Image Acquisition Research Articles

Optical coherence tomography to guide percutaneous coronary intervention.

Percutaneous coronary intervention (PCI) has been most commonly guided by coronary angiography. However, to overcome the inherent limitations of conventional coronary angiography, there has been an increasing interest in the adjunctive tools of intracoronary imaging for PCI guidance. Recently, optical coherence tomography (OCT) has garnered substantial attention as a valid intravascular imaging modality for guiding PCI. However, despite the unparalleled high-resolution imaging capability of OCT, which offers detailed anatomical information on coronary lesion morphology and PCI optimisation, its broad application in routine PCI practice remains limited. Several factors may have curtailed the widespread adoption of OCT-guided PCI in daily practice, including the transitional challenge from intravascular ultrasound (IVUS), the experienced skill required for image acquisition and interpretation, the lack of a uniform algorithm for OCT-guided PCI optimisation, and the limited clinical evidence. Herein, we provide an in-depth review of OCT-guided PCI, involving the technical aspects, optimal strategies for OCT-guided PCI, and the wide application of OCT-guided PCI in various anatomical subsets. Special attention is given to the latest clinical evidence from recent randomised clinical trials with respect to OCT-guided PCI.

Applications of artificial intelligence in radiology

Artificial intelligence (AI) is increasingly finding its way into routine radiological work. Presentation of the current advances and applications of AI along the entire radiological patient journey. Systematic literature review of established AI techniques and current research projects, with reference to consensus recommendations. The applications of AI in radiology cover awide range, starting with AI-supported scheduling and indications assessment, extending to AI-enhanced image acquisition and reconstruction techniques that have the potential to reduce radiation doses in computed tomography (CT) or acquisition times in magnetic resonance imaging (MRI), while maintaining comparable image quality. These include computer-aided detection and diagnosis, such as fracture recognition or nodule detection. Additionally, methods such as worklist prioritization and structured reporting facilitated by large language models enable arethinking of the reporting process. The use of AI promises to increase the efficiency of all steps of the radiology workflow and an improved diagnostic accuracy. To achieve this, seamless integration into technical workflows and proven evidence of AI systems are necessary. Applications of AI have the potential to profoundly influence the role of radiologists in the future.

Determining Simple and Effective Cost Functions for an Efficient Volumetric-Modulated Arcs-Based Stereotactic Radiosurgery for Single Brain Metastases Using Monaco® Planning System.

Introduction Volumetric-modulated arcs (VMA) can producedose distributions suitable for stereotactic radiosurgery (SRS) with a multi-leaf collimator (MLC) for brain metastases (BMs). The treatment planning and verification for VMA are more complicatedthan for dynamic conformal arcs. The longer the preparation timefrom image acquisition to the start of irradiation, the higher the risk of tumor growth and/or displacement. This planning study aimed to exploit the simple and effective cost function (CF) for establishingsemi-automatic efficient VMA optimizationfor SRS ofsingle BMs. Materials and methods The study population included 30 clinical BMs with a gross tumor volume (GTV) of 0.72-44.30 cc (median 9.81 cc) and a depth of 20-79 mm (median 41 mm). The treatment platform included a 5-mm leaf-width MLC Agility® (Elekta AB, Stockholm, Sweden) and a planning system Monaco® (Elekta AB). Among various physical and biological CFs available, three combinations consisting of just two or three physical CFs were compared. The Target Penalty CF was uniformly used for ensuring the GTV dose. Three different CF combinations were applied for reducing the surrounding tissue doses: (1) the Conformality alone with the 4-cm margin around target (MAT) that optimizes the limited voxels around the GTV (woQO); (2) the Conformality with the 4-cm MAT and the Quadratic Overdose (wQO_4 cm); and (3) the Conformality with the 8-cm MAT that optimizes the overall voxels around the GTV and the Quadratic Overdose (wQO_8 cm). The prescribed dose was uniformly assigned to each GTV D V-0.01 cc, the minimum dose of GTV minus 0.01 cc. Results Adding the Quadratic Overdose (wQO_4 cm and wQO_8 cm) significantly improved the overall dose distribution in comparison to the woQO, while no significant difference was observed between the wQO_4 cm and wQO_8 cm overall. However, for the GTVs of ≥14 cc, the GTV dose conformity and dose gradient outside the GTV boundary, including the dose attenuation margin, were significantly superior in the wQO_8 cm than wQO_4 cm. In addition, for the GTV depth of ≥41 mm, the GTV dose conformity and the dose concentric lamellarity at 2 mm outside the GTV were significantly superior in the wQO_8 cm than wQO_4 cm. Meanwhile, for the GTVs of ≥10 cc, the GTV dose was significantly more inhomogeneous inthe wQO_4 cm than thewQO_8 cm. In addition, for the GTVs of <10 cc and the depth of ≤40 mm, the dose concentric lamellarity at 4 mm inside the GTV surface was significantly higher in the wQO_4 cm than thewQO_8 cm. Conclusions Applying at least three physical CFs to a GTV and the head surface contouris recommended as an effective and efficient optimization method using Monaco for VMA-based SRS of single BMs. In addition, optimizing the overall voxels around the GTV is suitable for reducing the surrounding tissue dose, especially for large and deeply located lesions. Templating the combination of the three CFs with the detailed settings allows for semi-automated and rapid treatment planning, facilitating the prompt start of irradiation after image acquisition.

In vitro and in vivo evaluation of a novel wired transmission pH-combined photographic catheter for ambulatory gastroesophageal reflux monitoring (with videos).

Twenty-four-hour pH-impedance monitoring is an important diagnostic approach for gastroesophageal reflux disease (GERD). Reflux monitoring results cannot be synchronized with ambulatory motility imaging of the esophageal sphincter. We have designed a novel wired transmission pH-combined photographic catheter (WT-CPC) for the synchronous acquisition of reflux image and pH. Different patterns of reflux events were simulated to perform in a porcine gastroesophageal reflux model in vitro. The live porcine model of gastroesophageal reflux was established in three Bama pigs. Monitoring was conducted with the WT-CPC and pH-impedance catheter simultaneously. Measurements included the number and proportion of reflux events, as well as acid exposure time (AET). The detection rates of WT-CPC for distal and horizontal acid reflux events were significantly higher compared to those of pH-impedance catheters (100% vs. 14.29%, 100% vs. 57.14%, P < 0.05). There was no significant difference between the two methods in proximal acid reflux events (P = 0.217). Regarding mixed reflux events, WT-CPC exhibited higher detection rates for distal events than pH-impedance catheter (100% vs. 42.86%, P < 0.05). However, there was no significant difference between the two methods for proximal reflux events (P > 0.05). Both methods showed similar results for horizontal reflux events. A porcine gastroesophageal reflux model was successfully established and utilized for reflux monitoring. A total of 28 episodes of reflux were detected within 6.5min. The detection rate achieved by WT-CPC for reflux events was significantly higher than that obtained by pH-impedance (100% vs. 78.57%, P = 0.023). The WT-CPC has demonstrated reflux monitoring capabilities in an isolated reflux organ model. It also showed good operability and performance in the porcine model. The WT-CPC holds promising potential to provide valuable diagnostic evidence for GERD.

Evolving capabilities of computed tomography imaging for transcatheter valvular heart interventions - new opportunities for precision medicine.

The last decade has witnessed a substantial growth in percutaneous treatment options for heart valve disease. The development in these innovative therapies has been mirrored by advances in multi-detector computed tomography (MDCT). MDCT plays a central role in obtaining detailed pre-procedural anatomical information, helping to inform clinical decisions surrounding procedural planning, improve clinical outcomes and prevent potential complications. Improvements in MDCT image acquisition and processing techniques have led to increased application of advanced analytics in routine clinical care. Workflow implementation of patient-specific computational modeling, fluid dynamics, 3D printing, extended reality, extracellular volume mapping and artificial intelligence are shaping the landscape for delivering patient-specific care. This review will provide an insight of key innovations in the field of MDCT for planning transcatheter heart valve interventions.

Simultaneous assessment of NAD(P)H and flavins with multispectral fluorescence lifetime imaging microscopy at a single excitation wavelength of 750nm.

Autofluorescence characteristics of the reduced nicotinamide adenine dinucleotide and oxidized flavin cofactors are important for the evaluation of the metabolic status of the cells. The approaches that involve a detailed analysis of both spectral and time characteristics of the autofluorescence signals may provide additional insights into the biochemical processes in the cells and biological tissues and facilitate the transition of spectral fluorescence lifetime imaging into clinical applications. We present the experiments on multispectral fluorescence lifetime imaging with a detailed analysis of the fluorescence decays and spectral profiles of the reduced nicotinamide adenine dinucleotide and oxidized flavin under a single excitation wavelength aimed at understanding whether the use of multispectral detection is helpful for metabolic imaging of cancer cells. We use two-photon spectral fluorescence lifetime imaging microscopy. Starting from model solutions, we switched to cell cultures treated by metabolic inhibitors and then studied the metabolism of cells within tumor spheroids. The use of a multispectral detector in combination with an excitation at a single wavelength of 750nm allows the identification of fluorescence signals from three components: free and bound NAD(P)H, and flavins based on the global fitting procedure. Multispectral data make it possible to assess not only the lifetime but also the spectral shifts of emission of flavins caused by chemical perturbations. Altogether, the informative parameters of the developed approach are the ratio of free and bound NAD(P)H amplitudes, the decay time of bound NAD(P)H, the amplitude of flavin fluorescence signal, the fluorescence decay time of flavins, and the spectral shift of the emission signal of flavins. Hence, with multispectral fluorescence lifetime imaging, we get five independent parameters, of which three are related to flavins. The approach to probe the metabolic state of cells in culture and spheroids using excitation at a single wavelength of 750nm and a fluorescence time-resolved spectral detection with the consequent global analysis of the data not only simplifies image acquisition protocol but also allows to disentangle the impacts of free and bound NAD(P)H, and flavin components evaluate changes in their fluorescence parameters (emission spectra and fluorescence lifetime) upon treating cells with metabolic inhibitors and sense metabolic heterogeneity within 3D tumor spheroids.

Lymph node metastasis diagnosis of postoperative OSCC patients by analyzing extracellular vesicles in drainage fluid based on microfluidic isolation

Lymph node metastasis (LNM) is a typical marker in oral squamous cell carcinoma (OSCC) indicating poor prognosis. Pathological examination by artificial image acquisition and analysis, as the main diagnostic method for LNM, often takes a week or longer which may cause great anxiety of the patient and also retard timely treatment. However, there are few efficient fast LNM diagnosis methods in clinical applications currently. Our previous study profiled the proteomics of extracellular vesicles (EVs) derived from postoperative drainage fluid (PDF) and showed the potential of detecting specific EVs that expressed aspartate β-hydroxylase (ASPH) for LNM diagnosis in OSCC patients. Considering that the analysis of ASPH+ PDF-EVs is challenging due to their low abundance (counting less than 10% of total EVs in PDF) and the complex EV isolation process of ultra-centrifugation, we developed a facile platform containing two microfluidic chips filled with antibody-modified microbeads to isolate ASPH+ PDF-EVs, with both the capture and retrieval rate reaching around 90%. Clinical sample analysis based on our method revealed that a mean of 6 × 106 /mL ASPH+ PDF-EVs could be isolated from LNM+ OSCC patients compared to 2.5 × 106 /mL in LNM- OSCC ones. When combined with enzyme-linked immunosorbent assay (ELISA) technique that was commonly used in clinical laboratories in hospitals, this microfluidic platform could precisely distinguish postoperative OSCC patients with LNM or not in several hours, which were validated by a double-blind test containing 6 OSCC patients. We believe this strategy has promise for early diagnosis of LNM in postoperative OSCC patients and finally helps guiding timely and reasonable treatment in clinic.

Plant-based egg washes for use in baked goods: Machine learning and visual parameter analysis.

Pea protein is one potential environmentally sustainable way of recreating the functionality of eggs in coatings for baked goods. These coatings are commonly applied to enhance visual properties of baked goods that consumers desire, especially color and gloss. This project used a simplified pie crust formulation (flour, shortening, and water) as a model baked good system to investigate color and gloss properties of pea protein coatings comprising three pea protein glycerol ratios, prepared at 2pH values, and applied at three different weights, baked for either 15 or 30 min. Pictures were also taken and used to estimate L*, a*, and b* colorimeter values from images using random forest and artificial neural network models developed using SciKit-Learn. Image acquisition was also conducted at different lighting conditions and the results are corrected utilizing a principal component analysis (PCA) weighting approach. Results show that pea protein glycerol solutions are all glossier than egg washes, and the 80%/20% ratio gives the highest glossiness. Furthermore, there are not significant differences in any of the colorimeter values at any concentration when applying 1 g of protein. The artificial neural network model coupled with PCA weighting led to reasonable estimates of color at different lighting conditions. Together, these results offer an alternative for the use of pea protein/glycerol in lieu of egg washes.

Identifying defects and varieties of Malting Barley Kernels

This study introduces a comprehensive approach for classifying individual malting barley kernels, involving dual-sided kernel imaging, a specifically designed image processing algorithm, an optimized deep neural network architecture, and a mechanical sorting system. The proposed method achieves precise classification into multiple classes, aligning with quality standards for malting material assessment. Throughout the study, various image analysis techniques were assessed, including traditional feature engineering, established transfer learning deep neural network architectures, and our custom-designed convolutional neural network tailored for barley kernel image analysis. Comparative analysis underscores the superior performance of our network model. The study reveals that our proposed deep learning network achieves a 94% accuracy in classifying barley kernel defects and varieties, outperforming well-established transfer learning models to complex architectures that attain 93% accuracy. Additionally, it surpasses the traditional machine learning approach involving feature extraction and support vector machine classifiers, which achieve accuracy below 90% in detecting defective kernels and below 70% in varietal classification. However, we also noted the traditional approach’s advantage in morphological feature recognition. This observation guides new research toward integrating morphological feature extraction techniques with modern convolutional networks. This paper presents a deep neural network designed specifically for the analysis of cereal kernel images in two applications: defect and variety classification. It emphasizes the importance of standardizing kernel orientation and merging images from both sides of the kernel, and introduces a device for image acquisition that fulfills this need.

Development of AI-assisted microscopy frameworks through realistic simulation with pySTED

The integration of artificial intelligence into microscopy systems significantly enhances performance, optimizing both image acquisition and analysis phases. Development of artificial intelligence-assisted super-resolution microscopy is often limited by access to large biological datasets, as well as by difficulties to benchmark and compare approaches on heterogeneous samples. We demonstrate the benefits of a realistic stimulated emission depletion microscopy simulation platform, pySTED, for the development and deployment of artificial intelligence strategies for super-resolution microscopy. pySTED integrates theoretically and empirically validated models for photobleaching and point spread function generation in stimulated emission depletion microscopy, as well as simulating realistic point-scanning dynamics and using a deep learning model to replicate the underlying structures of real images. This simulation environment can be used for data augmentation to train deep neural networks, for the development of online optimization strategies and to train reinforcement learning models. Using pySTED as a training environment allows the reinforcement learning models to bridge the gap between simulation and reality, as showcased by its successful deployment on a real microscope system without fine tuning.

Visualization of stepwise electrode decomposition in a nail penetrated commercial lithium-ion cell using low-temperature synchrotron X-ray computed tomography

The transition towards zero carbon emissions in power generation hinges on the integration of efficient electrical energy storage systems, with lithium-ion batteries (LIBs) positioned as a pivotal technology. While generally safe, deviations in their operational guidelines due to manufacturing defects or misuse can lead to critical safety concerns, notably thermal runaway (TR) events. Internal short circuits (ISCs) are primary initiators of TR within LIBs. For abuse testing, ISCs are often triggered by nail penetration. This study explores the morphological changes and mechanisms underlying ISC-induced TR in LIBs using operando synchrotron X-ray computed tomography (SXCT) at subzero temperatures. A novel cryogenic setup was developed to control a stepwise temperature increase in the damaged sample while monitoring electrochemical characteristics and simultaneously enabling acquisition of high-resolution SXCT images. The findings reveal that conducting nail penetration at minus 80 °C prevents immediate TR, enabling detailed analysis of subsequent structural and electrochemical behavior during controlled thawing. Thus, the initiation of TR processes at localized ISC sites has been observed, evidenced by voltage fluctuations and morphological changes, such as cathode material cracking and decomposition. These results underscore the importance of temperature control in mitigating TR risks and provide critical insights into the internal dynamics of LIBs under abusive conditions. The developed cryogenic SXCT methodology offers a powerful tool for non-destructive, high-resolution investigation of battery failure mechanisms, contributing to the enhancement of LIB safety.

Light-field tomographic fluorescence lifetime imaging microscopy

Fluorescence lifetime imaging microscopy (FLIM) is a powerful imaging technique that enables the visualization of biological samples at the molecular level by measuring the fluorescence decay rate of fluorescent probes. This provides critical information about molecular interactions, environmental changes, and localization within biological systems. However, creating high-resolution lifetime maps using conventional FLIM systems can be challenging, as it often requires extensive scanning that can significantly lengthen acquisition times. This issue is further compounded in three-dimensional (3D) imaging because it demands additional scanning along the depth axis. To tackle this challenge, we developed a computational imaging technique called light-field tomographic FLIM (LIFT-FLIM). Our approach allows for the acquisition of volumetric fluorescence lifetime images in a highly data-efficient manner, significantly reducing the number of scanning steps required compared to conventional point-scanning or line-scanning FLIM imagers. Moreover, LIFT-FLIM enables the measurement of high-dimensional data using low-dimensional detectors, which are typically low cost and feature a higher temporal bandwidth. We demonstrated LIFT-FLIM using a linear single-photon avalanche diode array on various biological systems, showcasing unparalleled single-photon detection sensitivity. Additionally, we expanded the functionality of our method to spectral FLIM and demonstrated its application in high-content multiplexed imaging of lung organoids. LIFT-FLIM has the potential to open up broad avenues in both basic and translational biomedical research.

Exploring Digital Zoom in Scanning Electron Microscopy

Abstract The concept of zoom in scanning electron microscopy (SEM) once referred to increasing the magnification of the sample by reducing the scanned field size. However, with the introduction of digital technology, the concept of zooming has become a little more complicated. Historically, older SEMs operated at a 512 × 512-pixel lateral resolution (262 kB pixel density) at TV scan rates. Today's newest SEMs can achieve up to 32k × 24k (4:3 ratio) pixel lateral resolution, equating to an 805 MB pixel density. In this article, we explore the factors involved in selecting an appropriate frame grabber size for SEM imaging. Specifically, we will delve into the implications of pixel density, including its impact on digital zoom capabilities, image file sizes, and overall image acquisition times.

Motion-free high-resolution on-chip microscopy using LED matrix

Lensless microscopy is an imaging technique that allows high-resolution imaging over a large field of view with a cost-effective design. Conventional lensless microscopy often utilizes multi-height phase retrieval and pixel-super-resolution algorithms to reconstruct high-resolution images, requiring mechanical stages for three-dimensional relative movements between a light source, camera, and sample. However, the excessive use of stages inevitably increases the bulkiness of the system and extends the image acquisition time. Here, we propose a motion-free lensless microscope that incorporates an RGB LED matrix array. A high-resolution holographic image is reconstructed from subpixel-shifted color images obtained with LED illuminations without any mechanical movement. Using a prototype system, we have demonstrated a spatial-bandwidth product of 30 megapixels with a resolution of 0.87 µm and a field of view of 24 mm2. The usability of the proposed method has been further tested for histopathologic examination. Our system features a compact and high-performance design with inexpensive optoelectronic elements, a conventional CMOS sensor and an LED matrix, which are well-aligned with the original design motivation of lensless imaging methods.

The effect of multiband sequences on statistical outcome measures in functional magnetic resonance imaging using a gustatory stimulus

Recent technical developments have led to the invention of multiband functional magnetic resonance imaging (fMRI) sequences that allow for faster sampling rates. However, some studies have highlighted problems with these sequences, leading to a decreased temporal signal-to-noise ratio (tSNR). In addition, this temporal noise may interfere with detecting reward-related responses in mesolimbic regions. The blood-oxygen-level-dependent signal utilized in the majority of fMRI measurements is relatively slow. Furthermore, the cerebral response to gustatory stimuli would also be relatively slow. Therefore, given the temporal noise issues with multiband sequences, it is unclear whether multiband sequences are necessary for fMRI studies using gustatory stimuli. We thus conducted an fMRI experiment using a gustatory stimulus to investigate the effects of multiband sequences and increased sampling rates on statistical outcome measures. A single-band sequence with a repetition time (TR) of 2 s of phantom fMRI data and gustatory fMRI data from the gustatory regions exhibited the highest tSNR, although the tSNR of this sequence of gustatory fMRI was not statistically different from tSNR of multiband sequences with a TR of 2 s in any of the selected region of interests. Conventional general linear model analysis of fMRI showed that single-band sequences are more advantageous than multiband sequences for detecting brain responses to gustatory stimuli in the primary gustatory cortex. In addition, a Bayesian data comparison showed that data derived from a single-band sequence with a TR of 2 s was optimal for inferring neuronal connectivity in gustatory processing. Therefore, a conventional single-band sequence with a TR of 2 s is more appropriate for fMRI with gustatory stimuli. Image acquisition sequences should be selected aligned with the study objectives and target brain regions.

Fast X-ray trimodal computed tomography with an improved diffraction enhanced imaging method

The rapid acquisition of projective images with low radiation dose is essential in computed tomography with diffraction enhanced imaging to extract absorption, refraction, and scattering images from weakly absorbing specimens. This plays a critical role in applying diffraction enhanced imaging to biological and medical imaging. In this study, an improved diffraction enhanced imaging method is proposed to rapidly implement X-ray trimodal computed tomography. This method positions the sample in a specific region near the analyzer crystal, allowing the simultaneous acquisition of transmitted and diffracted images in a single exposure. When combined with computed tomography, three different sample properties, known as absorption, phase and scattering, can be reconstructed simultaneously by collecting the projective images as the sample rotates from 0° to 360°. The experimental results demonstrate that this straightforward method enables the quantitative extraction of sample information and simplifies the data acquisition procedure in computed tomography with diffraction enhanced imaging. Therefore, this novel method offers the advantages of straightforward imaging device, rapid CT data acquisition, low radiation dose and more comprehensive sample information extraction.

Design of a Novel Microlens Array and Imaging System for Light Fields.

Light field cameras are unsuitable for further acquisition of high-quality images due to their small depth of field, insufficient spatial resolution, and poor imaging quality. To address these issues, we proposed a novel four-focal-square microlens and light field system. A square aspheric microlens array with four orthogonal focal lengths was designed, in which the aperture of a single lens was 100 μm. The square arrangement improves pixel utilization, the four focal lengths increase the depth of field, and the aspheric improves image quality. The simulations demonstrate pixel utilization rates exceeding 90%, depth-of-field ranges 6.57 times that of a single focal length, and image quality is significantly improved. We have provided a potential solution for improving the depth of field and image quality of the light field imaging system.

Design of an FPGA-Based Controller for Fast Scanning Probe Microscopy.

Atomic-scale imaging using scanning probe microscopy is a pivotal method for investigating the morphology and physico-chemical properties of nanostructured surfaces. Time resolution represents a significant limitation of this technique, as typical image acquisition times are on the order of several seconds or even a few minutes, while dynamic processes-such as surface restructuring or particle sintering, to be observed upon external stimuli such as changes in gas atmosphere or electrochemical potential-often occur within timescales shorter than a second. In this article, we present a fully redesigned field programmable gate array (FPGA)-based instrument that can be integrated into most commercially available standard scanning probe microscopes. This instrument not only significantly accelerates the acquisition of atomic-scale images by orders of magnitude but also enables the tracking of moving features such as adatoms, vacancies, or clusters across the surface ("atom tracking") due to the parallel execution of sophisticated control and acquisition algorithms and the fast exchange of data with an external processor. Each of these measurement modes requires a complex series of operations within the FPGA that are explained in detail.

Using of high spatial resolution images to evaluate the thematic accuracy of land use and occupation maps with the Kappa index

The objective from this article evaluates the thematic quality of automatic mappings with supervised classification for land use and land cover, using high spatial resolution satellite images as "ground truth". This consider the advancement of remote sensing technologies has enabled the acquisition of satellite images with various spatial resolutions, which are essential for thematic mapping and automatic classifiers in the context of land use and land cover mapping. In fact, the wide availability of high, medium, and low spatial resolution satellite images has significantly optimized the time and resources required by using more accurate classifiers during data processing. The image used for verification of this paper was GeoEye, with a spatial resolution of 0.5m, dated October 2023. The images submitted to the automatic classifier were Sentinel-2A with a spatial resolution of 10m and Planet with a spatial resolution of 5m, both from the same satellite pass period (October 2023) over the study area, aiming to avoid seasonal and phenological variations in vegetation, as well as changes in the environment due to anthropogenic intervention. The classification method adopted was Maximum Likelihood (MAXVER). The classification accuracy was rigorously evaluated to ensure the reliability of the results using the Kappa index, assessing the agreement between the observed and expected classifications. Based on the methods presented, the set of mapped classes in this study showed good accuracy for the Planet image and very good accuracy for the Sentinel image.

Priority radiomic parameters for computed tomography of head and neck malignancies: A systematic review

BACKGROUND: Radiomics is the newest and most promising direction in modern radiographic diagnostics. The number of head and neck cancer studies employing radiomics is increasing annually. A systematic review of recent publications (2021–2023) on computed tomography (CT) of head and neck malignancies was performed. AIM: To present systematized data on parameters for radiomic analysis for head and neck malignancies identified by CT data. MATERIALS AND METHODS: The literature search was carried out in PubMed. The basic characteristics of the selected articles were extracted, and their quality was assessed using RQS 2.0 and the modified QUADAS-CAD questionnaire. The reproducibility level of radiomic parameters selected for predictive models in different studies was assessed. Eleven articles were selected for the review. In most cases, a high risk of systematic error associated with data imbalance in terms of demographic parameters and level of pathologies was noted. RESULTS: The range of RQS 2.0 scores for the included articles varied from 19.44% to 50.00% of the maximum possible score. The decreasing research quality was mainly caused by the lack of external result validation (73% of the analyzed articles) and data accessibility and transparency (82%). Inter-study reproducibility of radiomic parameters was low owing to the wide variety of techniques used for image acquisition, image post-processing, extraction, and statistical processing of radiomic parameters. CONCLUSION: A set of stable radiomic parameters must be successfully introduced into clinical practice. The standardization of radiomics method and creation of an open radiomics database are necessary for this purpose.