Publications by Type: Journal Article

2022
Duan, H., Shen, W., Min, X., Tian, Y., Jung, J. - H., Yang, X., & Zhai, G. (2022). Develop then Rival: A Human Vision-Inspired Framework for Superimposed Image Decomposition. IEEE Transactions on Multimedia. Publisher's Version
2021
Avraham, D., Jung, J. - H., Yitzhaky, Y., & Peli, E. (2021). Retinal prosthetic vision simulation: temporal aspects. Journal of Neural Engineering , 18 (4), 0460d9. Publisher's VersionAbstract
Objective. The perception of individuals fitted with retinal prostheses is not fully understood, although several retinal implants have been tested and commercialized. Realistic simulations of perception with retinal implants would be useful for future development and evaluation of such systems. Approach. We implemented a retinal prosthetic vision simulation, including temporal features, which have not been previously simulated. In particular, the simulation included temporal aspects such as persistence and perceptual fading of phosphenes and the electrode activation rate. Main results. The simulated phosphene persistence showed an effective reduction in flickering at low electrode activation rates. Although persistence has a positive effect on static scenes, it smears dynamic scenes. Perceptual fading following continuous stimulation affects prosthetic vision of both static and dynamic scenes by making them disappear completely or partially. However, we showed that perceptual fading of a static stimulus might be countered by head-scanning motions, which together with the persistence revealed the contours of the faded object. We also showed that changing the image polarity may improve simulated prosthetic vision in the presence of persistence and perceptual fading. Significance. Temporal aspects have important roles in prosthetic vision, as illustrated by the simulations. Considering these aspects may improve the future design, the training with, and evaluation of retinal prostheses.
Jung, J. - H., Kurukuti, N. M., & Peli, E. (2021). Photographic Depiction of the Field of View with Spectacles-mounted Low Vision Aids. Optometry and Vision Science , 98 (10), 1210-1226. Publisher's VersionAbstract

SIGNIFICANCE 

Photographic depiction helps to illustrate the primary and secondary field of view effects of low vision devices along with their utility to clinicians, patients, and caretakers. This technique may also be helpful for designers and researchers in improving the design and fitting of low vision devices.

PURPOSE 

The field of view through spectacles-mounted low vision devices has typically been evaluated using perimetry. However, the perimetric field diagram is different from the retinal image and often fails to represent the important aspects of the field of view and visual parameters. We developed a photographic depiction method to record and veridically show the field of view effects of these devices.

METHODS 

We used a 3D-printed holder to place spectacles-mounted devices at the same distance from the empirically determined reference point of the field of view in a camera lens (f = 16 mm) as they would be from an eye, when in use. The field of view effects of a bioptic telescope, a minifier (reverse telescope), and peripheral prisms were captured using a conventional camera, representing retinal images. The human eye pupil size (adjusting the F number: f/2.8 to f/8 and f/22 in the camera lens) and fitting parameters (pantoscopic tilt and back vertex distance) varied.

RESULTS 

Real-world indoor and outdoor walking and driving scenarios were depicted as retinal images illustrating the field of view through low vision devices, distinguishing optical and obscuration scotomas, and demonstrating secondary effects (spatial distortions, viewpoint changes, diplopia, spurious reflection, and multiplexing effects) not illustrated by perimetric field diagrams.

CONCLUSIONS 

Photographic depiction illustrates the primary and secondary field of view effects of the low vision devices. These images highlight the benefit and possible trade-offs of the low vision devices and may be beneficial in education and training.

[PDF]
2020
Peli, E., Vargas-Martin, F., Kurukuti, N. M., & Jung, J. - H. (2020). Multi-periscopic prism device for field expansion. Biomed. Opt. Express , 11 (9), 4872–4889 . OSA. Publisher's VersionAbstract
Patients with visual field loss frequently collide with other pedestrians, with the highest risk being from pedestrians at a bearing angle of 45&\#x00B0;. Current prismatic field expansion devices (&\#x2248;30&\#x00B0;) cannot cover pedestrians posing the highest risk and are limited by poor image quality and restricted eye scanning range (<5&\#x00B0;). A new field expansion device: multi-periscopic prism (MPP); comprising a cascade of half-penta prisms provides wider shifting power (45&\#x00B0;) with dramatically better image quality and wider eye scanning range (15&\#x00B0;) is presented. Spectacles-mounted MPPs were implemented using 3D printing. The efficacy of the MPP is demonstrated using perimetry, photographic depiction, and analyses of the collision risk covered by the devices.
[PDF]
Jung, J. - H., Castle, R., Kurukuti, N. M., Manda, S., & Peli, E. (2020). Field Expansion with Multiplexing Prism Glasses Improves Pedestrian Detection for Acquired Monocular Vision. Translational Vision Science & Technology , 9 (8), Article 35. Publisher's VersionAbstract
Patients with acquired monocular vision (AMV) lose vision in the temporal crescent on the side of the blind eye. This visual field loss affects patients’ ability to detect potential hazards in the blind field. Mounting a base-in multiplexing prism (MxP) on the nasal side of the seeing eye can provide true field expansion and enable detection of potential collision hazards. We evaluated the efficacy of the MxP glasses in a virtual reality walking environment. A three-dimensional printed clip-on MxP holder that can be adjusted for an individual user's facial parameters was developed. Virtual reality walking scenarios were designed to evaluate the effect of MxP field expansion on the detection of a pedestrian approaching from different initial bearing angles and courses. The pedestrian detection rates and response times of 10 participants with simulated AMV (normally sighted participants with one eye patched) and three patients with AMV were measured. The MxP provided true field expansion of about 25°. Participants performed significantly better with the MxP than without the MxP in the pedestrian detection task on their blind field, while their seeing field performance was not significantly different. The MxP glasses for patients with AMV improved the detection of potential collision hazards in the blind field. The MxP with an adjustable clip-on holder may help patients with AMV to decrease the risk of collision with other pedestrians.
jung_2020_tvst.xlsx [PDF]
Pamir, Z., Jung, J. - H., & Peli, E. (2020). Preparing participants for the use of the tongue visual sensory substitution device. Disability and Rehabilitation: Assistive Technology . Publisher's Version
Pamir, Z., Canoluk, M. U., Jung, J. - H., & Peli, E. (2020). Poor resolution at the back of the tongue is the bottleneck for spatial pattern recognition. Scientific Reports , 10 (1), 2435. Publisher's VersionAbstract
Spatial patterns presented on the tongue using electro-tactile sensory substitution devices (SSDs) have been suggested to be recognized better by tracing the pattern with the tip of the tongue. We examined if the functional benefit of tracing is overcoming the poor sensitivity or low spatial resolution at the back of the tongue or alternatively compensating for limited information processing capacity by fixating on a segment of the spatial pattern at a time. Using a commercially available SSD, the BrainPort, we compared letter recognition performance in three presentation modes; tracing, static, and drawing. Stimulation intensity was either constant or increased from the tip to the back of the tongue to partially compensate for the decreasing sensitivity. Recognition was significantly better for tracing, compared to static and drawing conditions. Confusion analyses showed that letters were confused based on their characteristics presented near the tip in static and drawing conditions. The results suggest that recognition performance is limited by the poor spatial resolution at the back of the tongue, and tracing seems to be an effective strategy to overcome this. Compensating for limited information processing capacity or poor sensitivity by drawing or increasing intensity at the back, respectively, does not improve the performance.
pamir-etal-2020-scirep-tracingstudy-data_3.xlsx [PDF]
Choi, H. - J., Peli, E., Park, M., & Jung, J. - H. (2020). Design of 45∘ periscopic visual field expansion device for peripheral field loss. Optics Communications , 454, 124364. Publisher's VersionAbstract
Patients with visual field loss have difficulty in mobility due to collision with pedestrians/obstacles from the blind side. In order to compensate for the visual field loss, prisms which deflect the field from the blind to the seeing side, have been widely used. However, the deflection power of current clinical Fresnel prisms is limited to ∼30°and only allows a 5°eye scanning range to the blind side. This is not sufficient to avoid most collisions and results in demands for a device with a higher power. In this paper, we propose a novel design and optimization of a higher power prism-like device (cascaded structure of mirror pairs filled with refractive medium) and verify enhanced field expansion of up to 45°in optical ray tracing and photorealistic simulations.
2019
Jung, J. - H., Castle, R., Kurukuti, N. M., Manda, S., & Peli, E. (2019). Improved Collision Detection with Multiplexing Prisms by Patients with Acquired Monocular Vision. Optometry and Vision Science , 96, E-abstract 190018.
Pamir, Z., Canoluk, M. U., Jung, J. - H., & Peli, E. (2019). The variable spatial resolution is compensated by tip of the tongue tracing in pattern recognition. Investigative Ophthalmology & Visual Science , 60 (9), 3310-3310.Abstract
The tongue has been suggested to be an ideal platform for visual sensory substitution devices because of its high spatial resolution and high sensitivity. However, inhomogeneities in resolution and sensitivity across its dorsal surface have been reported. In the present study, using the BrainPort Vision Pro, we investigated how the inhomogeneity in the spatial resolution affects letter recognition performance of normally sighted and blind individuals. Stimuli were 26 computer-generated English alphabet capital letters streamed directly to the intra-oral device (IOD), not through the camera. Stimulation intensity was adjusted to be above threshold at the lowest resolution (back of the tongue). We compared three display conditions; static, drawing, and tracing with the tip of the tongue. In the static condition, the complete letter was displayed on the tongue at once and participants were not allowed to move their tongue. In the drawing condition, the letter was divided into smaller parts and sequentially presented as in the case of hand lettering. In the tracing condition, the letter is presented at once, as in the static, but participants were encouraged to trace the pattern with their tip of the tongue. In all conditions, participants were required to name the letter. Data was collected from 10 participants. Recognition performance was significantly better for the tracing condition (M=67%, SEM=8.54) compared to both static (M=38%, SEM =5.07) and drawing (M=33%, SEM =5.93) conditions. There was no difference between the static and drawing conditions. The performance was well above chance (3.84%) for all conditions. There was no significant difference in time required to recognize a letter across conditions, and between normally sighted and blind participants’ performance. Tracing the spatial pattern with the tip of tongue provides far better performance in recognizing letters. The results suggest that inhomogeneity in spatial resolution reduces the recognition performance through aliasing when the spatial pattern is presented across the inhomogenous tongue surface; and the higher resolution at the tip enables motor tracing for recognition. This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019. Individual (N: Normally sighted, B: Blind) and mean percent correct responses for letter recognition in static, drawing and tracing conditions. The horizontal red line shows chance level (1/26). Error bars show SEM. * p < 0.001.
Jung, J. - H., Castle, R., Kurukuti, N. M., Manda, S., & Peli, E. (2019). Evaluation of Field Expansion for Acquired Monocular Vision with Multiplexing Prisms. Investigative Ophthalmology & Visual Science , 60 (9), 4026-4026.Abstract
Individuals with acquired monocular vision (AMV) report collisions with pedestrians appearing in the temporal crescent of the blind eye. Our previous analyses show that loss of nasal field beyond ~55° may result in missing ~30% of possible collisions with other pedestrians. We developed a 3D-printed clip-on multiplexing prism (MxP) holder for AMV that expands nasal field ~30° into the missing temporal crescent. We evaluated pedestrian detection performance with the MxP in AMV using a virtual walking simulator. AMV was simulated with an eye patch in normal vision participants (N=8). Each subject was fitted with MxP in the 3D-printed clip-on holder. Visual fields were measured using the Goldmann perimeter for each viewing condition (normal, simulated AMV, and simulated AMV with MxP). Virtual walking scenarios were developed in a driving simulator, where pedestrians walked towards collision with the subject (walking speed 1.4m/s) from the left or right side with initial bearing angles of 45°, 65°, or 80°. Subjects fixated a target in the center of the screen and were to press the horn when they saw a pedestrian (48 trials per viewing condition). Detection rates and response times (between pedestrian appearance and horn press) were analyzed using a 3 (viewing condition)×6 (initial bearing) two-way repeated measures Analysis of Variance (ANOVA). The nasal field in simulated AMV expanded from 55.9°±1.5 to 81.3°±0.9 with the MxP. The ANOVA showed significant interaction between viewing condition and initial bearing for detection rate and response time (all p<0.001). Detection rate and response time on the blind nasal side of simulated AMV (61% with 4.1s at 65°; 50% with 5.1s at 80°) improved significantly (p<0.001) with MxP (94% with 1.8s at 65°; 64% with 4.1s at 80°) but was worse than normal vision (100% with 1.1s in 65°; 98% with 1.8s in 80°). Visual confusion caused by the MxP (superimposing seeing and shifted fields) did not interfere with detection performance at 45° since there were no significant differences in performance (all p=0.393) with normal vision (100% with 1.1s), simulated AMV (95% with 1.4s), and simulated AMV with MxP (100% with 1.2s). The 3D-printed clip-on MxP holder allowed for individualized prism fitting, which expanded visual field in simulated AMV and significantly improved pedestrian detection on the blind side with minimal interference in the seeing side. This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.
2018
Jung, J. - H., & Peli, E. (2018). No Useful Field Expansion with Full-field Prisms. Optometry and Vision Science , 95 (9), 805-813. Publisher's VersionAbstract
SIGNIFICANCEFull-field prisms that fill the entire spectacle eye wire have been considered as field expansion devices for homonymous hemianopia (HH) and acquired monocular vision (AMV). Although the full-field prism is used for addressing binocular dysfunction and for prism adaptation training after brain injury as treatment for spatial hemineglect, we show that the full-field prism for field expansion does not effectively expand the visual field in either HH or AMV.PURPOSEFull-field prisms may shift a portion of the blind side to the residual seeing side. However, foveal fixation on an object of interest through a full-field prism requires head and/or eye rotation away from the blind side, thus negating the shift of the field toward the blind side.METHODSWe fit meniscus and flat full-field 7Δ and 12Δ yoked prisms and conducted Goldmann perimetry in HH and AMV. We compared the perimetry results with ray tracing calculations.RESULTSThe rated prism power was in effect at the primary position of gaze for all prisms, and the meniscus prisms maintained almost constant power at all eccentricities. To fixate on the perimetry target, the subjects needed to turn their head and/or eyes away from the blind side, which negated the field shift into the blind side. In HH, there was no difference in the perimetry results on the blind side with any of the prisms. In AMV, the lower nasal field of view was slightly shifted into the blind side with the flat prisms, but not with the meniscus prisms.CONCLUSIONSFull-field prisms are not an effective field expansion device owing to the inevitable fixation shift. There is potential for a small field shift with the flat full-field prism in AMV, but such lenses cannot incorporate refractive correction. Furthermore, in considering the apical scotoma, the shift provides a mere field substitution at best.
datasharing_fullfield_1.xlsx
Jung, J. - H., & Peli, E. (2018). Field Expansion for Acquired Monocular Vision Using a Multiplexing Prism. Optometry and Vision Science , 95 (9), 814-828. Publisher's VersionAbstract
SIGNIFICANCEAcquired monocular vision (AMV) is a common visual field loss. Patients report mobility difficulties in walking due to collisions with objects or other pedestrians on the blind side.PURPOSEThe visual field of people with AMV extends more than 90° temporally on the side of the seeing eye but is restricted to approximately 55° nasally. We developed a novel field expansion device using a multiplexing prism (MxP) that superimposes the see-through and shifted views for true field expansion without apical scotoma. We present various designs of the device that enable customized fitting and improved cosmetics.METHODSA partial MxP segment is attached (base-in) near the nose bridge. To avoid total internal reflection due to the high angle of incidence at nasal field end (55°), we fit the MxP with serrations facing the eye and tilt the prism base toward the nose. We calculated the width of the MxP (the apex location) needed to prevent apical scotoma and monocular diplopia. We also consider the effect of spectacle prescriptions on these settings. The results are verified perimetrically.RESULTSWe documented the effectivity of various prototype glasses designs with perimetric measurements. With the prototypes, all patients with AMV had field-of-view expansions up to 90° nasally without any loss of seeing field.CONCLUSIONSThe novel and properly mounted MxP in glasses has the potential for meaningful field-of-view expansion up to the size of normal binocular vision in cosmetically acceptable form.
datasharing_amv_mxp_4.xlsx
Han, S. 'E., Qiu, C., Lee, K. R., Jung, J. - H., & Peli, E. (2018). Word recognition: re-thinking prosthetic vision evaluation. Journal of Neural Engineering , 15 (5), 055003 . IOP Publishing. Publisher's VersionAbstract
Objective. Evaluations of vision prostheses and sensory substitution devices have frequently relied on repeated training and then testing with the same small set of items. These multiple forced-choice tasks produced above chance performance in blind users, but it is unclear if the observed performance represents restoration of vision that transfers to novel, untrained items. Approach. Here, we tested the generalizability of the forced-choice paradigm on discrimination of low-resolution word images. Extensive visual training was conducted with the same 10 words used in previous BrainPort tongue stimulation studies. The performance on these 10 words and an additional 50 words was measured before and after the training sessions. Main results. The results revealed minimal performance improvement with the untrained words, demonstrating instead pattern discrimination limited mostly to the trained words. Significance. These findings highlight the need to reconsider current evaluation practices, in particular, the use of forced-choice paradigms with a few highly trained items. While appropriate for measuring the performance thresholds in acuity or contrast sensitivity of a functioning visual system, performance on such tasks cannot be taken to indicate restored spatial pattern vision.
wordrecognition2018datasharing_1.xlsx
Qiu, C., Jung, J. - H., Tuccar-Burak, M., Spano, L., Goldstein, R., & Peli, E. (2018). Measuring Pedestrian Collision Detection With Peripheral Field Loss and the Impact of Peripheral Prisms. Translational Vision Science & Technology , 7 (5), 1-1. Publisher's VersionAbstract
Peripheral field loss (PFL) due to retinitis pigmentosa, choroideremia, or glaucoma often results in a highly constricted residual central field, which makes it difficult for patients to avoid collision with approaching pedestrians. We developed a virtual environment to evaluate the ability of patients to detect pedestrians and judge potential collisions. We validated the system with both PFL patients and normally sighted subjects with simulated PFL. We also tested whether properly placed high-power prisms may improve pedestrian detection. A virtual park-like open space was rendered using a driving simulator (configured for walking speeds), and pedestrians in testing scenarios appeared within and outside the residual central field. Nine normally sighted subjects and eight PFL patients performed the pedestrian detection and collision judgment tasks. The performance of the subjects with simulated PFL was further evaluated with field of view expanding prisms. The virtual system for testing pedestrian detection and collision judgment was validated. The performance of PFL patients and normally sighted subjects with simulated PFL were similar. The prisms for simulated PFL improved detection rates, reduced detection response times, and supported reasonable collision judgments in the prism-expanded field; detections and collision judgments in the residual central field were not influenced negatively by the prisms. The scenarios in a virtual environment are suitable for evaluating PFL and the impact of field of view expanding devices. This study validated an objective means to evaluate field expansion devices in reproducible near-real-life settings.
[PDF]
Qiu, C., Lee, K. R., Jung, J. - H., Goldstein, R., & Peli, E. (2018). Motion Parallax Improves Object Recognition in the Presence of Clutter in Simulated Prosthetic Vision. Translational Vision Science & Technology , 7 (5), 29-29. Publisher's VersionAbstract
Efficacy of current visual prostheses in object recognition is limited. Among various limitations to be addressed, such as low resolution and low dynamic range, here we focus on reducing the impact of background clutter on object recognition. We have proposed the use of motion parallax via head-mounted camera lateral scanning and computationally stabilizing the object of interest (OI) to support neural background decluttering. Simulations in head-mounted displays (HMD), mimicking the proposed effect, were used to test object recognition in normally sighted subjects. Images (24° field of view) were captured from multiple viewpoints and presented at a low resolution (20 × 20). All viewpoints were centered on the OI. Experimental conditions (2 × 3) included clutter (with or without) × head scanning (single viewpoint, 9 coherent viewpoints corresponding to subjects' head positions, and 9 randomly associated viewpoints). Subjects used lateral head movements to view OIs in the HMD. Each object was displayed only once for each subject. The median recognition rate without clutter was 40% for all head scanning conditions. Performance with synthetic background clutter dropped to 10% in the static condition, but it was improved to 20% with the coherent and random head scanning (corrected P = 0.005 and P = 0.049, respectively). Background decluttering using motion parallax cues but not the coherent multiple views of the OI improved object recognition in low-resolution images. The improvement did not fully eliminate the impact of background. Motion parallax is an effective but incomplete decluttering solution for object recognition with visual prostheses.
motion_parallax_data_sharing_2.xlsx [PDF]
2017
Peli, E., & Jung, J. - H. (2017). Multiplexing Prisms for Field Expansion. Optometry and Vision Science , 94 (8), 817-829. Publisher's VersionAbstract
PURPOSEPrisms used for field expansion are limited by the optical scotoma at a prism apex (apical scotoma). For a patient with two functioning eyes, fitting prisms unilaterally allows the other eye to compensate for the apical scotoma. A monocular patient’s field loss cannot be expanded with a conventional or Fresnel prism because of the apical scotoma. A newly invented optical device, the multiplexing prism (MxP), was developed to overcome the apical scotoma limitation in monocular field expansion.METHODSA Fresnel–prism–like device with alternating prism and flat elements superimposes shifted and see-through views, thus creating the (monocular) visual confusion required for field expansion and eliminating the apical scotoma. Several implementations are demonstrated and preliminarily evaluated for different monocular conditions with visual field loss. The field expansion of the MxP is compared with the effect of conventional prisms using calculated and measured perimetry.RESULTSField expansion without apical scotomas is shown to be effective for monocular patients with hemianopia or constricted peripheral field. The MxPs are shown to increase the nasal field for a patient with only one eye and for patients with bitemporal hemianopia. The MxPs placed at the far temporal field are shown to expand the normal visual field. The ability to control the contrast ratio between the two images is verified.CONCLUSIONSA novel optical device is demonstrated to have the potential for field expansion technology in a variety of conditions. The devices may be inexpensive and can be constructed in a cosmetically acceptable format.
datasharing_mxp_fov_3.xlsx
2016
Jung, J. - H., Pu, T., & Peli, E. (2016). Comparing object recognition from binary and bipolar edge images for visual prostheses. Journal of Electronic Imaging , 25 (6), 061619. Publisher's VersionAbstract

Visual prostheses require an effective representation method due to the limited display condition which has only 2 or 3 levels of grayscale in low resolution. Edges derived from abrupt luminance changes in images carry essential information for object recognition. Typical binary (black and white) edge images have been used to represent features to convey essential information. However, in scenes with a complex cluttered background, the recognition rate of the binary edge images by human observers is limited and additional information is required. The polarity of edges and cusps (black or white features on a gray background) carries important additional information; the polarity may provide shape from shading information missing in the binary edge image. This depth information may be restored by using bipolar edges. We compared object recognition rates from 16 binary edge images and bipolar edge images by 26 subjects to determine the possible impact of bipolar filtering in visual prostheses with 3 or more levels of grayscale. Recognition rates were higher with bipolar edge images and the improvement was significant in scenes with complex backgrounds. The results also suggest that erroneous shape from shading interpretation of bipolar edges resulting from pigment rather than boundaries of shape may confound the recognition.

bipolar_filtering_dataset_1.zip
Peli, E., Bowers, A., Keeney, K., & Jung, J. - H. (2016). High Power Prismatic Devices for Oblique Peripheral Prisms. Optometry and Vision Science , 93 (5). data_sharing_high_power_3.xlsx
2015
Jung, J. - H., Aloni, D., Yitzhaky, Y., & Peli, E. (2015). Active Confocal Imaging for Visual Prostheses. Vision Research , 111 (June), 182-196. Publisher's VersionAbstract

There are encouraging advances in prosthetic vision for the blind, including retinal and cortical implants, and other “sensory substitution devices” that use tactile or electrical stimulation. However, they all have low resolution, limited visual field, and can display only few gray levels (limited dynamic range), severely restricting their utility. To overcome these limitations, image processing or the imaging system could emphasize objects of interest and suppress the background clutter. We propose an active confocal imaging system based on light-field technology that will enable a blind user of any visual prosthesis to efficiently scan, focus on, and “see” only an object of interest while suppressing interference from background clutter. The system captures three-dimensional scene information using a light-field sensor and displays only an in-focused plane with objects in it. After capturing a confocal image, a de-cluttering process removes the clutter based on blur difference. In preliminary experiments we verified the positive impact of confocal-based background clutter removal on recognition of objects in low resolution and limited dynamic range simulated phosphene images. Using a custom-made multiple-camera system, we confirmed that the concept of a confocal de-cluttered image can be realized effectively using light field imaging.

[PDF] activeconfocalimaging_datasharing_2.xlsx
Best paper of the year by a trainee in SERI, Harvard Medical School

Pages