Automated Visual Evaluation (AVE)
Overview
Automated Visual Evaluation (AVE) is an artificial-intelligence-based cervical screening tool developed by the Clinton Health Access Initiative (CHAI) and Global Health Labs (GHL). The tool is designed to assist healthcare providers in identifying precancerous cervical lesions using a standard smartphone.
AVE is primarily intended for clinicians, gynecologists, and community health workers operating in low- and middle-income countries (LMICs) and resource-limited care settings. It serves as an objective decision aid to augment or replace traditional, subjective Visual Inspection with Acetic Acid (VIA). The tool can be utilized either as a primary screening method or as a triage test for patients who have already screened positive for human papillomavirus (HPV).
In a clinical workflow, the provider performs a standard gynecological exam, applies acetic acid, and captures an image of the cervix using a dedicated smartphone application. The machine learning algorithm processes the image locally on the device, generating an evaluative score in less than one minute. Notable capabilities of AVE include:
- Offline Functionality: The algorithm runs entirely on-device, requiring no active internet connectivity to analyze images.
- Hardware Compatibility: The software is designed to run on low-cost, widely available smartphone models.
- Objective Decision Support: It provides an immediate, automated second opinion to help reduce human subjectivity in visual assessments.
Reviewed by Pouyan Golshani, MD — Interventional Radiologist
Key Features
- AI model for precancerous cervical lesion detection
- Utilizes smartphone images of the cervix
- Outperforms traditional visual inspection with acetic acid (VIA)
- Provides immediate results (within 30-60 seconds)
- Operates offline without internet connectivity
- Intended for use as a triage test for HPV-positive women or primary screening
- High sensitivity (85%) and specificity (86%) for CIN2+ lesions
- Designed for low-resource settings
- Decision aid for health workers
Use Cases
- Cervical cancer screening in low- and middle-income countries (LMICs)
- Triage for women who have screened positive for HPV
- Primary cervical cancer screening where HPV testing is unavailable
- Augmenting the performance of Visual Inspection with Acetic Acid (VIA)
What Physicians Need to Know
AVE acts as a decision aid, providing a rapid 'second opinion' on cervical images. While highly accurate, the healthcare worker remains the ultimate judge of clinical management. The tool is particularly valuable in settings where access to advanced screening methods or specialized professionals is limited, as it enhances the accuracy of visual inspections. Consider its use for primary screening or as a triage for HPV-positive individuals to identify those with a high probability of precancer. Ensure proper training on image capture using smartphones, as algorithm robustness depends on the quality of processed images.
AVE is designed as a smartphone application, allowing for immediate image analysis at the point of care without internet connectivity. Future scaling efforts aim to transfer AVE to widely available and inexpensive smartphone devices. The development of a robust algorithm for smartphones requires training on images produced by various smartphone models. There is also ongoing work to design pre-applications to ensure image adequacy and post-applications to assess treatment options.
Details
| Category | OB/GYN AI, Oncology AI, Radiology & Imaging AI |
| Pricing | Not available publicly |
| Deployment | Smartphone app |
| Compliance | |
| BAA Available | Unknown AI-estimated |
| HIPAA Compliant | Unknown AI-estimated |
| FDA Status | No AI-estimated — The developing partners are identifying pathways to secure regulatory approvals. |
| Integrations | |
| EHR | Not specified |
| Specialties | Obstetrics & Gynecology, Oncology, Radiology |
Ratings & Reviews
No reviews yet. Be the first to review this tool!
Rate Automated Visual Evaluation (AVE)
Videos
Product demos, reviews, and walkthroughs for Automated Visual Evaluation (AVE).
