Development of an embedded diagnostic tool for visual misalignment screening
This article presents the design, implementation, and validation of a low-cost embedded system for preliminary strabismus screening, based on computer vision and deep learning. The hardware integrates a Raspberry Pi 4, a USB camera, and a 3D-printed chin rest to ensure consistent facial positioning....
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2025-09-01
|
| Series: | HardwareX |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2468067225000707 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849226947778838528 |
|---|---|
| author | Daniel Soto Rodriguez Andres Eduardo Rivera Gomez Ruthber Rodriguez Serrezuela |
| author_facet | Daniel Soto Rodriguez Andres Eduardo Rivera Gomez Ruthber Rodriguez Serrezuela |
| author_sort | Daniel Soto Rodriguez |
| collection | DOAJ |
| description | This article presents the design, implementation, and validation of a low-cost embedded system for preliminary strabismus screening, based on computer vision and deep learning. The hardware integrates a Raspberry Pi 4, a USB camera, and a 3D-printed chin rest to ensure consistent facial positioning. The software, developed in Python using PyQt5 and OpenCV, incorporates a NASNetLarge convolutional neural network converted to TensorFlow Lite for real-time inference. The graphical interface allows users to capture or upload images, perform automated analysis, generate diagnostic PDF reports, and access a gamified treatment module. Functional validation included a proprietary dataset of 27 images, achieving a 96.30 % classification accuracy. Additionally, a stratified 10-fold cross-validation on a balanced dataset of 1000 images yielded an average accuracy of 95.6 % with strong generalization metrics (F1-score, precision, and recall above 94 %). A novel treatment validation mechanism was implemented by analyzing pupil-to-stimulus distance frame-by-frame, confirming reliable eye tracking and the system’s potential for detecting microstrabismus. This open-source, portable prototype is suitable for community health screening and educational use, particularly in low-resource settings. |
| format | Article |
| id | doaj-art-93094875b6bf4bc18e752bef9d7f7b60 |
| institution | Kabale University |
| issn | 2468-0672 |
| language | English |
| publishDate | 2025-09-01 |
| publisher | Elsevier |
| record_format | Article |
| series | HardwareX |
| spelling | doaj-art-93094875b6bf4bc18e752bef9d7f7b602025-08-24T05:13:52ZengElsevierHardwareX2468-06722025-09-0123e0069210.1016/j.ohx.2025.e00692Development of an embedded diagnostic tool for visual misalignment screeningDaniel Soto Rodriguez0Andres Eduardo Rivera Gomez1Ruthber Rodriguez Serrezuela2Facultad de ingeniería, ingeniería de software, universidad surcolombiana, Neiva, Huila, ColombiaMaster en inteligencia artificial, Universidad de la Rioja, Logroño, España; Corresponding author.Facultad de ingeniería, ingeniería mecatrónica, Corporación universitaria del huila “corhuila”, Neiva, Huila, ColombiaThis article presents the design, implementation, and validation of a low-cost embedded system for preliminary strabismus screening, based on computer vision and deep learning. The hardware integrates a Raspberry Pi 4, a USB camera, and a 3D-printed chin rest to ensure consistent facial positioning. The software, developed in Python using PyQt5 and OpenCV, incorporates a NASNetLarge convolutional neural network converted to TensorFlow Lite for real-time inference. The graphical interface allows users to capture or upload images, perform automated analysis, generate diagnostic PDF reports, and access a gamified treatment module. Functional validation included a proprietary dataset of 27 images, achieving a 96.30 % classification accuracy. Additionally, a stratified 10-fold cross-validation on a balanced dataset of 1000 images yielded an average accuracy of 95.6 % with strong generalization metrics (F1-score, precision, and recall above 94 %). A novel treatment validation mechanism was implemented by analyzing pupil-to-stimulus distance frame-by-frame, confirming reliable eye tracking and the system’s potential for detecting microstrabismus. This open-source, portable prototype is suitable for community health screening and educational use, particularly in low-resource settings.http://www.sciencedirect.com/science/article/pii/S2468067225000707Strabismus detectionMedical AIConvolutional neural networks (CNN)Low-cost diagnostic device |
| spellingShingle | Daniel Soto Rodriguez Andres Eduardo Rivera Gomez Ruthber Rodriguez Serrezuela Development of an embedded diagnostic tool for visual misalignment screening HardwareX Strabismus detection Medical AI Convolutional neural networks (CNN) Low-cost diagnostic device |
| title | Development of an embedded diagnostic tool for visual misalignment screening |
| title_full | Development of an embedded diagnostic tool for visual misalignment screening |
| title_fullStr | Development of an embedded diagnostic tool for visual misalignment screening |
| title_full_unstemmed | Development of an embedded diagnostic tool for visual misalignment screening |
| title_short | Development of an embedded diagnostic tool for visual misalignment screening |
| title_sort | development of an embedded diagnostic tool for visual misalignment screening |
| topic | Strabismus detection Medical AI Convolutional neural networks (CNN) Low-cost diagnostic device |
| url | http://www.sciencedirect.com/science/article/pii/S2468067225000707 |
| work_keys_str_mv | AT danielsotorodriguez developmentofanembeddeddiagnostictoolforvisualmisalignmentscreening AT andreseduardoriveragomez developmentofanembeddeddiagnostictoolforvisualmisalignmentscreening AT ruthberrodriguezserrezuela developmentofanembeddeddiagnostictoolforvisualmisalignmentscreening |