maximum tail angle.DOI:http://dx.doi.org/10.7554/eLife.04878.004

Mentions: (A) Overlay of 50 frames (167 ms) of high-speed video showing examples of behavior in head fixed larvae. Larvae performed forward swims in response to a 3° dot. j-turns were observed when the same dot was to the right or left. Spontaneous swims were often observed in the absence of any stimulus. (B) Example video frame showing points assigned by the digitization algorithm. (C) The position of the tip of the tail over time for the videos in (A). (D) The distribution of tail beat amplitudes for each bout in expert-classified prey capture and spontaneous swim videos. (E) Duration of the longest bend greater than 20° during each bout. (F) Overview of support vector machine (SVM) based bout classification procedure, displaying only two parameters (maximum tail bend and mean tail tip deflection) for clarity. Bouts are extracted using a threshold on the normalized and smoothed first derivative of tail bend angles. Values for each parameter are calculated for all bouts and used to train an SVM. The SVM is then used to classify unlabeled bouts. See Figure 1—figure supplement 1 for plots of each of the five parameters, and accuracy of the SVM vs number of parameters.


A dedicated visual pathway for prey detection in larval zebrafish.

Semmelhack JL, Donovan JC, Thiele TR, Kuehn E, Laurell E, Baier H - Elife (2014)

Prey capture and spontaneous behavior can be classified using five parameters.(A) Each parameter used in the virtual prey SVM is plotted vs the other four. (B) Accuracy of the virtual prey SVM as more parameters are used to classify the bouts as prey capture or spontaneous swims. The dotted line represents the number of parameters used in the final version. The parameters plotted were: 1. Maximum tail curvature, 2. Number of peaks in tail angle, 3. Mean tip angle, 4. Maximum tail angle, 5. Mean tip position. (C). Accuracy of the paramecium SVM plotted vs number of parameters. The parameters shown here were 1. Mean tip angle, 2. Maximum tail curvature, 3. Number of peaks in tail angle, 4. Mean tip position, 5. Mean number of frames between peaks, 6. Maximum tail angle.DOI:http://dx.doi.org/10.7554/eLife.04878.004
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4281881&req=5

fig1s1: Prey capture and spontaneous behavior can be classified using five parameters.(A) Each parameter used in the virtual prey SVM is plotted vs the other four. (B) Accuracy of the virtual prey SVM as more parameters are used to classify the bouts as prey capture or spontaneous swims. The dotted line represents the number of parameters used in the final version. The parameters plotted were: 1. Maximum tail curvature, 2. Number of peaks in tail angle, 3. Mean tip angle, 4. Maximum tail angle, 5. Mean tip position. (C). Accuracy of the paramecium SVM plotted vs number of parameters. The parameters shown here were 1. Mean tip angle, 2. Maximum tail curvature, 3. Number of peaks in tail angle, 4. Mean tip position, 5. Mean number of frames between peaks, 6. Maximum tail angle.DOI:http://dx.doi.org/10.7554/eLife.04878.004
Mentions: (A) Overlay of 50 frames (167 ms) of high-speed video showing examples of behavior in head fixed larvae. Larvae performed forward swims in response to a 3° dot. j-turns were observed when the same dot was to the right or left. Spontaneous swims were often observed in the absence of any stimulus. (B) Example video frame showing points assigned by the digitization algorithm. (C) The position of the tip of the tail over time for the videos in (A). (D) The distribution of tail beat amplitudes for each bout in expert-classified prey capture and spontaneous swim videos. (E) Duration of the longest bend greater than 20° during each bout. (F) Overview of support vector machine (SVM) based bout classification procedure, displaying only two parameters (maximum tail bend and mean tail tip deflection) for clarity. Bouts are extracted using a threshold on the normalized and smoothed first derivative of tail bend angles. Values for each parameter are calculated for all bouts and used to train an SVM. The SVM is then used to classify unlabeled bouts. See Figure 1—figure supplement 1 for plots of each of the five parameters, and accuracy of the SVM vs number of parameters.

Bottom Line: Two-photon calcium imaging revealed a small visual area, AF7, that was activated specifically by the optimal prey stimulus.We identified neurons with arbors in AF7 and found that they projected to multiple sensory and premotor areas: the optic tectum, the nucleus of the medial longitudinal fasciculus (nMLF) and the hindbrain.These findings indicate that computations in the retina give rise to a visual stream which transforms sensory information into a directed prey capture response.

View Article: PubMed Central - PubMed

Affiliation: Department Genes-Circuits-Behavior, Max Planck Institute of Neurobiology, Martinsried, Germany.

ABSTRACT
Zebrafish larvae show characteristic prey capture behavior in response to small moving objects. The neural mechanism used to recognize objects as prey remains largely unknown. We devised a machine learning behavior classification system to quantify hunting kinematics in semi-restrained animals exposed to a range of virtual stimuli. Two-photon calcium imaging revealed a small visual area, AF7, that was activated specifically by the optimal prey stimulus. This pretectal region is innervated by two types of retinal ganglion cells, which also send collaterals to the optic tectum. Laser ablation of AF7 markedly reduced prey capture behavior. We identified neurons with arbors in AF7 and found that they projected to multiple sensory and premotor areas: the optic tectum, the nucleus of the medial longitudinal fasciculus (nMLF) and the hindbrain. These findings indicate that computations in the retina give rise to a visual stream which transforms sensory information into a directed prey capture response.

Show MeSH
Related in: MedlinePlus