FDA Clearance for AI-Assisted Thyroid Ultrasound Analysis Software: A New Era in Radiology

FDA Marketing Clearance for AI “Software”

An important development has highlighted the power of artificial intelligence (AI) in detecting and assessing thyroid nodules. The US Food and Drug Administration (FDA) granted See-Mode Technologies’ AI-based thyroid ultrasound analysis and reporting software 510(k) marketing clearance. This clearance indicates that the software is equivalent to similar products already on the market. The technology stands out as the first FDA-cleared system to automate the detection and diagnosis of thyroid nodules. As AI becomes increasingly important in healthcare, this development may enable faster and more accurate assessment of thyroid nodules.

NOTE: the marketing clearance See-Mode Technologies received from the FDA is for AI software that can be integrated into ultrasound systems, rather than a device. This software is a system that analyzes thyroid ultrasound images and generates automated reports. It detects thyroid nodules, classifies them using the American College of Radiology’s (ACR) TI-RADS system, and then completes the report. After clinical review, the report is sent to radiology reporting systems.

One of the software’s most important features is fully automated nodule detection and characterization, allowing radiologists to review and approve results quickly without manual user input. Its cloud-based design also allows easy integration into existing ultrasound devices without complex hardware installations.

This FDA-cleared software stands out as the first to provide both detection and diagnosis (CADe/x), aiming to speed up ultrasound reports while reducing variability between operators.

• Related topic: What Are Thyroid Nodules and Thyroid Biopsy? Which Nodules Should Be Biopsied?

Thyroid Nodules and Artificial Intelligence: Why Does It Matter?

Thyroid nodules are common masses that are usually benign but can sometimes be malignant (cancerous). Early detection of nodules carrying a cancer risk can significantly affect treatment and outcomes. Conventional ultrasound assessment is time-consuming and can produce inconsistent results between specialists. AI technologies accelerate this process, saving time while providing diagnostic consistency. This new software from See-Mode Technologies enables clinicians to classify thyroid nodules more quickly and accurately.

How Does See-Mode Technologies’ AI-Based System Work?

See-Mode’s software detects nodules in thyroid ultrasound images containing single or multiple nodules and automatically classifies them in line with the American College of Radiology’s (ACR) Thyroid Imaging Reporting and Data System (TI-RADS). The software automatically classifies the detected nodules and prepares a report, which is sent to reporting systems after a physician’s approval. This saves report preparation time while allowing standardization in radiology practice.

• Related topic: Artificial Intelligence in Oncology: What Is Being Done Now and What Lies Ahead?

FDA Marketing Clearance and the MRMC Study

Receiving FDA 510(k) marketing clearance indicates that a product is “equivalent” to an existing medical device and is therefore considered safe and effective. See-Mode’s software was assessed within this framework and received FDA marketing clearance. The multireader, multicase (MRMC) study included in the FDA submission showed that, with AI assistance, radiologists performed better in thyroid nodule detection, characterization, and agreement on ACR TI-RADS levels. The study showed that AI-assisted analyses allowed a more precise distinction between benign (noncancerous) and malignant nodules.

Using AI in Clinical Practice: Time Savings and Consistency

Commenting on the FDA marketing clearance, See-Mode co-founder Sadaf Monajemi said, “AI produced major improvements in radiologists’ nodule localization and agreement on ACR TI-RADS levels. By integrating AI into clinical practice, we aim to reduce reporting times and variability between operators.” The system automatically detects and characterizes thyroid nodules, eliminating manual input. Reviewing clinical findings and completing reports becomes much faster and more efficient for physicians.

Looking Ahead: AI-Assisted Reporting

See-Mode’s AI-based system has the potential to reduce workload and improve the consistency of thyroid ultrasound analyses. Existing CPT (Current Procedural Terminology) codes facilitate consideration of AI use for reimbursement, encouraging wider adoption. However, independent research into the system’s effectiveness could clarify its clinical benefits.

Conclusion: AI-Assisted Assessment of Thyroid Nodules

AI-based technologies in thyroid nodule assessment and management have the potential to reduce radiologists’ workload and improve diagnostic accuracy. See-Mode’s FDA-cleared software may herald a new era in medicine. This advance in medical imaging AI could help physicians make faster, more accurate decisions and help patients achieve better outcomes.