Search

Classifying pixels according to color, and segmenting the respective areas, are necessary steps in any computer vision task that involves color images. The gap between human color perception, linguistic color terminology, and digital representation are the main challenges for developing methods that properly classify pixels based on color. To address these challenges, we propose a novel method combining geometric analysis, color theory, fuzzy color theory, and multi-label systems for the automatic classification of pixels into 12 conventional color categories, and the subsequent accurate description of each of the detected colors. This method presents a robust, unsupervised, and unbiased strategy for color naming, based on statistics and color theory. The proposed model, "ABANICCO" (AB ANgular Illustrative Classification of COlor), was evaluated through different experiments: its color detection, classification, and naming performance were assessed against the standardized ISCC–NBS color system; its usefulness for image segmentation was tested against state-of-the-art methods. This empirical evaluation provided evidence of ABANICCO's accuracy in color analysis, showing how our proposed model offers a standardized, reliable, and understandable alternative for color naming that is recognizable by both humans and machines. Hence, ABANICCO can serve as a foundation for successfully addressing a myriad of challenges in various areas of computer vision, such as region characterization, histopathology analysis, fire detection, product quality prediction, object description, and hyperspectral imaging. [ABSTRACT FROM AUTHOR]

Simple Summary: Four decades ago, intraoperative electron radiation therapy (IOeRT) was developed to improve precision in local cancer treatment by combining real-time surgical exploration and resection with high-energy electron irradiation. The technology of ultra-high dose rate electron and other radiation beams known as FLASH irradiation sharply increases its interests, as data from preclinical experiments have proven a marked favorable effect on the therapeutic index: similar cancer control with a clearly improved tolerance of many normal tissues to high doses of irradiation. The knowledge and tools regarding technology, physics, biology, and preclinical results in heterogeneous cancers opens great opportunities towards the path of developing the first clinical applications of the emerging FLASH technology via clinical trials based on state-of-the-art medical practice with IOeRT. Introduction: The clinical practice and outcome results of intraoperative electron radiation therapy (IOeRT) in cancer patients have been extensively reported over 4 decades. Electron beams can be delivered in the promising FLASH dose rate. Methods and Materials: Several cancer models were approached by two alternative radiobiological strategies to optimize local cancer control: boost versus exclusive IOeRT. Clinical outcomes are revisited via a bibliometric search performed for the elaboration of ESTRO/ACROP IORT guidelines. Results: In the period 1982 to 2020, a total of 19,148 patients were registered in 116 publications concerning soft tissue sarcomas (9% of patients), unresected and borderline-resected pancreatic cancer (22%), locally recurrent and locally advanced rectal cancer (22%), and breast cancer (45%). Clinical outcomes following IOeRT doses in the range of 10 to 25 Gy (with or without external beam fractionated radiation therapy) show a wide range of local control from 40 to 100% depending upon cancer site, histology, stage, and treatment intensity. Constraints for normal tissue tolerance are important to maintain tumor control combined with acceptable levels of side effects. Conclusions: IOeRT represents an evidence-based approach for several tumor types. A specific risk analysis for local recurrences supports the identification of cancer models that are candidates for FLASH studies. [ABSTRACT FROM AUTHOR]

This work analyzed the use of Microsoft HoloLens 2 in orthopedic oncological surgeries and compares it to its predecessor (Microsoft HoloLens 1). Specifically, we developed two equivalent applications, one for each device, and evaluated the augmented reality (AR) projection accuracy in an experimental scenario using phantoms based on two patients. We achieved automatic registration between virtual and real worlds using patient-specific surgical guides on each phantom. They contained a small adaptor for a 3D-printed AR marker, the characteristic patterns of which were easily recognized using both Microsoft HoloLens devices. The newest model improved the AR projection accuracy by almost 25%, and both of them yielded an RMSE below 3 mm. After ascertaining the enhancement of the second model in this aspect, we went a step further with Microsoft HoloLens 2 and tested it during the surgical intervention of one of the patients. During this experience, we collected the surgeons' feedback in terms of comfortability, usability, and ergonomics. Our goal was to estimate whether the improved technical features of the newest model facilitate its implementation in actual surgical scenarios. All of the results point to Microsoft HoloLens 2 being better in all the aspects affecting surgical interventions and support its use in future experiences. [ABSTRACT FROM AUTHOR]

Surgical navigation techniques have shown potential benefits in orthopedic oncologic surgery. However, the translation of these results to acral tumor resection surgeries is challenging due to the large number of joints with complex movements of the affected areas (located in distal extremities). This study proposes a surgical workflow that combines an intraoperative open-source navigation software, based on a multi-camera tracking, with desktop three-dimensional (3D) printing for accurate navigation of these tumors. Desktop 3D printing was used to fabricate patient-specific 3D printed molds to ensure that the distal extremity is in the same position both in preoperative images and during image-guided surgery (IGS). The feasibility of the proposed workflow was evaluated in two clinical cases (soft-tissue sarcomas in hand and foot). The validation involved deformation analysis of the 3D-printed mold after sterilization, accuracy of the system in patient-specific 3D-printed phantoms, and feasibility of the workflow during the surgical intervention. The sterilization process did not lead to significant deformations of the mold (mean error below 0.20 mm). The overall accuracy of the system was 1.88 mm evaluated on the phantoms. IGS guidance was feasible during both surgeries, allowing surgeons to verify enough margin during tumor resection. The results obtained have demonstrated the viability of combining open-source navigation and desktop 3D printing for acral tumor surgeries. The suggested framework can be easily personalized to any patient and could be adapted to other surgical scenarios. [ABSTRACT FROM AUTHOR]

Patient-specific instruments (PSIs) have become a valuable tool for osteotomy guidance in complex surgical scenarios such as pelvic tumor resection. They provide similar accuracy to surgical navigation systems but are generally more convenient and faster. However, their correct placement can become challenging in some anatomical regions, and it cannot be verified objectively during the intervention. Incorrect installations can result in high deviations from the planned osteotomy, increasing the risk of positive resection margins. In this work, we propose to use augmented reality (AR) to guide and verify PSIs placement. We designed an experiment to assess the accuracy provided by the system using a smartphone and the HoloLens 2 and compared the results with the conventional freehand method. The results showed significant differences, where AR guidance prevented high osteotomy deviations, reducing maximal deviation of 54.03 mm for freehand placements to less than 5 mm with AR guidance. The experiment was performed in two versions of a plastic three-dimensional (3D) printed phantom, one including a silicone layer to simulate tissue, providing more realism. We also studied how differences in shape and location of PSIs affect their accuracy, concluding that those with smaller sizes and a homogeneous target surface are more prone to errors. Our study presents promising results that prove AR's potential to overcome the present limitations of PSIs conveniently and effectively. [ABSTRACT FROM AUTHOR]

Deep learning is a recent technology that has shown excellent capabilities for recognition and identification tasks. This study applies these techniques in open cranial vault remodeling surgeries performed to correct craniosynostosis. The objective was to automatically recognize surgical tools in real-time and estimate the surgical phase based on those predictions. For this purpose, we implemented, trained, and tested three algorithms based on previously proposed Convolutional Neural Network architectures (VGG16, MobileNetV2, and InceptionV3) and one new architecture with fewer parameters (CranioNet). A novel 3D Slicer module was specifically developed to implement these networks and recognize surgical tools in real time via video streaming. The training and test data were acquired during a surgical simulation using a 3D printed patient-based realistic phantom of an infant's head. The results showed that CranioNet presents the lowest accuracy for tool recognition (93.4%), while the highest accuracy is achieved by the MobileNetV2 model (99.6%), followed by VGG16 and InceptionV3 (98.8% and 97.2%, respectively). Regarding phase detection, InceptionV3 and VGG16 obtained the best results (94.5% and 94.4%), whereas MobileNetV2 and CranioNet presented worse values (91.1% and 89.8%). Our results prove the feasibility of applying deep learning architectures for real-time tool detection and phase estimation in craniosynostosis surgeries. [ABSTRACT FROM AUTHOR]

During the last decade, orthopedic oncology has experienced the benefits of computerized medical imaging to reduce human dependency, improving accuracy and clinical outcomes. However, traditional surgical navigation systems do not always adapt properly to this kind of interventions. Augmented reality (AR) and three-dimensional (3D) printing are technologies lately introduced in the surgical environment with promising results. Here we present an innovative solution combining 3D printing and AR in orthopedic oncological surgery. A new surgical workflow is proposed, including 3D printed models and a novel AR-based smartphone application (app). This app can display the patient's anatomy and the tumor's location. A 3D-printed reference marker, designed to fit in a unique position of the affected bone tissue, enables automatic registration. The system has been evaluated in terms of visualization accuracy and usability during the whole surgical workflow. Experiments on six realistic phantoms provided a visualization error below 3 mm. The AR system was tested in two clinical cases during surgical planning, patient communication, and surgical intervention. These results and the positive feedback obtained from surgeons and patients suggest that the combination of AR and 3D printing can improve efficacy, accuracy, and patients' experience. [ABSTRACT FROM AUTHOR]

We present a novel method to assess the variations in protein expression and spatial heterogeneity of tumor biopsies with application in computational pathology. This was done using different antigen stains for each tissue section and proceeding with a complex image registration followed by a final step of color segmentation to detect the exact location of the proteins of interest. For proper assessment, the registration needs to be highly accurate for the careful study of the antigen patterns. However, accurate registration of histopathological images comes with three main problems: the high amount of artifacts due to the complex biopsy preparation, the size of the images, and the complexity of the local morphology. Our method manages to achieve an accurate registration of the tissue cuts and segmentation of the positive antigen areas. [ABSTRACT FROM AUTHOR]

Radiotherapy is one of the main treatments for localized head and neck (HN) cancer. To design a personalized treatment with reduced radio-induced toxicity, accurate delineation of organs at risk (OAR) is a crucial step. Manual delineation is time- and labor-consuming, as well as observer-dependent. Deep learning (DL) based segmentation has proven to overcome some of these limitations, but requires large databases of homogeneously contoured image sets for robust training. However, these are not easily obtained from the standard clinical protocols as the OARs delineated may vary depending on the patient's tumor site and specific treatment plan. This results in incomplete or partially labeled data. This paper presents a solution to train a robust DL-based automated segmentation tool exploiting a clinical partially labeled dataset. We propose a two-step workflow for OAR segmentation: first, we developed longitudinal OAR-specific 3D segmentation models for pseudo-contour generation, completing the missing contours for some patients; with all OAR available, we trained a multi-class 3D convolutional neural network (nnU-Net) for final OAR segmentation. Results obtained in 44 independent datasets showed superior performance of the proposed methodology for the segmentation of fifteen OARs, with an average Dice score coefficient and surface Dice similarity coefficient of 80.59% and 88.74%. We demonstrated that the model can be straightforwardly integrated into the clinical workflow for standard and adaptive radiotherapy.