Hybrid Intelligence-Driven Medical Image Recognition for Remote Patient Diagnosis in Internet of Medical Things

IEEE Journal of Biomedical and Health Informatics, Год журнала: 2021, Номер 26(12), С. 5817 - 5828

Опубликована: Дек. 31, 2021

In ear of smart cities, intelligent medical image recognition technique has become a promising way to solve remote patient diagnosis in IoMT. Although deep learning-based approaches have received great development during the past decade, explainability always acts as main obstacle promote higher levels. Because it is hard clearly grasp internal principles learning models. contrast, conventional machine (CML)-based methods are well explainable, they give relatively certain meanings parameters. Motivated by above view, this paper combines with CML, and proposes hybrid intelligence-driven framework On one hand, convolution neural network utilized extract abstract features for initial images. other CML-based techniques employed reduce dimensions extracted construct strong classifier that output results. A real dataset about pathologic myopia selected establish simulative scenario, order assess proposed framework. Results reveal proposal improves accuracy two three percent.

Язык: Английский

---

Robust Zero Watermarking Algorithm for Medical Images Based on Improved NasNet-Mobile and DCT

Electronics, Год журнала: 2023, Номер 12(16), С. 3444 - 3444

Опубликована: Авг. 15, 2023

In the continuous progress of mobile internet technology, medical image processing technology is also always being upgraded and improved. this field, digital watermarking significant provides a strong guarantee for information security. This paper offers robustness zero strategy pictures based on an Improved NasNet-Mobile convolutional neural network discrete cosine transform (DCT) to address lack existing algorithms. First, structure pre-training adjusted by using fully connected layer with 128 output regression instead original Softmax classification layer, thus generating output, whereby features are extracted from images migration learning. Migration learning then performed modified obtain trained network, which used extract features, finally subjected DCT low frequency data, perceptual hashing algorithm processes data 32-bit binary feature vector. Before performing watermark embedding, encrypted chaos mapping increase Next, technique allow embed without changing contained in image. The experimental findings demonstrate algorithm’s resistance both conventional geometric assaults. some practical application value realm medicine when compared other approaches.

Язык: Английский

---

Deep Learning Based Image Processing for Robot Assisted Surgery: A Systematic Literature Survey

IEEE Access, Год журнала: 2022, Номер 10, С. 122627 - 122657

Опубликована: Янв. 1, 2022

The recent advancements in the surging field of Deep Learning (DL) have revolutionized every sphere life, and healthcare domain is no exception. enormous success DL models, particularly with image data, has led to development image-guided Robot Assisted Surgery (RAS) systems. By large, number studies concerning image-driven computer assisted surgical systems using increased exponentially. Additionally, contemporary availability datasets also boosted applications RAS. Inspired by latest trends contributions surgery, this literature survey presents a summarized analysis innovations RAS After thorough review, sum 184 articles are selected grouped into four categories, based on relevancy task articles, comprising 1) Surgical Tools, 2) Processes, 3) Surveillance, 4) Performance. discusses publicly available highlights basics models. Furthermore, legal, ethical, technological challenges together intuitive predictions recommendations related autonomous presented. study reveals that Convolutional Neural Network (CNN) most widely adopted architecture, whereas, JIGSAWS employed dataset suggests fusing kinematic data along which produces better accuracy precision, gesture trajectory segmentation tasks. CNN Long Short Term Memory (LSTM) networks shown remarkable performance, however, authors recommend employing these gigantic architectures only when simpler models failed produce satisfactory results. despite their limitations, time cost effective yield considerable outcomes even smaller datasets.

Язык: Английский

---

Pulmonary arteries segmentation from CT images using PA‐Net with attention module and contour loss

Medical Physics, Год журнала: 2023, Номер 50(8), С. 4887 - 4898

Опубликована: Фев. 8, 2023

Pulmonary embolism is a kind of cardiovascular disease that threatens human life and health. Since pulmonary exists in the artery, improving segmentation accuracy artery key to diagnosis embolism. Traditional medical image methods have limited effectiveness segmentation. In recent years, deep learning been gradually adopted solve complex problems field segmentation.Due irregular shape adjacent-complex tissues, existing based on needs be improved. Therefore, purpose this paper develop network, which can obtain higher further improve effect.In study, performance from network model loss function improved, proposing (PA-Net) segment region 2D CT images. Reverse Attention edge attention are used enhance expression ability boundary. addition, better use feature information, channel module introduced decoder highlight important features suppress unimportant channels. Due blurred boundaries, pixels near boundaries may difficult segment. new contour active proposed study target by assigning dynamic weights false positive negative regions accurately predict boundary structure.The experimental results show method significantly improved comparison with state-of-the-art methods, Dice coefficient 0.938 ± 0.035, also confirmed 3D reconstruction results.Our structure. This development will provide possibility for rapid diseases such as Code available at https://github.com/Yuanyan19/PA-Net.

Язык: Английский

---

StratMed: Relevance stratification between biomedical entities for sparsity on medication recommendation

Knowledge-Based Systems, Год журнала: 2023, Номер 284, С. 111239 - 111239

Опубликована: Ноя. 29, 2023

Язык: Английский

---

Self-supervised visual–textual prompt learning for few-shot grading of gastric intestinal metaplasia

Knowledge-Based Systems, Год журнала: 2024, Номер 301, С. 112303 - 112303

Опубликована: Авг. 3, 2024

Язык: Английский

---

Recurrent generative adversarial networks for unsupervised WCE video summarization

Knowledge-Based Systems, Год журнала: 2021, Номер 222, С. 106971 - 106971

Опубликована: Март 20, 2021

Язык: Английский

---

Artificial intelligence in ophthalmopathy and ultra-wide field image: A survey

Expert Systems with Applications, Год журнала: 2021, Номер 182, С. 115068 - 115068

Опубликована: Апрель 20, 2021

Язык: Английский

---

Generative Adversarial Network Based Automatic Segmentation of Corneal Subbasal Nerves on In Vivo Confocal Microscopy Images

Translational Vision Science & Technology, Год журнала: 2021, Номер 10(6), С. 33 - 33

Опубликована: Май 26, 2021

Purpose: In vivo confocal microscopy (IVCM) is a noninvasive, reproducible, and inexpensive diagnostic tool for corneal diseases. However, widespread effortless image acquisition in IVCM creates serious analysis workloads on ophthalmologists, neural networks could solve this problem quickly. We have produced novel deep learning algorithm based generative adversarial (GANs), we compare its accuracy automatic segmentation of subbasal nerves images with fully convolutional network (U-Net) method. Methods: collected from 85 subjects. U-Net GAN-based methods were trained tested under the supervision three clinicians nerves. Nerve results GAN U-Net-based compared by using Pearson's R correlation, Bland-Altman analysis, receiver operating characteristics (ROC) statistics. Additionally, different noises applied to evaluate performances algorithms biomedical imaging. Results: The demonstrated similar correlation U-Net. method showed significantly higher ROC curves. performance deteriorated noises, especially speckle noise, GAN. Conclusions: This study first application images. than nerve images, patient-acquired noise can be used as facilitating ophthalmology clinics. Translational Relevance: Generative are emerging models medical processing, which important clinical tools rapid

Язык: Английский

---

MDFA-Net: Multiscale dual-path feature aggregation network for cardiac segmentation on multi-sequence cardiac MR

Knowledge-Based Systems, Год журнала: 2021, Номер 215, С. 106776 - 106776

Опубликована: Янв. 15, 2021

Язык: Английский

---