AISAP, an FDA-cleared artificial intelligence platform designed for point-of-care ultrasound (POCUS), has been highlighted as Case Study #5 in the World Health Organization's latest report, Bridging Theory and Practice (page 21). The report showcases practical implementations of AI in healthcare systems across the globe. Selected from among thousands of AI companies, AISAP was recognized for its collaboration with Sheba Medical Center, one of the world's foremost healthcare institutions.

The featured case study describes AISAP's implementation in Sheba Medical Center's internal medicine and emergency departments. Between July 2022 and December 2023, physicians without specialized ultrasound training used the AI-guided cardiac ultrasound system to assess 660 patients. Findings published in Mayo Clinic Proceedings: Digital Health (2026;4(2):100355) revealed that AI-assisted focused cardiac ultrasound (FoCUS) detected clinically significant cardiac abnormalities in 193 patients (29%), including 147 previously unrecognized conditions (22%).

The diagnostic information led to medication adjustments for 49 patients (7%) and interventional procedures for 9 patients (1.4%). In a subgroup of 117 patients whose results were compared with formal echocardiography, the AI-assisted FoCUS achieved 100% sensitivity and 96.9% specificity in identifying significant left ventricular dysfunction, while demonstrating 100% sensitivity and 87.6% specificity for detecting aortic stenosis.

Commenting on the recognition, Robert Klempfner, Chief Medical Officer and Co-Founder of AISAP, said that the WHO acknowledgment reinforces the clinical experience at Sheba Medical Center, where AI-guided POCUS has enabled clinicians across departments to perform specialist-quality cardiac assessments at the bedside within minutes. He noted that validated AI technology has the potential to make expert-level cardiac diagnostics rapidly accessible throughout the hospital.

According to the WHO report, the prospective clinical study conducted at Sheba showed that more than 30% of patient examinations resulted in treatment modifications, expedited discharge, or transitions in care, while each ultrasound examination was completed in less than five minutes. By the second quarter of 2025, the AISAP platform had been used for over 3,000 examinations across six hospital departments and the emergency department, supported by approximately 150 clinicians who had incorporated the technology into their routine practice.

AISAP's CARDIO V1.0 system received FDA 510(k) clearance (K234141) in September 2024. The clearance was supported by a separate multicenter U.S. clinical study demonstrating sensitivity and specificity exceeding 93% for identifying significant ventricular and valvular abnormalities. The study's findings were independently validated by a core laboratory at Mass General Brigham. In addition, Sheba Medical Center has reported operational outcomes from the broader deployment of the platform across the institution, separate from the peer-reviewed clinical research.

The complete WHO report, Bridging Theory and Practice, provides additional information on AISAP's case study and other examples of AI implementation in healthcare.

The Ever-Changing World of AI in Software Testing
Everywhere you turn, artificial intelligence (AI) is the hot topic of discussion. It’s no surprise that some people are growing tired of the AI buzz. For skeptics, AI seems more like an overhyped spreadsheet than a genuine leap in cognitive technology.

Just when you think you’ve grasped the latest advancements, a new term emerges. First, there was basic machine learning and AI, then Generative AI took center stage, and now Agentic AI is the newest trend. If you feel like you're always trying to catch up, you’re not alone.

Love it or hate it, AI is here to stay. In fact, some modern AI tools can think, create, and learn—just like Keysight’s Eggplant Intelligence.

The Thinking, Creating, and Learning Framework
To make AI easier to understand, it can be broken down into three primary functions:

But how do these AI types differ? What role do they play in software testing? And should you care? The short answer: Yes, they’re significantly different, they have a huge impact, and understanding them is essential.

Before diving into these concepts, let’s explore how AI in software testing has evolved.

The Origins of AI in Software Testing – Keysight Eggplant’s Legacy
The idea of machines exhibiting intelligence and learning like humans dates back to a 1947 lecture by Alan Turing. Since then, AI has come a long way. In 2018, Keysight Eggplant revolutionized the field by integrating AI into its Digital Automation Intelligence (DAI) platform, now known as Eggplant Test. This advancement transformed software testing by enabling:

This is far beyond standard test automation. Imagine modifying a payment gateway on an eCommerce platform—Eggplant automatically generates new test cases, eliminating the need for hours of manual script adjustments. That’s the power of intelligent automation.

But AI’s role in software testing extends beyond executing test cases. Eggplant Test has also pioneered image-based testing, optical character recognition (OCR), and computer vision, crucial for automating graphical user interface (GUI) testing in highly secure environments.

Generative AI – Automating Test Case Creation
Generative AI, which falls under the "Creating" category, focuses on understanding and generating human-like language through natural language processing (NLP) and large language models (LLMs).

This technology can automate test case creation, reducing manual effort while improving accuracy. However, Keysight takes it a step further—once launched, its Generative AI will generate test case frameworks directly from software requirement documents, allowing testers to refine them instead of building from scratch.

Security is a top priority. When Eggplant Test integrates Generative AI, it will operate with secure, offline, and technology-agnostic LLMs, ensuring sensitive data stays protected. Unlike cloud-based solutions, Eggplant's on-premises AI model guarantees data privacy and regulatory compliance.

Cloud-based AI tools that rely on platforms like ChatGPT pose risks, including "shadow prompting," where unmonitored inputs can generate inaccurate results. While prompt engineering can help mitigate this, an on-premises AI solution eliminates the risk entirely.

Agentic AI – The Next Frontier in Software Testing
Agentic AI, the "Thinking" component of the framework, introduces intelligent agents capable of autonomously designing, executing, and optimizing test cases. It leverages chain-of-thought reasoning, a method that stacks multiple commands to handle complex tasks, ensuring every potential user interaction is covered.

One key breakthrough is computer use agents (CUAs), including large action models (LAMs), which automate browser-based processes by interacting with web applications as human testers would. This is crucial for comprehensive end-to-end testing across different devices and browsers.

Another advancement is large vision models (LLaVA), which extend traditional computer vision capabilities to interpret and validate visual elements in applications, ensuring UI accuracy.

If this sounds familiar, it’s because Eggplant Intelligence already integrates AI, Generative AI, and Agentic AI into a unified platform. By optimizing test coverage, automating interactions, and executing tests like a human, Eggplant ensures compliance with AI governance laws in the UK, EU, and US—all while operating offline.

AI Testing Compliance – Why Security Matters
Many cloud-based AI testing tools fail to comply with strict regulations such as the EU AI Act. Industries like aerospace, defense, and healthcare demand high-security standards, making cloud solutions a liability due to data privacy risks.

Keysight Eggplant is the only AI-powered testing platform that prioritizes security, transparency, and governance. Its on-premises approach ensures data remains within secure firewalls, meeting even the most stringent compliance requirements.

Using cloud-based AI for test script generation or reporting isn't just risky—it’s illegal in many jurisdictions. GDPR and other data protection laws prohibit storing customer data outside secure networks, making cloud-based AI tools a non-compliant option for many organizations.
The Future of AI in Software Testing

AI in software testing isn’t just about keeping up with the latest trends—it’s about making strategic, future-proof choices that balance innovation, security, and compliance.

Keysight Eggplant has been at the forefront of AI-driven testing since 2017, long before many of today’s players entered the field. As AI evolves, Eggplant continues to lead the way, ensuring its platform remains the gold standard in secure, AI-powered software testing.

So, if you’re serious about automated testing and need an AI-driven solution that prioritizes security, compliance, and flexibility, it’s time to explore what Keysight Eggplant has to offer.

Click here to start a 14-day free trial.