Mice recognize 3D objects from recalled 2D pictures, support for picture-object equivalence

Abstract

Picture-object equivalence or recognizing a three-dimensional (3D) object after viewing a two-dimensional (2D) photograph of that object, is a higher-order form of visual cognition that may be beyond the perceptual ability of rodents. Behavioral and neurobiological mechanisms supporting picture-object equivalence are not well understood. We used a modified visual recognition memory task, reminiscent of those used for primates, to test whether picture-object equivalence extends to mice. Mice explored photographs of an object during a sample session, and 24 h later were presented with the actual 3D object from the photograph and a novel 3D object, or the stimuli were once again presented in 2D form. Mice preferentially explored the novel stimulus, indicating recognition of the "familiar" stimulus, regardless of whether the sample photographs depicted radially symmetric or asymmetric, similar, rotated, or abstract objects. Discrimination did not appear to be guided by individual object features or low-level visual stimuli. Inhibition of CA1 neuronal activity in dorsal hippocampus impaired discrimination, reflecting impaired memory of the 2D sample object. Collectively, results from a series of experiments provide strong evidence that picture-object equivalence extends to mice and is hippocampus-dependent, offering important support for the appropriateness of mice for investigating mechanisms of human cognition.

Figure 1

Recognition of a 3D object from a 2D picture is hippocampal dependent regardless of symmetry. (a) Representative infusion sites within the CA1 region of dorsal hippocampus and representative photomicrograph of cannula placement (inset). (b) Each mouse explored two identical novel pictures of a radially symmetric metal leveling foot during sample session (top left). Upon acquiring sample session picture exploration criterion, the mouse was removed and received bilateral intrahippocampal saline or muscimol. During the test session 24 h later (top right), a novel picture replaced one of the familiar pictures. Saline-treated mice explored the novel picture significantly more than the familiar picture during the test session, behavior consistent with visual recognition memory. However, post-sample muscimol-treated mice explored both pictures equivalently, indicating a failure of recognition memory. Discrimination ratio scores were significantly different between the post-sample treatment groups; this difference between the groups was not due to a difference in overall object exploration during the test session. *P < 0.01 versus the respective saline condition. (c) The sample session was conducted as in Fig. 1b. During the test session 24 h later (top right), the 2D familiar pictures were replaced with a 3D “familiar” object (i.e., viewed previously in picture form) and a 3D novel object. Saline-treated mice explored the novel object significantly more than the "familiar" during the 5-min test session; behavior consistent with picture-object correspondence. Muscimol-treated mice explored both objects equivalently, implying that hippocampal inactivation impaired memory for the pictured object, and consequently these mice failed to exhibit test session behavior consistent with picture-object correspondence. A second cohort of saline- and muscimol-treated mice explored the novel object more than the familiar when given a 10-min test session. *P < 0.05 versus respective saline condition. (d) From within a Plexiglas insert, the sample session was conducted as in Fig. 1b&c; however, the stimuli were pictures of a radially asymmetric monkey. The 5-min test session was conducted as in Fig. 1c. Saline-treated mice explored the novel object significantly more than the "familiar" during the test session; behavior consistent with picture-object correspondence. Muscimol-treated mice explored both objects equivalently, implying that hippocampal inactivation impaired memory for the pictured object, and consequently these mice failed to exhibit test session behavior consistent with picture-object correspondence.

Figure 2

Discrimination of an individual object presented in both 2D and 3D forms. It is possible that limitations of the mouse visual system may preclude mice from truly perceiving the difference between an actual 3D object and a 2D picture of that object. This experiment confirmed that the mice identify the 2D pictures and 3D objects as separate entities, as opposed to perceiving them as the same stimulus. During the sample session, naïve mice (n = 9) were placed within the Plexiglas arena insert for a maximum of 10 min where they visually explored two identical 3D objects. During the test session 24 h later, the mice were returned to the insert and allowed 5 min to visually explore the familiar 3D monkey and a 2D picture of the “familiar” monkey. The mice preferentially explored the “familiar” 2D picture over the familiar 3D object [t(8) = − 6.79, P < 0.01, d = 2.13]. This result indicates that rodents identify the 2D “familiar” picture as a stimulus visually distinct from that of the familiar 3D object. *P < 0.01 versus the familiar 3D object.

Figure 3

Recognition of a 3D object from a 2D picture is not affected by low-level visual properties. (a) During the sample session, mice viewed 2D pictures of a stimulus visually similar to the test session novel object. During the test session 24 h later, the mice explored the “familiar” object significantly less than the novel object. Inset, mice exhibited a significant discrimination of the novel object over the “familiar” object. (b) During the sample session, mice viewed 2D pictures of a stimulus visually dissimilar to the test session novel object in configuration only. During the test session 24 h later, the mice explored the “familiar” object significantly less than the novel object. Inset, mice exhibited a significant discrimination of the novel object over the “familiar” object. (c) To account for low-level luminance differences that may be guiding novel object preference, mice viewed 2D pictures of a chess piece (rook or bishop that were either black or white in color) during the sample session. During the test session 24 h later, mice exhibited a significant discrimination of the novel object over the “familiar” object regardless of the color in which the objects were presented. (d) During the sample session, mice viewed 2D pictures of a chess piece (rook or bishop that were either black or white). During the test session 24 h later, mice entered the familiar arena with the ‘familiar’ object in a novel color (i.e., if the mice received a picture of the white rook during sample, then during test, the rook was black) and the novel object in the same color as the sample 2D pictures. Regardless of the color in which the object was represented in the sample pictures, the mice preferentially explored the novel object and there was no difference in discrimination performance as a result of the color in which the stimuli were presented.

Figure 4

Recognition of a 3D object from a 2D picture is not affected by viewing angle. During the sample session, mice viewed pictures depicting the monkey in a rotated or profile view from within the Plexiglas insert. During a 5-min test session 24 h later, mice explored the novel object significantly more than the (non-rotated) “familiar” object. Next, we tested whether extending the duration of the test session would affect the expression of picture-object equivalence in mice. Naïve mice (n = 10) explored pictures of the monkey in a rotated or profile view during the sample session. During a 10-min test session 24 h later, the mice exhibited a significant preference for the novel 3D object over the “familiar” monkey [t(9) = − 5.12, P < 0.01, d = 1.60]. Inset, mice exhibited a significant discrimination of the novel object over the “familiar” object [t(9) = 4.86, P < 0.01, d = 1.54]. Compared to the object discrimination elicited with a 5-min test session, extending the duration of the test session increased object discrimination [t(18) = − 2.54, P = 0.02, d = 1.14]. Thus, this experiment also indicates that mice do not lose their proclivity to explore novel items after short intervals of time. *P < 0.05 versus 5-min test.

Figure 5

Mice rely on composite images for subsequent object recognition. (a) Mice are incapable of matching scrambled images of an object to its actual 3D form or its holistic image. During a sample session, mice explored pictures of an asymmetric object, and a blank picture. During a test session 24 h later, the mice explored both the novel picture and scrambled picture of the "familiar" object significantly more than the familiar picture. This result suggests that mice likely recall the sample session picture as a composite image. (b) Mice explored pictures of a scrambled object from within the Plexiglas insert, then 24 h later, visually explored the “familiar” and the novel objects equivalently. Inset, mice did not discriminate between test objects, indicating that accurate identification of the “familiar” object is only possible when the picture representation is sufficiently similar to the actual object. *P < 0.05 versus familiar picture.

Figure 6

Evidence of picture-object equivalence in mice. (a) During a sample session, mice explored abstract silhouette pictures of a symmetrical object, and 24 h later, explored the novel object significantly more than the “familiar” object. Inset, mice demonstrate object discrimination, indicating that the retrieved memory of the viewed silhouette conveyed enough information to permit recognition of the “familiar” object; furthering support for picture-object equivalence in mice. (b) Mice explored abstract silhouette pictures of a rotated asymmetrical object, and 24 h later, explored the novel object significantly more than the “familiar” object. Inset, mice demonstrate object discrimination, indicating that the retrieved memory of the viewed silhouette conveyed enough information to permit recognition of the “familiar” object; furthering support for picture-object equivalence in mice. *P < 0.05 versus “familiar” object.