Dados do Trabalho

Título

The impact of AI-assisted interpretation on breast ultrasound: current and emerging clinical applications

Introdução e objetivo(s)

Women with dense breast tissue often require an alternative to mammography for diagnosis. Ultrasound (US) is a widely available and effective complementary method.
Artificial intelligence (AI) algorithms have been developed to assist in the interpretation of breast US scans. These algorithms analyze the images captured by the hand held device and provide automated analysis that allows physicians to better detect and characterize breast abnormalities. The original AI-based interpretation can aid in the detection of suspicious lesions, malignancy risk calculation and clinical management decision making.
Our aim is to highlight both current and emerging clinical applications of AI in hand-held US breast examinations.
To illustrate the impact of original and artificially enhanced AI on the interpretation of breast US.

Método(s)

Retrospective pictorial essay of clinical cases selected from our digital archive of ultrasound breast lesions. Description of clinical AI applications . Bibliographic review

Discussão

The impact of original AI-based interpretation lies in its potential to improve diagnostic accuracy. AI applications can improve the consistency of breast cancer treatment recommendations by reducing intra- and inter-observer variability.
The findings highlight that issues related to user confidence in AI need to be considered in the development and implementation of AI, as well as in the training of radiologists, as the interaction between humans and AI will ultimately influence the impact of AI on patient care. This can lead to earlier intervention, better patient outcomes and potentially lower healthcare costs.
The increasing use of AI in radiology has raised concerns in clinical applications such as the detection, characterization and classification of breast lesions. Emerging challenges relate to prediction of tumor biology and molecular subtypes of breast cancer, prediction of axillary nodal metastases, and AI - breast US in low resource settings.

Conclusões

AI-based detection has the potential to play an important clinical role in handheld breast US. There is evidence that AI breast ultrasound could soon be used in clinical practice to detect, characterize, and classify breast lesions and determine prognosis.
Radiologists and healthcare professionals play a critical role in reviewing and validating AI-generated findings to make informed decisions for patient care.

Palavras Chave

Artificial intelligence; breast neoplasms; Ultrasound

Arquivos