Detecting and Localizing Objects With Color-Contrast With a Second-Generation (44-Channel) Suprachoroidal Retinal Prosthesis.
Аннотация
PURPOSE: To assess the detection of color contrast in retinal prosthesis users by evaluating a novel Detecting Objects with Color-Contrast (DOCC) vision processing algorithm.
METHODS
Three participants monocularly implanted with a suprachoroidal retinal prosthesis (NCT03406416) were asked to locate and identify objects on a tabletop that were color-contrasted with their backgrounds using the DOCC algorithm, a control algorithm (Lanczos2 [L2]) and system off. The task involved two setups; (i) high-contrast white-on-black objects, and (ii) low-grayscale-contrast (color-only contrast) red-on-green objects. Participants were required to verbally localize (nine alternative forced choice), then touch the object, with the distance measured from their finger to the true object location.
RESULTS
For low- (color-only) contrast objects, DOCC (58.3% ± 43.0%) performed better than L2 (27.9% ± 42.5%) for verbally localizing objects (P < 0.0001) and both DOCC (P < 0.0001) and L2 (P = 0.0018) performed better than system off (9.1% ± 33.1%). Furthermore, DOCC (9.7 ± 10.6 cm) had a closer final proximity to the object than L2 (28.0 ± 24.7 cm; P < 0.0001) and system off (45.0 ± 23.9 cm; P < 0.0001). For high-contrast objects, DOCC (59.8% ± 47.8%) performed similarly to L2 (60.8% ± 48.3%) and system off (25.4% ± 41.8%) for verbally localizing objects (P = 0.32, three-condition). Likewise, for final object proximity, DOCC (8.0 ± 7.5 cm) also performed similarly to L2 (6.2 ± 8.4 cm, P = 0.43), whereas both DOCC (P = 0.0014) and L2 (P < 0.0001) performed better than system off (25.6 ± 22.3 cm).
CONCLUSIONS
Vision processing software can improve object detection and localization for objects of one color against a low-contrast background of another color.
TRANSLATIONAL RELEVANCE: A color-contrast algorithm may assist retinal prosthesis users with detecting and localizing objects that are color-contrasted to their background.
Предметные рубрики MeSH
Подготовка и перевод публикации
Обработка статьи · этапы, ошибки и управление
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Тип публикации указан по данным источника. Уровень доказательности автоматически не присваивается.