Khosravi, Bardia

Agentic artificial intelligence marks a shift from large language models to autonomous, goal-driven clinical assistants. The framework presented in this report outlines implementation in radiology workflows and addresses cybersecurity, bias, and accountability challenges.

Large language models pose unique cybersecurity risks in health care, including vulnerability to malicious attacks and data breaches. This article equips professionals with strategies to mitigate these threats for safe use.

Abstract Background Automated data extraction from echocardiography reports could facilitate large-scale registry creation and clinical surveillance of valvular heart diseases (VHD). We evaluated the performance of open-source large language models (LLMs) guided by prompt instructions and chain of thought (CoT) for this task. Methods From consecutive transthoracic echocardiographies performed in our center, we utilized 200 random reports from 2019 for prompt optimization and 1000 from 2023 for evaluation. Five instruction-tuned LLMs (Qwen2.0-72B, Llama3.0-70B, Mixtral8-46.7B, Llama3.0-8B, and Phi3.0-3.8B) were guided by prompt instructions with and without CoT to classify prosthetic valve presence and VHD severity. Performance was evaluated using classification metrics against expert-labeled ground truth. Mean squared error (MSE) was also calculated for predicted severity’s deviation from actual severity. Results With CoT prompting, Llama3.0-70B and Qwen2.0 achieved the highest performance (accuracy: 99.1% and 98.9% for VHD severity; 100% and 99.9% for prosthetic valve; MSE: 0.02 and 0.05, respectively). Smaller models showed lower accuracy for VHD severity (54.1%-85.9%) but maintained high accuracy for prosthetic valve detection (>96%). Chain of thought reasoning yielded higher accuracy for larger models while increasing processing time from 2-25 to 67-154 seconds per report. Based on CoT reasonings, the wrong predictions were mainly due to model outputs being influenced by irrelevant information in the text or failure to follow the prompt instructions. Conclusions Our study demonstrates the near-perfect performance of open-source LLMs for automated echocardiography report interpretation with the purpose of registry formation and disease surveillance. While larger models achieved exceptional accuracy through prompt optimization, practical implementation requires balancing performance with computational efficiency.

Although generative artificial intelligence can create podcast summaries of radiology articles, summaries are limited by hallucinations and lack of audience appropriateness, requiring further investigation to justify use as an educational tool.

Abstract Deep learning is revolutionizing medical imaging analysis by enabling the classification of various pathoanatomical conditions at scale. Unfortunately, there have been a limited number of accurate and efficient machine learning (ML) algorithms that have been developed for the diagnostic workup of morphological hip pathologies, including developmental dysplasia of the hip and femoroacetabular impingement. The current study reports on the performance of a novel ML model with YOLOv5 and ConvNeXt-Tiny architecture in predicting the morphological features of these conditions, including cam deformity, ischial spine sign, dysplastic appearance, and other abnormalities. The model achieved 78.0% accuracy for detecting cam deformity, 87.2% for ischial spine sign, 76.6% for dysplasia, and 71.6% for all abnormalities combined. The model achieved an Area under the Receiver Operating Curve of 0.89 for ischial spine sign, 0.80 for cam deformity, 0.80 for dysplasia, and 0.81 for all abnormalities combined. Inter-rater agreement among surgeons, assessed using Gwet’s AC1, was substantial for dysplasia (0.83) and all abnormalities (0.88), and moderate for ischial spine sign (0.75) and cam deformity (0.61).

Thyroid Ultrasound (US) is the primary method to evaluate thyroid nodules. Deep learning (DL) has been playing a significant role in evaluating thyroid cancer. We propose a DL-based pipeline to detect and classify thyroid nodules into benign or malignant groups relying on two views of US imaging. Transverse and longitudinal US images of thyroid nodules from 983 patients were collected retrospectively. Eighty-one cases were held out as a testing set, and the rest of the data were used in five-fold cross-validation (CV). Two You Look Only Once (YOLO) v5 models were trained to detect nodules and classify them. For each view, five models were developed during the CV, which was ensembled by using non-max suppression (NMS) to boost their collective generalizability. An extreme gradient boosting (XGBoost) model was trained on the outputs of the ensembled models for both views to yield a final prediction of malignancy for each nodule. The test set was evaluated by an expert radiologist using the American College of Radiology Thyroid Imaging Reporting and Data System (ACR-TIRADS). The ensemble models for each view achieved a mAP0.5 of 0.797 (transverse) and 0.716 (longitudinal). The whole pipeline reached an AUROC of 0.84 (CI 95%: 0.75–0.91) with sensitivity and specificity of 84% and 63%, respectively, while the ACR-TIRADS evaluation of the same set had a sensitivity of 76% and specificity of 34% (p-value = 0.003). Our proposed work demonstrated the potential possibility of a deep learning model to achieve diagnostic performance for thyroid nodule evaluation.