“Artificial intelligence systems are used in clinical practice to assist with medical imaging diagnosis, such as detecting cancers on radiology images.”

“AI is helping to improve the speed, accuracy, and reliability of some cancer screening and detection methods. For example: The Food and Drug Administration has authorized the marketing of AI-based software to help pathologists identify areas of prostate biopsy images that may contain cancer. Medical images such as mammograms can be rapidly processed with the help of AI, allowing radiologists to focus their time on other tasks that require their technical judgement. NCI-supported research has shown that AI imaging algorithms not only improve breast cancer detection on mammography but can also help predict long-term risk of invasive breast cancers.”

The US Food and Drug Administration (FDA) has approved several next-generation AI products with an indication for breast cancer screening (BCS) in recent years. These AI systems are used as adjunct tools in clinical mammography to detect or mark suspicious lesions, support radiologists in their interpretation, and in some cases serve as a second-reader. The review describes multiple commercially available, FDA-approved AI algorithms that have been implemented in routine breast cancer screening practice.

The authors note that radiology is the leading application field of FDA-approved medical AI devices. The outlook for AI deployment in radiology is promising. Large and general foundation and generative AI models, including vision and language models, can impact the clinical adoption of AI tools and may help reduce burnout in radiology. The report explores clinical, cultural, computational, and regulatory considerations essential to adopt AI technology successfully in radiology and emphasizes that AI tools can play a key role in medical imaging if radiologists trust their design and deploy them with adequate training.

This review states that "AI systems for breast imaging, particularly deep learning algorithms for mammography, have moved from research into clinical practice". It notes that commercially available AI tools are "being used as concurrent or second readers to assist radiologists in detecting breast cancers" and that several have received regulatory clearance for use in routine screening.

The paper explains that AI in cancer imaging "may automate processes in the initial interpretation of images" and "assist in radiographic detection, management decisions on whether or not to biopsy or resect, and treatment response assessment". It highlights that AI tools have shown promise in "detecting and characterizing lung nodules, breast lesions, and other malignancies" on radiology images and that some systems have been integrated into clinical workflows.

“AI in medical imaging uses advanced computer algorithms to help radiologists analyze scans with greater speed, accuracy, and precision… Real-World Examples – Faster scans with improved image quality: AI tools now allow doctors to scan their patients up to 75% faster, while also improving image quality… Cancer detection: AI is a powerful tool which helps radiologists find cancer earlier. AI can also precisely measure tumor size, assist in biopsy planning, and track response to treatment across cancers like breast, lung, prostate, and brain.”

This comprehensive review scrutinises the interplay of AI and ML in radiology, exploring their foundational principles, historical progression, practical clinical applications, and challenges. It describes multiple AI tools that have been integrated into clinical workflows, such as algorithms for breast cancer screening, lung nodule detection on CT, and identification of intracranial haemorrhage, which assist radiologists in diagnosis and triage rather than replacing them. The review highlights that clinical adoption is growing but remains uneven across institutions and regions.

Artificial intelligence (AI) applications in diagnostic radiology have demonstrated remarkable accuracy on institutional datasets; however, their performance often degrades when deployed across different populations and scanners. We review evidence from real-world implementations of AI tools for tasks such as cancer detection on mammography and CT, showing that while these systems are used in clinical practice, careful monitoring and local validation are necessary to ensure safe and effective diagnostic support.

“The patient sought a second opinion from a radiologist who does thyroid ultrasound exams using artificial intelligence (AI), which provides a more detailed image and analysis than a traditional ultrasound… While AI is not commonly used in cancer diagnoses, more and more doctors are deploying it to help them determine what might be cancer, predict what might develop into cancer, and devise personalized treatment plans when cancer is found. The Food and Drug Administration has approved AI-assisted tools to help detect cancers of the brain, breast, lung, prostate, skin, and thyroid.”

“In the case of lung cancer, AI-based processing of low-dose CT (LDCT) images can enhance early diagnosis of pulmonary nodules and stratify the risk of malignancy. Multiple AI tools for mammography, chest CT, and brain MRI have received regulatory clearance and are currently deployed in clinical practice to assist radiologists in detecting cancers and other pathologies. Early clinical studies show that AI assistance can increase cancer detection rates and reduce reading times compared with radiologist-only interpretation.”

The main focus of this paper is to examine the existing condition of radiology workflow and identify the challenges hindering the implementation of AI in hospitals and imaging centers. Our systematic review found several community-driven AI tools that have been deployed for clinical use, including systems for chest radiograph triage, lung nodule detection, and mammography decision support. Nevertheless, the overall penetration of AI into routine radiology practice remains limited, with many deployments confined to pilot projects or specific clinical indications.

The U.S. Food and Drug Administration (FDA) has granted De Novo authorization to Clairity Breast, the first-ever AI-powered platform that predicts a woman’s risk of developing breast cancer over the next five years—using only a standard mammogram. Unlike current risk models, Clairity Breast analyzes the mammogram itself using advanced artificial intelligence to detect subtle imaging patterns in breast tissue that correlate with future cancer development. The result is a validated five-year risk score that can guide personalized follow-up care in clinical practice.

A new technology that harnesses AI to analyze mammograms and improve the accuracy of predicting a woman’s personalized five-year risk of developing breast cancer has received Breakthrough Device designation from the Food and Drug Administration (FDA). The system analyzes mammograms to produce a risk score estimating the likelihood that a woman will develop breast cancer over the next five years. The software is a pre-trained machine learning system that analyzes mammogram images and provides an estimate of how likely a patient is to develop breast cancer based solely on images and a woman’s age.

AI-driven radiology workflows can automate image analyses that are manual and time intensive, optimize processes to shortlist for review, and identify potential abnormalities such as cancers or fractures on imaging studies. In clinical sites that have deployed these tools, radiologists use AI outputs as decision-support rather than definitive diagnoses, integrating algorithm findings into their own interpretation to improve efficiency and consistency.

As of late 2025, hundreds of AI-enabled tools have received regulatory clearance for medical imaging tasks, and adoption by clinicians is growing. Breast imaging has seen extensive AI deployment; tools such as Mirai predict long-term breast cancer risk directly from a patient’s mammogram, and several AI lesion detectors automatically highlight suspect masses or calcifications on mammograms to prompt radiologist review. Observational studies suggest that combined human+AI reading slightly outperforms radiologists alone in cancer detection, and these tools have been integrated into many screening programs.

“AI diagnostic tools can exceed 95% accuracy in areas like lung cancer detection and retinal disease screening. High diagnostic accuracy depends on well-curated training data, robust validation, and appropriate clinical integration. In many hospitals and imaging centers, these AI tools are now embedded into PACS workflows, where they flag suspicious lesions, prioritize critical findings, and provide quantitative measurements that radiologists use when diagnosing cancers and other conditions.”

Across peer-reviewed reviews and regulatory databases, there is consistent reporting that AI-enabled tools are actively used in clinical radiology to assist in detecting cancers, especially in breast cancer screening (mammography) and lung cancer screening (CT). However, these tools are typically deployed as decision-support systems that flag or prioritize suspicious findings for radiologists rather than acting as stand-alone diagnostic systems, and adoption is uneven across institutions and countries.