Comparative evaluation of image-based vs. text-based vs. multimodal AI approaches for automatic breast density assessment in mammograms.

BACKGROUND AND OBJECTIVES: In the last decade, there has been a growing interest in applying artificial intelligence (AI) systems to breast cancer assessment, including breast density evaluation. However, few models have been developed to integrate textual mammographic reports and mammographic images. Our aims are (1) to generate a natural language processing (NLP)-based AI system, (2) to evaluate an external image-based software, and (3) to develop a multimodal system, using the late fusion approach, by integrating image and text inferences for the automatic classification of breast density according to the American College of Radiology (ACR) guidelines in mammograms and radiological reports. METHODS: We first compared different NLP models, three based on n-gram term frequency - inverse document frequency and two transformer-based architectures, using 1533 unstructured mammogram reports as a training set and 303 reports as a test set. Subsequently, we evaluated an external image-based software using 303 mammogram images. Finally, we assessed our multimodal system taking into account both text and mammogram images. RESULTS: Our best NLP model achieved 88 % accuracy, while the external software and the multimodal system achieved 75 % and 80 % accuracy, respectively, in classifying ACR breast densities. CONCLUSION: Although our multimodal system outperforms the image-based tool, it currently does not improve the results offered by the NLP model for ACR breast density classification. Nevertheless, the promising results observed here open the possibility to more comprehensive studies regarding the utilization of multimodal tools in the assessment of breast density.

Role of advanced MRI sequences for thyroid lesions assessment. A narrative review.

Despite not being the first imaging modality for thyroid gland assessment, Magnetic Resonance Imaging (MRI), thanks to its optimal tissue contrast and spatial resolution, has provided some advancements in detecting and characterizing thyroid abnormalities. Recent research has been focused on improving MRI sequences and employing advanced techniques for a more comprehensive understanding of thyroid pathology. Although not yet standard practice, advanced MRI sequences have shown high accuracy in preliminary studies, correlating well with histopathological results. They particularly show promise in determining malignancy risk in thyroid lesions, which may reduce the need for invasive procedures like biopsies. In this line, functional MRI sequences like Diffusion Weighted Imaging (DWI), Dynamic Contrast-Enhanced MRI (DCE-MRI), and Arterial Spin Labeling (ASL) have demonstrated their potential usefulness in evaluating both diffuse thyroid conditions and focal lesions. Multicompartmental DWI models, such as Intravoxel Incoherent Motion (IVIM) and Diffusion Kurtosis Imaging (DKI), and novel methods like Amide Proton Transfer (APT) imaging or artificial intelligence (AI)-based analyses are being explored for their potential valuable insights into thyroid diseases. This manuscript reviews the critical physical principles and technical requirements for optimal functional MRI sequences of the thyroid and assesses the clinical utility of each technique. It also considers future prospects in the context of advanced MR thyroid imaging and analyzes the current role of advanced MRI sequences in routine practice.

Utilidad de un modelo de reconstrucción e impresión 3D en Cirugía Hepática Compleja

Introducción: La anatomía hepática siempre ha sido un reto por su complejidad y variabilidad. En los últimos años, el abaratamiento de los costes ha permitido la generación de modelos 3D individualizados para cada paciente que pueden facilitar el abordaje quirúrgico de las lesiones. El objetivo principal fue determinar la utilidad del modelado 3D preoperatorio para la planificación quirúrgica en pacientes con lesiones hepáticas. Métodos: Se trata de un estudio de casos de 38 pacientes intervenidos por lesiones hepáticas múltiples ocupantes de espacio, en el cual, en un grupo seleccionado, en 19 pacientes se utilizó un modelo impreso 3D para planificar la cirugía (grupo 3D) y el otro grupo sin el modelo impreso 3D (grupo control). Resultados: Se observó una diferencia de medias significativa en el número de lesiones; mayor en el grupo 3D al realizar el test de Wilcoxon (p < 0,001) y un mayor número de casos con afectación vascular en este mismo grupo al realizar Chi cuadrado Pearson (p = 0,008). El resto de variables no mostraron diferencias estadísticamente significativas. A pesar de esto, la mortalidad se redujo a 0 cuando se usan modelos impresos en 3D. Conclusión: La impresión 3D permite planear, de manera más precisa, cirugías complejas del hígado, ayuda a la inclusión y exclusión de los pacientes para la cirugía, disminuyendo el tiempo de la sala de operaciones, la posterior hospitalización y las complicaciones quirúrgicas.

Introduction: Liver anatomy has always been a challenge due to its complexity and variability. In recent years, lower costs has allowed the generation of individualized 3D models for each patient, which can facilitate the surgical approach to liver lesions. The main objective was to determine usefulness of preoperative 3D modeling for surgical planning in patients with liver lesions. Methods: Quasi-experimental before-after study. 19 cases were included in which surgery was planned using a 3D printed model (13 bilobar hepatectomies, 3 of them with vascular involvement, and 6 unilobar hepatectomies, 1 of them with vascular involvement), and another 19 cases whose planning was carried out without a 3D printed model (7 bilobar segmental hepatic resections and 12 unilobar segmental resections. None of these cases had vascular involvement). Results: A significant difference in mean lesion count was observed, higher in the group of cases when performing the Wilcoxon test (p < 0.001), and a higher number of cases with vascular involvement in the same group when performing the Pearson chi-square test (p = 0.008). The rest of the variables did not show statistically significant differences. Despite this, mortality was reduced to 0 when 3D printed models were used. Conclusion: 3D printing allows for more precise planning of complex liver surgeries, helps with the inclusion and exclusion of patients for surgery, reduces operating room time, postoperative hospitalization, and surgical complications.