Combining classical radiomics and autoencoder-based features extracted from tumor bounding boxes improves US-based categorization of benign and malignant breast tumors compared with segmentation-based approaches.
Automation of medical data analysis is an important topic in modern cancer diagnostics, aiming at robust and reproducible workflows. Therefore, we used a dataset of breast US images (252 malignant and 253 benign cases) to realize and compare different strategies for CAD support in lesion detection and classification. Eight different datasets (including pre-processed and spatially augmented images) were prepared, and machine learning algorithms (i.e., Viola–Jones; YOLOv3) were trained for lesion detection. The radiomics signature (RS) was derived from detection boxes and compared with RS derived from manually obtained segments. Finally, the classification model was established and evaluated concerning accuracy, sensitivity, specificity, and area under the Receiver Operating Characteristic curve. After training on a dataset including logarithmic derivatives of US images, we found that YOLOv3 obtains better results in breast lesion detection (IoU: 0.544 ± 0.081; LE: 0.171 ± 0.009) than the Viola–Jones framework (IoU: 0.399 ± 0.054; LE: 0.096 ± 0.016). Interestingly, our findings show that the classification model trained with RS derived from detection boxes and the model based on the RS derived from a gold standard manual segmentation are comparable (p-value = 0.071). Thus, deriving radiomics signatures from the detection box is a promising technique for building a breast lesion classification model, and may reduce the need for the lesion segmentation step in the future design of CAD systems.
Objective: The aim of the present study was to investigate the influence of three grafting materials for cleft repair on orthodontic tooth movement in rats. Materials and methods: Artificial alveolar clefts were created in 21 Wistar rats and were repaired 4 weeks later using autografts, human xenografts and synthetic bone substitute (beta-tricalcium phosphate/hydroxyapatite [β-TCP/HA]). A further 4 weeks later, the first molar was moved into the reconstructed maxilla. Microfocus computed tomography (μCT) was performed six times (T0–T5) to assess the tooth movement and root resorption. After 8 weeks, the affected reconstructed jaw was resected for histopathological investigation. Results: Total distances reached ranged from 0.82 ± 0.72 mm (β-TCP/HA) to 0.67 ± 0.27 mm (autograft). The resorption was particularly determined at the mesiobuccal root. Descriptive tooth movement slowed and root resorption increased slightly. However, neither the radiological changes during tooth movement (µCT T1 vs. µCT T5: autograft 1.85 ± 0.39 mm3 vs. 2.38 ± 0.35 mm3, p = 0.30; human xenograft 1.75 ± 0.45 mm3 vs. 2.17 ± 0.26 mm3, p = 0.54; β-TCP/HA: 1.52 ± 0.42 mm3 vs. 1.88 ± 0.41 mm3, p = 0.60) nor the histological differences after tooth movement (human xenograft: 0.078 ± 0.05 mm2; β-TCP/HA: 0.067 ± 0.049 mm2; autograft: 0.048 ± 0.015 mm2) were statistically significant. Conclusion: The autografts, human xenografts or synthetic bone substitute used for cleft repair seem to have a similar effect on the subsequent orthodontic tooth movement and the associated root resorptions. Clinical relevance: Development of root resorptions seems to have a secondary role in choosing a suitable grafting material for cleft repair.
27 November 2024 in IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control
Angiogenesis - the formation of new blood vessels from pre-existing ones - is one of the hallmarks of cancer, regardless of subtype. However, the development of a specific tumor type is a highly heterogeneous process that influences the morphology of the tumor vasculature, which has a direct impact on the malignancy and invasiveness of the lesions. Therefore, the analysis of tumor vascularity without the need for invasive procedures is of fundamental interest for the classification of tumor tissue and the monitoring of therapies. Ultrasound localization microscopy is a promising new technique that breaks the resolution limits of conventional ultrasound imaging and allows to detect vascular structures and blood flow down to the capillary level. In this article, we discuss this emerging technique in the context of cancer imaging, focusing on crucial implementation aspects as well as on initial basic research in preclinical and clinical settings.
In addition to post-extraction bleeding, pronounced alveolar bone resorption is a very common complication after tooth extraction in patients undergoing anticoagulation therapy. The novel, biodegenerative, polyurethane adhesive VIVO has shown a positive effect on soft tissue regeneration and hemostasis. However, the regenerative potential of VIVO in terms of bone regeneration has not yet been explored. The present rodent study compared the post-extraction bone healing of a collagen sponge (COSP) and VIVO in the context of ongoing anticoagulation therapy. According to a split-mouth design, a total of 178 extraction sockets were generated under rivaroxaban treatment, of which 89 extraction sockets were treated with VIVO and 89 with COSP. Post-extraction bone analysis was conducted via in vivo micro-computed tomography (µCT), scanning electron microscopy (SEM), and energy-dispersive X-ray spectroscopy (EDX) after 5, 10, and 90 days. During the observation time of 90 days, µCT analysis revealed that VIVO and COSP led to significant increases in both bone volume and bone density (p ≤ 0.001). SEM images of the extraction sockets treated with either VIVO or COSP showed bone regeneration in the form of lamellar bone mass. Ratios of Ca/C and Ca/P observed via EDX indicated newly formed bone matrixes in both treatments after 90 days. There were no statistical differences between treatment with VIVO or COSP. The hemostatic agents VIVO and COSP were both able to prevent pronounced bone loss, and both demonstrated a strong positive influence on the bone regeneration of the alveolar ridge post-extraction.
Regeneration of jawbone defects poses major challenges. The combination of dental pulp stem cells (DPSCs) or DPSC-derived extracellular vesicles (EVs) with bone substitute materials shows promising potential for bone tissue engineering in vitro. This study evaluated the in vivo bone regeneration potential of DPSCs and EVs with bone graft substitutes in a novel intrabony defect model. DPSCs were isolated from 35 male Sprague–Dawley rat incisors, and EVs were collected from the cell culture medium. DPSCs were seeded onto alloplastic and xenogeneic bone graft materials and implanted into bone defects. Control groups received bone substitutes without DPSCs or EVs. Micro-computed tomography (µCT) was performed at 12 and 24 weeks post-implantation to assess bone volume (BV), bone density (BD), trabecular thickness (Tr.Th), bone growth rate (BGR), and bone-to-mineral ratio (BMR). Both graft types increased BV and BD, with no significant differences between them. Tr.Th increased across all treatments after 24 weeks, indicating ongoing bone remodeling. Notably, xenogeneic grafts combined with DPSCs and EVs significantly improved BGR (p = 0.034) and BMR (p = 0.021) compared to alloplastic grafts with DPSCs. Xenogeneic bone grafts combined with DPSCs and EVs appear to be a promising approach for bone regeneration of alveolar bone defects.