Medical Imaging Applications Research Articles

Emerging New-Generation Semiconductor Single Crystals of Metal Halide Perovskites for Radiation Detection

Radiation detection uses semiconductor materials to convert high-energy photons into charge (direct detection) or low-energy photons (indirect detection), and it has a wide range of applications in nuclear physics, medical imaging, astronomical detection, homeland security, and other fields. Metal halide perovskites have the advantages of high frequency number, high carrier mobility, high defect tolerance, low defect density, adjustable band gap, and fast light response, and they have wide application prospects in the field of radiation detection. However, the research is still in its infancy stage, and it is far from meeting the requirements of industrial application. This paper focuses on the advantages of metal halide perovskite single-crystal materials in both semiconductors-based direct conversion detection and scintillator-based indirect detection as well as the latest progress in this promising field. This paper not only introduces the latest application of lead halide perovskite monocrystalline materials in high-energy electromagnetic radiation detection (X-ray and γ-rays), but it also introduces the latest development of α-particle/β-particle/neutron detection. Finally, this paper points out the challenges and future prospects of metal halide perovskite single-crystal materials in radiation detection.

Security of Magnetic Resonance Medical Images using Region-based Lossless Image Compression in Healthcare Information Systems

Purpose The purpose of region-based medical image compression is to optimize the compression process by focusing on specific regions of interest within medical images. Unlike traditional compression methods that treat the entire image uniformly, region-based compression techniques identify and prioritize certain areas or regions within the image that are deemed more diagnostically significant or relevant. By allocating more resources to compressing these critical regions while reducing compression in less important areas, region-based compression methods aim to achieve higher compression efficiency while preserving diagnostic quality. This approach is particularly valuable in medical imaging, where accurate representation of anatomical structures or pathological findings is paramount for clinical diagnosis and decision-making. Region-based compression can help reduce storage requirements, transmission bandwidth, and processing time without compromising the diagnostic integrity of medical images, thereby facilitating more efficient healthcare delivery and telemedicine applications. Methods In this study, we utilized distortion-limiting compression techniques to optimize the compression process for specific regions within medical images. We employed lossless scalable RBC (Region-Based Compression) using Discrete Wavelet Transform (DWT) for Digital Imaging and Communication in Medicine (DICOM) images. The initial step involved medical image pre-processing, followed by segmentation to separate the image into regions of interest (ROI) and non-ROI. Compression techniques were then applied to reduce network bandwidth and storage requirements. Fractal lossy compression was employed for the non-ROI portion, while context-tree weighting lossless compression was proposed for the ROI portion, effectively compressing the image while rejecting noisy background elements. During decompression, the original medical image can be reconstructed using the reverse process. This approach optimizes storage and transmission efficiency while preserving diagnostic integrity in medical imaging applications. Results The experiment involved testing various medical images, and the proposed method outperformed previous techniques in terms of results. According to the findings, the improvement in Peak Signal-to-Noise Ratio (PSNR) over current techniques reached up to 24.23 dB compared to the Joint Photographic Experts Group (JPEG). Additionally, it achieved up to 12.22 dB improvement compared to other transform approaches. These significant enhancements prompted the development of a web and mobile platform for compressing and sending medical images, particularly microscopic ones, in real time. Conclusion This research focuses on employing wavelet transform techniques to compress the Region of Interest (ROI) within medical images. This ROI-based compression approach is particularly valuable as it retains essential diagnostic information while reducing the overall file size. Such a technique holds significant promise for telemedicine systems, especially in rural regions where network resources may be limited or constrained. By selectively compressing the most diagnostically relevant areas of medical images, this approach ensures that critical information is preserved while optimizing data transmission and storage efficiency. This can ultimately enhance access to medical imaging services and facilitate remote diagnosis and treatment in underserved areas with limited network infrastructure.

ValveVision AI - a multimodal language model for qualitative reporting of the aortic valve in echocardiography

Abstract Background The precise assessment of the aortic valve via echocardiography is critical for early detection and management of aortic valve diseases. Until recently, previous studies have examined machine learning models to estimate individual measurements and severity of aortic stenosis (AS) from echocardiographic images. These image processing algorithms, while precise in their narrow focus, fall short in mirroring the holistic and interconnected clinical judgment typical of human echocardiographers in producing a qualitative report. Large language models (LLMs), particularly image-to-text multi-modal LLMs, are a fundamental advance in the field of deep learning with implications for a host of applications in medical imaging. They promise to encapsulate not just discrete data points typical in traditional machine learning, but also the complex contextual interrelations in clinical diagnosis. Methods In this study, a large-scale heterogeneous database of echocardiographic images containing over 90,681 studies with textual descriptors of the aortic valve was used to train a single, image-to-text multimodal LLM called ValveVision AI. The ground truth textual summaries were drafted by level III echocardiographers in a clinical setting between 2015-2020. BLEU and ROUGE score was calculated. The models were retrospectively assessed on a holdout dataset. Reviewing physicians compared the generated summary to the ground truth and binarily agreed or rejected it. Receiver Operating Characteristics (ROC) for distinct pathologies were also assessed (Figure I). Results ValveVision AI performed with a BLEU score of 0.45 and a ROUGE score of 0.49. The performance of the model in reporting on classification of moderate/severe vs none/mild AS in concordance with the validation protocol described above achieved a specificity of 91.98% and a sensitivity of 83.89%, along with more precise qualitative description. Qualitatively, the model exhibited the capability of zero-shot learning in certain instances, however, this result remains an area of exploration. Conclusion This study represents to our knowledge, the first attempt at an image-representation to text-tokenizer deep learning model architecture to mimic the thought and subtlety of echocardiographic qualitative analysis of the aortic valve. The results suggest that this multimodal LLM has sufficient accuracy to create a preliminary textual summary of the aortic valve that, if paired with a point of care ultrasound (POCUS) device in a primary care setting, may facilitate case triage, increase efficiency, and determine a more precise care pathway for patients.

Transfer learning classification of suspicious lesions on breast ultrasound: is there room to avoid biopsies of benign lesions?

BackgroundBreast cancer (BC) is the most common malignancy in women and the second cause of cancer death. In recent years, there has been a strong development in artificial intelligence (AI) applications in medical imaging for several tasks. Our aim was to evaluate the potential of transfer learning with convolutional neural networks (CNNs) in discriminating suspicious breast lesions on ultrasound images.MethodsTransfer learning performances of five different CNNs (Inception V3, Xception, Densenet121, VGG 16, and ResNet50) were evaluated on a public and on an institutional dataset (526 and 392 images, respectively), customizing the top layers for the specific task. Institutional images were contoured by an expert radiologist and processed to feed the CNNs for training and testing. Postimaging biopsies were used as a reference standard for classification. The area under the receiver operating curve (AUROC) was used to assess diagnostic performance.ResultsNetworks performed very well on the public dataset (AUROC 0.938–0.996). The direct generalization to the institutional dataset resulted in lower performances (max AUROC 0.676); however, when tested on BI-RADS 3 and BI-RADS 5 only, results were improved (max AUROC 0.792). Good results were achieved on the institutional dataset (AUROC 0.759–0.818) and, when selecting a threshold of 2% for classification, a sensitivity of 0.983 was obtained for three of five CNNs, with the potential to spare biopsy in 15.3%–18.6% of patients.ConclusionIn conclusion, transfer learning with CNNs may achieve high sensitivity and might be used as a support tool in managing suspicious breast lesions on ultrasound images.Relevance statementTransfer learning is a powerful technique to exploit the performances of well-trained CNNs for image classification. In a clinical scenario, it might be useful for the management of suspicious breast lesions on breast ultrasound, potentially sparing biopsy in a non-negligible number of patients.Key PointsProperly trained CNNs with transfer learning are highly effective in differentiating benign and malignant lesions on breast ultrasound.Setting clinical thresholds increased sensitivity.CNNs might be useful as support tools in managing suspicious lesions on breast ultrasound.Graphical

Deep transfer learning with improved crayfish optimization algorithm for oral squamous cell carcinoma cancer recognition using histopathological images

Oral Squamous Cell Carcinoma (OSCC) causes a severe challenge in oncology due to the lack of diagnostic devices, leading to delays in detecting the disorder. The OSCC diagnosis through histopathology demands a pathologist expert because the cellular presentation is variable and highly complex. Existing diagnostic approaches for OSCC have specific efficiency and accuracy restrictions, highlighting the necessity for more reliable techniques. The increase of deep neural networks (DNN) model and their applications in medical imaging have been instrumental in disease diagnosis and detection. Automatic detection systems using deep learning (DL) approaches show tremendous promise in investigating medical imagery with speed, efficiency, and accuracy. In terms of OSCC, this system allows the diagnostic method to be streamlined, facilitating earlier diagnosis and enhancing survival rates. Automatic analysis of histopathological image (HI) can assist in accurately detecting and identifying tumorous tissue, reducing diagnostic turnaround times and increasing the efficacy of pathologists. This study presents a Squeeze-Excitation with Hybrid Deep Learning for Oral Squamous Cell Carcinoma Recognition (SEHDL-OSCCR) on HIs. The presented SEHDL-OSCCR technique mainly focuses on detecting oral cancer (OC) using hybrid DL models. The bilateral filtering (BF) technique is initially used to remove the noise. Next, the SEHDL-OSCCR technique employs the SE-CapsNet model to recognize the feature extractors. An improved crayfish optimization algorithm (ICOA) technique is utilized to improve the performance of the SE-CapsNet model. At last, the classification of the OSCC technique is performed by employing a convolutional neural network with a bidirectional long short-term memory (CNN-BiLSTM) model. The simulation results obtained using the SEHDL-OSCCR technique are investigated using a benchmark medical image dataset. The experimental validation of the SEHDL-OSCCR technique illustrated a greater accuracy outcome of 98.75% compared to recent approaches.

Deep Learning Approach for Arm Fracture Detection Based on an Improved YOLOv8 Algorithm

Artificial intelligence (AI)-assisted computer vision is an evolving field in medical imaging. However, accuracy and precision suffer when using the existing AI models for small, easy-to-miss objects such as bone fractures, which affects the models’ applicability and effectiveness in a clinical setting. The proposed integration of the Hybrid-Attention (HA) mechanism into the YOLOv8 architecture offers a robust solution to improve accuracy, reliability, and speed in medical imaging applications. Experimental results demonstrate that our HA-modified YOLOv8 models achieve a 20% higher Mean Average Precision (mAP 50) and improved processing speed in arm fracture detection.

Enhancing the Sensitivity and Spatial Imaging Resolution of a Hybrid X-Ray Imaging Screen via Energy Transfer at the ZnS (Ag) and a Thermally Activated Delayed Fluorescence Interface.

Novel scintillation materials have played an indispensable role in the recent remarkable progress witnessed for X-ray imaging technology. Herein, a high-performance X-ray scintillation screen was developed based on a highly efficient hybrid system combining inorganic ZnS (Ag) with thermally activated delayed fluorescence (TADF) scintillator materials via an interfacial energy transfer (EnT) mechanism. ZnS (Ag) has a high X-ray absorption capacity and functions as the initial layer for efficiently converting high-energy X-ray photons into low-energy visible light (acting as a sensitizer) while also serving as an energy donor. The TADF component, on the contrary, is an energy acceptor and forms an active scintillating layer. By harnessing TADF chromophores that can efficiently capture both singlet and triplet excitons, our composite material offers a remarkable spatial imaging resolution of 24 line pairs per millimeter, surpassing those of the majority of existing organic and inorganic scintillators. Further, our interfacial energy transfer strategy effectively amplifies the radioluminescence intensity of the TADF scintillator by a factor of 75, offering an outstanding light yield of 38,000 photons/MeV. This advancement represents a remarkable breakthrough in organic X-ray scintillation technology and is a notable achievement within the X-ray imaging field, paving the way for novel applications in medical imaging and security inspection.

ShiftTransUNet: An Efficient Deep Learning Model for Medical Image Segmentation Using ShiftViT Framework

Deep learning has significantly advanced the field of medical image segmentation. However, the complexity of network structures often leads to high computational demands, limiting their practical efficiency. To enhance the efficiency of image segmentation, this paper introduces an innovative, concise, and lightweight deep learning network. First, to reduce model complexity, we replaced the attention mechanism in the traditional vision transformer (ViT) structure with a shift operation, creating the ShiftViT architecture. This substitution significantly decreased computation and the number of parameters while preserving model performance. Second, to retain and enhance fine-grained features and facilitate more precise information transfer across different layers, we employed a full-scale progressive skip connection strategy. This approach effectively integrates multi-scale feature information, further enhancing model performance. Additionally, to further reduce network complexity, inspired by the independence of probabilities, we opted for depth-wise separable convolution over traditional convolution. This enhances the relative independence between layers. Together, these modifications achieved superior segmentation results on both the Synapse and Automated Cardiac Diagnostic Challenge (ACDC) datasets compared to mainstream models, with substantial advantages in terms of computational efficiency and parameter count. The proposed approach represents an effective solution for medical image applications with limited computational resources and holds great promise for clinical practice.

Assessing the feasibility of ChatGPT-4o and Claude 3-Opus in thyroid nodule classification based on ultrasound images.

Large language models (LLMs) are pivotal in artificial intelligence, demonstrating advanced capabilities in natural language understanding and multimodal interactions, with significant potential in medical applications. This study explores the feasibility and efficacy of LLMs, specifically ChatGPT-4o and Claude 3-Opus, in classifying thyroid nodules using ultrasound images. This study included 112 patients with a total of 116 thyroid nodules, comprising 75 benign and 41 malignant cases. Ultrasound images of these nodules were analyzed using ChatGPT-4o and Claude 3-Opus to diagnose the benign or malignant nature of the nodules. An independent evaluation by a junior radiologist was also conducted. Diagnostic performance was assessed using Cohen's Kappa and receiver operating characteristic (ROC) curve analysis, referencing pathological diagnoses. ChatGPT-4o demonstrated poor agreement with pathological results (Kappa = 0.116), while Claude 3-Opus showed even lower agreement (Kappa = 0.034). The junior radiologist exhibited moderate agreement (Kappa = 0.450). ChatGPT-4o achieved an area under the ROC curve (AUC) of 57.0% (95% CI: 48.6-65.5%), slightly outperforming Claude 3-Opus (AUC of 52.0%, 95% CI: 43.2-60.9%). In contrast, the junior radiologist achieved a significantly higher AUC of 72.4% (95% CI: 63.7-81.1%). The unnecessary biopsy rates were 41.4% for ChatGPT-4o, 43.1% for Claude 3-Opus, and 12.1% for the junior radiologist. While LLMs such as ChatGPT-4o and Claude 3-Opus show promise for future applications in medical imaging, their current use in clinical diagnostics should be approached cautiously due to their limited accuracy.

Artificial Intelligence in Medicine

Artificial intelligence in medicine refers to the use of machine learning models to help process medical data and provide medical professionals with important insights, improving health outcomes and patient experience. Thanks to recent advances in computer science and informatics, artificial intelligence (AI) is rapidly becoming an integral part of modern healthcare. Therefore, artificial intelligence algorithms and other AI-powered applications are now used to support medical professionals in clinical settings and in ongoing research. There are several applications of Artificial intelligence in medicine, including applications to help detect and diagnose diseases; applications to treat diseases with the help of an AI-powered virtual assistant; AI applications in medical imaging; applications to increase the efficiency of clinical trials; and applications to accelerate drug development. The benefits of Artificial intelligence in medicine can be summarized in providing informed patient care, reducing errors, reducing care costs, and increasing doctor-patient engagement.

Double-foci acoustic metamaterial Luneburg lens

In this study, an acoustic Luneburg lens is proposed to achieve double foci. High energy density between the two foci can be obtained, which enables the realization of an ultra long acoustic jet between the two foci. The acoustic Luneburg lens is designed based on the gradient index acoustic metamaterial with lattice unit cells. Analytical and numerical analyses were performed to investigate the performance of double foci Luneburg lens. Results show that the double foci can be realized with acoustic jet length goes up to 30 times compared with the wavelength, which has multiple potential applications in medical imaging and energy harvesting.

Assessing the Capacity of a Denoising Diffusion Probabilistic Model to Reproduce Spatial Context.

Diffusion models have emerged as a popular family of deep generative models (DGMs). In the literature, it has been claimed that one class of diffusion models-denoising diffusion probabilistic models (DDPMs)-demonstrate superior image synthesis performance as compared to generative adversarial networks (GANs). To date, these claims have been evaluated using either ensemble-based methods designed for natural images, or conventional measures of image quality such as structural similarity. However, there remains an important need to understand the extent to which DDPMs can reliably learn medical imaging domain-relevant information, which is referred to as 'spatial context' in this work. To address this, a systematic assessment of the ability of DDPMs to learn spatial context relevant to medical imaging applications is reported for the first time. A key aspect of the studies is the use of stochastic context models (SCMs) to produce training data. In this way, the ability of the DDPMs to reliably reproduce spatial context can be quantitatively assessed by use of post-hoc image analyses. Error-rates in DDPM-generated ensembles are reported, and compared to those corresponding to other modern DGMs. The studies reveal new and important insights regarding the capacity of DDPMs to learn spatial context. Notably, the results demonstrate that DDPMs hold significant capacity for generating contextually correct images that are 'interpolated' between training samples, which may benefit data-augmentation tasks in ways that GANs cannot.

Adding barium sulfate (BaSO4) to fly ash geopolymer increases its compressive strength as X-ray shielding for medical imaging applications

BackgroundGeopolymers, a novel cementitious material, have the potential for reducing carbon dioxide wastes resulting from the manufacture of cement. PurposeThis study presents an experimental inquiry conducted to produce a fly ash geopolymer mortar with a good of compressive strength of the batch. MethodsActivation of the mortar is accomplished through the use of sodium hydroxide and sodium silicate. The mortar is created from a mixture of fly ash and sand. In order to determine the maximum load that a material is capable of bearing before experiencing failure, the compressive strength test was utilized. ResultsAccording on the findings of the compressive strength testing device, the fly ash geopolymer with 15 % barium sulfate (BaSO4) demonstrates the highest compressive strength, which is measured at 56 MPa. When compared to the results obtained from cement mortar, this one is twice as high, which indicates that the strength improvement factor is 2. This study's findings highlighted the necessity of employing fly ash geopolymer that contains 15 % BaSO4 for the purpose of providing shielding protection. Conclusion: Therefore, in terms of its performance, fly ash geopolymer is superior to cement mortar. This is due to the fact that it is manufactured with a high compressive strength.

MEMS Micromirror Actuation Techniques: A Comprehensive Review of Trends, Innovations, and Future Prospects

Micromirrors have recently emerged as an essential component in optical scanning technology, attracting considerable attention from researchers. Their compact size and versatile capabilities, such as light steering, modulation, and switching, are leading them as potential alternatives to traditional bulky galvanometer scanners. The actuation of these mirrors is critical in determining their performance, as it contributes to factors such as response time, scanning angle, and power consumption. This article aims to provide a thorough exploration of the actuation techniques used to drive micromirrors, describing the fundamental operating principles. The four primary actuation modalities—electrostatic, electrothermal, electromagnetic, and piezoelectric—are thoroughly investigated. Each type of actuator’s operational principles, key advantages, and their limitations are discussed. Additionally, the discussion extends to hybrid micromirror designs that combine two types of actuation in a single device. A total of 208 closely related papers indexed in Web of Science were reviewed. The findings indicate ongoing advancements in the field, particularly in terms of size, controllability, and field of view, making micromirrors ideal candidates for applications in medical imaging, display projections, and optical communication. With a comprehensive overview of micromirror actuation strategies, this manuscript serves as a compelling resource for researchers and engineers aiming to utilize the appropriate type of micromirror in the field of optical scanning technology.

Enhancing the X-ray Sensitivity of Cs2AgBiBr6 Double Perovskite Single Crystals through Cation Engineering.

Owing to their outstanding optoelectronic properties, halide perovskite (HP) materials have been employed in a wide range of applications, including solar cells, light-emitting devices, and X-ray detectors. Among them, lead-free double HPs are characterized by enhanced stability and reduced toxicity compared with lead-based alternatives. Cs2AgBiBr6, in particular, has emerged as a promising candidate for direct X-ray detection. The detection sensitivity, on the other hand, cannot yet compete with that of lead-containing perovskites. Developing schemes to improve X-ray detection efficiency is critical for reducing radiation exposure in medical imaging applications. Here, we investigate the potential of controlled doping and cation substitution with either lanthanides or small organic cations to improve the X-ray detection performance of Cs2AgBiBr6. Our findings reveal that by growing the perovskite in a slightly Bi-poor and Eu-rich environment, the X-ray sensitivity significantly increases 7-fold (from 17 to 120 μC Gyair -1 cm-2) and simultaneously improves the phototo-dark current ratio (from 2.5 to 29). Additionally, Cs-site substitution with imidazolium remarkably enhances the sensitivity over 10-fold (180 μC Gyair -1 cm-2), and ammonium enhances the phototo-dark current ratio to 37. Terahertz photoconductivity measurements reveal a positive correlation between enhanced X-ray sensitivity and improved charge transport properties (e.g., increased scattering time and, thus, carrier mobility) by doping. This study outlines straightforward strategies for boosting X-ray detection and fundamental photoconductivity in lead-free double HP, with potential implications for broader optoelectronic applications.

A Low-Power, High-Resolution Analog Front-End Circuit for Carbon-Based SWIR Photodetector

Carbon nanotube field-effect transistors (CNT-FETs) have shown great promise in infrared image detection due to their high mobility, low cost, and compatibility with silicon-based technologies. This paper presents the design and simulation of a column-level analog front-end (AFE) circuit tailored for carbon-based short-wave infrared (SWIR) photodetectors. The AFE integrates a Capacitor Trans-impedance Amplifier (CTIA) for current-to-voltage conversion, coupled with Correlated Double Sampling (CDS) for noise reduction and operational amplifier offset suppression. A 10-bit/125 kHz Successive Approximation analog-to-digital converter (SAR ADC) completes the signal processing chain, achieving rail-to-rail input/output with minimized component count. Fabricated using 0.18 μm CMOS technology, the AFE demonstrates a high signal-to-noise ratio (SNR) of 59.27 dB and an Effective Number of Bits (ENOB) of 9.35, with a detectable current range from 500 pA to 100.5 nA and a total power consumption of 7.5 mW. These results confirm the suitability of the proposed AFE for high-precision, low-power SWIR detection systems, with potential applications in medical imaging, night vision, and autonomous driving systems.

Paediatric computed tomography diagnostic reference levels in Africa: A systematic review.

Improvements in computed tomography (CT) technology in terms of image quality and reduction in absorbed dose have increased its applications in medical imaging. Diagnostic reference levels (DRLs) help to identify high radiation doses that are unusually delivered to patients undergoing exposure to ionising radiation. The aim of this review was to provide an overview of published studies by African researchers towards establishing paediatric CT DRLs in Africa. The search for articles was conducted using some relevant literature search engines including PubMed, Scopus, Science Direct, Google Scholar and Web of Science. Two reviewers were involved in the article selection process which involved a three-stage screening process of identifying; article titles, abstracts and full-test reading. One hundred and seventy-four articles were identified from the database, PubMed (30), Scopus (21), Google Scholar (53), Web of Science (25) and Science Direct (45). Fifty duplicated articles were excluded before screening. Twelve peer-reviewed articles were included in this study based on the inclusion criteria. DRL values in terms of computed tomography dose index volume of head for the age groupings 0-1, 1-5, 5-10 and 10-15 were 27, 36.6, 39.5 and 47.5 mGy while the dose length product values were 461.6, 664, 872 and 978 mGy.cm respectively. The DRLs were calculated as 75th percentile of the local DRLs reported by the 12 articles included in this review. This review has shown that only few of the African countries (19%) have published studies on paediatric CT DRLs. There were variations in the DRLs published by the various authors which indicate that harmonisation and standardisation of paediatric CT protocols is essential for the optimisation of paediatric doses.

Activation function optimization scheme for image classification

Activation function has a significant impact on the dynamics, convergence, and performance of deep neural networks. The search for a consistent and high-performing activation function has always been a pursuit during deep learning model development. Existing state-of-the-art activation functions are manually designed with human expertise except for Swish. Swish was developed using a reinforcement learning-based search strategy. In this study, we propose an evolutionary approach for optimizing activation functions specifically for image classification tasks, aiming to discover functions that outperform current state-of-the-art options. Through this optimization framework, we obtain a series of high-performing activation functions denoted as Exponential Error Linear Unit (EELU). The developed activation functions are evaluated for image classification tasks from two perspectives: (1) five state-of-the-art neural network architectures, such as ResNet50, AlexNet, VGG16, MobileNet, and Compact Convolutional Transformer, which cover computationally heavy to light neural networks, and (2) eight standard datasets, including CIFAR10, Imagenette, MNIST, Fashion MNIST, Beans, Colorectal Histology, CottonWeedID15, and TinyImageNet which cover from typical machine vision benchmark, agricultural image applications to medical image applications. Finally, we statistically investigate the generalization of the resultant activation functions developed through the optimization scheme. With a Friedman test, we conclude that the optimization scheme is able to generate activation functions that outperform the existing standard ones in 92.8% cases among 28 different cases studied, and −x⋅erf(e−x) is found to be the best activation function for image classification generated by the optimization scheme.

PelviNet: A Collaborative Multi-agent Convolutional Network for Enhanced Pelvic Image Registration.

PelviNet introduces a groundbreaking multi-agent convolutional network architecture tailored for enhancing pelvic image registration. This innovative framework leverages shared convolutional layers, enabling synchronized learning among agents and ensuring an exhaustive analysis of intricate 3D pelvic structures. The architecture combines max pooling, parametric ReLU activations, and agent-specific layers to optimize both individual and collective decision-making processes. A communication mechanism efficiently aggregates outputs from these shared layers, enabling agents to make well-informed decisions by harnessing combined intelligence. PelviNet's evaluation centers on both quantitative accuracy metrics and visual representations to elucidate agents' performance in pinpointing optimal landmarks. Empirical results demonstrate PelviNet's superiority over traditional methods, achieving an average image-wise error of 2.8 mm, a subject-wise error of 3.2 mm, and a mean Euclidean distance error of 3.0 mm. These quantitative results highlight the model's efficiency and precision in landmark identification, crucial for medical contexts such as radiation therapy, where exact landmark identification significantly influences treatment outcomes. By reliably identifying critical structures, PelviNet advances pelvic image analysis and offers potential enhancements for broader medical imaging applications, marking a significant step forward in computational healthcare.

Harnessing AI for precision tonsillitis diagnosis: a revolutionary approach in endoscopic analysis.

Diagnosing and treating tonsillitis pose no significant challenge for otolaryngologists; however, it can increase the infection risk for healthcare professionals amidst the coronavirus pandemic. In recent years, with the advancement of artificial intelligence (AI), its application in medical imaging has also thrived. This research is to identify the optimal convolutional neural network (CNN) algorithm for accurate diagnosis of tonsillitis and early precision treatment. Semi-supervised learning with pseudo-labels used for self-training was adopted to train our CNN, with the algorithm including UNet, PSPNet, and FPN. A total of 485 pharyngoscopic images from 485 participants were included, comprising healthy individuals (133 cases), patients with the common cold (295 cases), and patients with tonsillitis (57 cases). Both color and texture features from 485 images are extracted for analysis. UNet outperformed PSPNet and FPN in accurately segmenting oropharyngeal anatomy automatically, with average Dice coefficient of 97.74% and a pixel accuracy of 98.12%, making it suitable for enhancing the diagnosis of tonsillitis. The normal tonsils generally have more uniform and smooth textures and have pinkish color, similar to the surrounding mucosal tissues, while tonsillitis, particularly the antibiotic-required type, shows white or yellowish pus-filled spots or patches, and shows more granular or lumpy texture in contrast, indicating inflammation and changes in tissue structure. After training with 485 cases, our algorithm with UNet achieved accuracy rates of 93.75%, 97.1%, and 91.67% in differentiating the three tonsil groups, demonstrating excellent results. Our research highlights the potential of using UNet for fully automated semantic segmentation of oropharyngeal structures, which aids in subsequent feature extraction, machine learning, and enables accurate AI diagnosis of tonsillitis. This innovation shows promise for enhancing both the accuracy and speed of tonsillitis assessments.