Applied Sciences

Editorial

Jump to: Research, Review, Other

2 pages, 180 KiB

Open AccessEditorial

Special Issue on Artificial Intelligence in Medical Imaging: The Beginning of a New Era

by Cosimo Nardi

Appl. Sci. 2023, 13(20), 11562; https://doi.org/10.3390/app132011562 - 23 Oct 2023

Cited by 1 | Viewed by 1022

Abstract

Artificial intelligence (AI) can be considered the real revolution of the 21st century [...] Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

Research

Jump to: Editorial, Review, Other

21 pages, 3694 KiB

Open AccessArticle

Improvement of the Performance of Scattering Suppression and Absorbing Structure Depth Estimation on Transillumination Image by Deep Learning

by Ngoc An Dang Nguyen, Hoang Nhut Huynh and Trung Nghia Tran

Appl. Sci. 2023, 13(18), 10047; https://doi.org/10.3390/app131810047 - 6 Sep 2023

Cited by 5 | Viewed by 1254

Abstract

The development of optical sensors, especially with regard to the improved resolution of cameras, has made optical techniques more applicable in medicine and live animal research. Research efforts focus on image signal acquisition, scattering de-blur for acquired images, and the development of image [...] Read more.

The development of optical sensors, especially with regard to the improved resolution of cameras, has made optical techniques more applicable in medicine and live animal research. Research efforts focus on image signal acquisition, scattering de-blur for acquired images, and the development of image reconstruction algorithms. Rapidly evolving artificial intelligence has enabled the development of techniques for de-blurring and estimating the depth of light-absorbing structures in biological tissues. Although the feasibility of applying deep learning to overcome these problems has been demonstrated in previous studies, limitations still exist in terms of de-blurring capabilities on complex structures and the heterogeneity of turbid medium, as well as the limit of accurate estimation of the depth of absorptive structures in biological tissues (shallower than 15.0 mm). These problems are related to the absorption structure’s complexity, the biological tissue’s heterogeneity, the training data, and the neural network model itself. This study thoroughly explores how to generate training and testing datasets on different deep learning models to find the model with the best performance. The results of the de-blurred image show that the Attention Res-UNet model has the best de-blurring ability, with a correlation of more than 89% between the de-blurred image and the original structure image. This result comes from adding the Attention gate and the Residual block to the common U-net model structure. The results of the depth estimation show that the DenseNet169 model shows the ability to estimate depth with high accuracy beyond the limit of 20.0 mm. The results of this study once again confirm the feasibility of applying deep learning in transmission image processing to reconstruct clear images and obtain information on the absorbing structure inside biological tissue. This allows the development of subsequent transillumination imaging studies in biological tissues with greater heterogeneity and structural complexity. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

11 pages, 1257 KiB

Open AccessArticle

Deep Learning Enhances Radiologists’ Detection of Potential Spinal Malignancies in CT Scans

by Leonard Gilberg, Bianca Teodorescu, Leander Maerkisch, Andre Baumgart, Rishi Ramaesh, Elmer Jeto Gomes Ataide and Ali Murat Koç

Appl. Sci. 2023, 13(14), 8140; https://doi.org/10.3390/app13148140 - 13 Jul 2023

Cited by 5 | Viewed by 3130

Abstract

Incidental spinal bone lesions, potential indicators of malignancies, are frequently underreported in abdominal and thoracic CT imaging due to scan focus and diagnostic bias towards patient complaints. Here, we evaluate a deep-learning algorithm (DLA) designed to support radiologists’ reporting of incidental lesions during [...] Read more.

Incidental spinal bone lesions, potential indicators of malignancies, are frequently underreported in abdominal and thoracic CT imaging due to scan focus and diagnostic bias towards patient complaints. Here, we evaluate a deep-learning algorithm (DLA) designed to support radiologists’ reporting of incidental lesions during routine clinical practice. The present study is structured into two phases: unaided and AI-assisted. A total of 32 scans from multiple radiology centers were selected randomly and independently annotated by two experts. The U-Net-like architecture-based DLA used for the AI-assisted phase showed a sensitivity of 75.0% in identifying potentially malignant spinal bone lesions. Six radiologists of varying experience levels participated in this observational study. During routine reporting, the DLA helped improve the radiologists’ sensitivity by 20.8 percentage points. Notably, DLA-generated false-positive predictions did not significantly bias radiologists in their final diagnosis. These observations clearly indicate that using a suitable DLA improves the detection of otherwise missed potentially malignant spinal cases. Our results further emphasize the potential of artificial intelligence as a second reader in the clinical setting. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

19 pages, 4097 KiB

Open AccessArticle

Region-of-Interest Optimization for Deep-Learning-Based Breast Cancer Detection in Mammograms

by Hoang Nhut Huynh, Anh Tu Tran and Trung Nghia Tran

Appl. Sci. 2023, 13(12), 6894; https://doi.org/10.3390/app13126894 - 7 Jun 2023

Cited by 7 | Viewed by 3217

Abstract

The early detection and diagnosis of breast cancer may increase survival rates and reduce overall treatment costs. The cancer of the breast is a severe and potentially fatal disease that impacts individuals worldwide. Mammography is a widely utilized imaging technique for breast cancer [...] Read more.

The early detection and diagnosis of breast cancer may increase survival rates and reduce overall treatment costs. The cancer of the breast is a severe and potentially fatal disease that impacts individuals worldwide. Mammography is a widely utilized imaging technique for breast cancer surveillance and diagnosis. However, images produced with mammography frequently contain noise, poor contrast, and other anomalies that hinder radiologists from interpreting the images. This study develops a novel deep-learning technique for breast cancer detection using mammography images. The proposed procedure consists of two primary steps: region-of-interest (ROI) (1) extraction and (2) classification. At the beginning of the procedure, a YOLOX model is utilized to distinguish breast tissue from the background and to identify ROIs that may contain lesions. In the second phase, the EfficientNet or ConvNeXt model is applied to the data to identify benign or malignant ROIs. The proposed technique is validated using a large dataset of mammography images from various institutions and compared to several baseline methods. The pF1 index is used to measure the effectiveness of the technique, which aims to establish a balance between the number of false positives and false negatives, and is a harmonic mean of accuracy and recall. The proposed method outperformed existing methods by an average of 8.0%, obtaining superior levels of precision and sensitivity, and area under the receiver operating characteristics curve (ROC AUC) and the precision–recall curve (PR AUC). In addition, ablation research was conducted to investigate the effects of the procedure’s numerous components. According to the findings, the proposed technique is another choice that could enhance the detection and diagnosis of breast cancer using mammography images. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

12 pages, 1586 KiB

Open AccessArticle

Ultrasound Intima-Media Complex (IMC) Segmentation Using Deep Learning Models

by Hanadi Hassen Mohammed, Omar Elharrouss, Najmath Ottakath, Somaya Al-Maadeed, Muhammad E. H. Chowdhury, Ahmed Bouridane and Susu M. Zughaier

Appl. Sci. 2023, 13(8), 4821; https://doi.org/10.3390/app13084821 - 12 Apr 2023

Cited by 7 | Viewed by 2583

Abstract

Common carotid intima-media thickness (CIMT) is a common measure of atherosclerosis, often assessed through carotid ultrasound images. However, the use of deep learning methods for medical image analysis, segmentation and CIMT measurement in these images has not been extensively explored. This study aims [...] Read more.

Common carotid intima-media thickness (CIMT) is a common measure of atherosclerosis, often assessed through carotid ultrasound images. However, the use of deep learning methods for medical image analysis, segmentation and CIMT measurement in these images has not been extensively explored. This study aims to evaluate the performance of four recent deep learning models, including a convolutional neural network (CNN), a self-organizing operational neural network (self-ONN), a transformer-based network and a pixel difference convolution-based network, in segmenting the intima-media complex (IMC) using the CUBS dataset, which includes ultrasound images acquired from both sides of the neck of 1088 participants. The results show that the self-ONN model outperforms the conventional CNN-based model, while the pixel difference- and transformer-based models achieve the best segmentation performance. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

14 pages, 7988 KiB

Open AccessArticle

Magnetic Resonance with Diffusion and Dynamic Perfusion-Weighted Imaging in the Assessment of Early Chemoradiotherapy Response of Naso-Oropharyngeal Carcinoma

by Michele Pietragalla, Eleonora Bicci, Linda Calistri, Chiara Lorini, Pierluigi Bonomo, Andrea Borghesi, Antonio Lo Casto, Francesco Mungai, Luigi Bonasera, Giandomenico Maggiore and Cosimo Nardi

Appl. Sci. 2023, 13(5), 2799; https://doi.org/10.3390/app13052799 - 22 Feb 2023

Cited by 1 | Viewed by 1686

Abstract

The purpose of this study was to differentiate post-chemoradiotherapy (CRT) changes from tumor persistence/recurrence in early follow-up of naso-oropharyngeal carcinoma on magnetic resonance (MRI) with diffusion (DWI) and dynamic contrast-enhanced perfusion-weighted imaging (DCE-PWI). A total of 37 patients were assessed with MRI both [...] Read more.

The purpose of this study was to differentiate post-chemoradiotherapy (CRT) changes from tumor persistence/recurrence in early follow-up of naso-oropharyngeal carcinoma on magnetic resonance (MRI) with diffusion (DWI) and dynamic contrast-enhanced perfusion-weighted imaging (DCE-PWI). A total of 37 patients were assessed with MRI both for tumor staging and 4-month follow-up from ending CRT. Mean apparent diffusion coefficient (ADC) values, area under the curve (AUC), and K(trans) values were calculated from DWI and DCE-PWI images, respectively. DWI and DCE-PWI values of primary tumor (ADC, AUC, K(trans)_pre), post-CRT changes (ADC, AUC, K(trans)_post), and trapezius muscle as a normative reference before and after CRT (ADC, AUC, K(trans)_muscle pre and _muscle post; AUC_{post/muscle post}:AUC_{pre/muscle pre} (AUC_{post/pre/muscle}); K(trans)_{post/muscle post}:K(trans)_{pre/muscle pre} (K(trans)_{post/pre/muscle}) were assessed. In detecting post-CRT changes, ADC_post > 1.33 × 10⁻³ mm²/s and an increase >0.72 × 10⁻³ mm²/s and/or >65.5% between ADC_post and ADC_pre values (ADC_post-pre; ADC_post-pre%) had 100% specificity, whereas hypointense signal intensity on DWIb800 images showed specificity 80%. Although mean AUC_{post/pre/muscle} and K(trans)_{post/pre/muscle} were similar both in post-CRT changes (1.10 ± 0.58; 1.08 ± 0.91) and tumor persistence/recurrence (1.09 ± 0.11; 1.03 ± 0.12), K(trans)_{post/pre/muscle} values < 0.85 and >1.20 suggested post-CRT fibrosis and inflammatory edema, respectively. In early follow-up of naso-oropharyngeal carcinoma, our sample showed that ADC_post > 1.33 × 10⁻³ mm²/s, ADC_post-pre% > 65.5%, and ADC_post-pre > 0.72 × 10⁻³ mm²/s identified post-CRT changes with 100% specificity. K(trans)_{post/pre/muscle} values less than 0.85 suggested post-CRT fibrosis, whereas K(trans)_{post/pre/muscle} values more than 1.20 indicated inflammatory edema. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

14 pages, 3446 KiB

Open AccessArticle

Hybrid System Mixed Reality and Marker-Less Motion Tracking for Sports Rehabilitation of Martial Arts Athletes

by Michela Franzò, Andrada Pica, Simona Pascucci, Franco Marinozzi and Fabiano Bini

Appl. Sci. 2023, 13(4), 2587; https://doi.org/10.3390/app13042587 - 17 Feb 2023

Cited by 11 | Viewed by 3547

Abstract

Rehabilitation is a vast field of research. Virtual and Augmented Reality represent rapidly emerging technologies that have the potential to support physicians in several medical activities, e.g., diagnosis, surgical training, and rehabilitation, and can also help sports experts analyze athlete movements and performance. [...] Read more.

Rehabilitation is a vast field of research. Virtual and Augmented Reality represent rapidly emerging technologies that have the potential to support physicians in several medical activities, e.g., diagnosis, surgical training, and rehabilitation, and can also help sports experts analyze athlete movements and performance. In this study, we present the implementation of a hybrid system for the real-time visualization of 3D virtual models of bone segments and other anatomical components on a subject performing critical karate shots and stances. The project is composed of an economic markerless motion tracking device, Microsoft Kinect Azure, that recognizes the subject movements and the position of anatomical joints; an augmented reality headset, Microsoft HoloLens 2, on which the user can visualize the 3D reconstruction of bones and anatomical information; and a terminal computer with a code implemented in Unity Platform. The 3D reconstructed bones are overlapped with the athlete, tracked by the Kinect in real-time, and correctly displayed on the headset. The findings suggest that this system could be a promising technology to monitor martial arts athletes after injuries to support the restoration of their movements and position to rejoin official competitions. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

20 pages, 4772 KiB

Open AccessArticle

Skin Cancer Classification Framework Using Enhanced Super Resolution Generative Adversarial Network and Custom Convolutional Neural Network

by Sufiyan Bashir Mukadam and Hemprasad Yashwant Patil

Appl. Sci. 2023, 13(2), 1210; https://doi.org/10.3390/app13021210 - 16 Jan 2023

Cited by 21 | Viewed by 9555

Abstract

Melanin skin lesions are most commonly spotted as small patches on the skin. It is nothing but overgrowth caused by melanocyte cells. Skin melanoma is caused due to the abnormal surge of melanocytes. The number of patients suffering from skin cancer is observably [...] Read more.

Melanin skin lesions are most commonly spotted as small patches on the skin. It is nothing but overgrowth caused by melanocyte cells. Skin melanoma is caused due to the abnormal surge of melanocytes. The number of patients suffering from skin cancer is observably rising globally. Timely and precise identification of skin cancer is crucial for lowering mortality rates. An expert dermatologist is required to handle the cases of skin cancer using dermoscopy images. Improper diagnosis can cause fatality to the patient if it is not detected accurately. Some of the classes come under the category of benign while the rest are malignant, causing severe issues if not diagnosed at an early stage. To overcome these issues, Computer-Aided Design (CAD) systems are proposed which help to reduce the burden on the dermatologist by giving them accurate and precise diagnosis of skin images. There are several deep learning techniques that are implemented for cancer classification. In this experimental study, we have implemented a custom Convolution Neural Network (CNN) on a Human-against-Machine (HAM10000) database which is publicly accessible through the Kaggle website. The designed CNN model classifies the seven different classes present in HAM10000 database. The proposed experimental model achieves an accuracy metric of 98.77%, 98.36%, and 98.89% for protocol-I, protocol-II, and protocol-III, respectively, for skin cancer classification. Results of our proposed models are also assimilated with several different models in the literature and were found to be superior than most of them. To enhance the performance metrics, the database is initially pre-processed using an Enhanced Super Resolution Generative Adversarial Network (ESRGAN) which gives a better image resolution for images of smaller size. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

15 pages, 2938 KiB

Open AccessArticle

Morphological, Functional and Texture Analysis Magnetic Resonance Imaging Features in the Assessment of Radiotherapy-Induced Xerostomia in Oropharyngeal Cancer

by Leonardo Calamandrei, Luca Mariotti, Eleonora Bicci, Linda Calistri, Eleonora Barcali, Martina Orlandi, Nicholas Landini, Francesco Mungai, Luigi Bonasera, Pierluigi Bonomo, Isacco Desideri, Leonardo Bocchi and Cosimo Nardi

Appl. Sci. 2023, 13(2), 810; https://doi.org/10.3390/app13020810 - 6 Jan 2023

Cited by 5 | Viewed by 1816

Abstract

The aim of this single-center, observational, retrospective study was to investigate magnetic resonance imaging (MRI) biomarkers for the assessment of radiotherapy (RT)-induced xerostomia. Twenty-seven patients who underwent radiation therapy for oropharyngeal cancer were divided into three groups according to the severity of their [...] Read more.

The aim of this single-center, observational, retrospective study was to investigate magnetic resonance imaging (MRI) biomarkers for the assessment of radiotherapy (RT)-induced xerostomia. Twenty-seven patients who underwent radiation therapy for oropharyngeal cancer were divided into three groups according to the severity of their xerostomia—mild, moderate, and severe—clinically confirmed with the Common Terminology Criteria for Adverse Events (CTCAE). No severe xerostomia was found. Conventional and functional MRI (perfusion- and diffusion- weighted imaging) performed both pre- and post-RT were studied for signal intensity, mean apparent diffusion coefficient (ADC) values, k-trans, and area under the perfusion curves. Contrast-enhanced T1 images and ADC maps were imported into 3D slicer software, and salivary gland volumes were segmented. A total of 107 texture features were derived. T-Student and Wilcoxon signed-rank tests were performed on functional MRI parameters and texture analysis features to identify the differences between pre- and post-RT populations. A p-value < 0.01 was defined as acceptable. Receiver operating characteristic (ROC) curves were plotted for significant parameters to discriminate the severity of xerostomia in the pre-RT population. Conventional and functional MRI did not yield statistically significant results; on the contrary, five texture features showed significant variation between pre- and post-RT on the ADC maps, of which only informational measure of correlation 1 (IMC 1) was able to discriminate the severity of RT-induced xerostomia in the pre-RT population (area under the curve (AUC) > 0.7). Values lower than the cut-off of −1.473 × 10⁻¹¹ were associated with moderate xerostomia, enabling the differentiation of mild xerostomia from moderate xerostomia with a 73% sensitivity, 75% specificity, and 75% diagnostic accuracy. Therefore, the texture feature IMC 1 on the ADC maps allowed the distinction between different degrees of severity of RT-induced xerostomia in the pre-RT population. Accordingly, texture analysis on ADC maps should be considered a useful tool to evaluate salivary gland radiosensitivity and help identify patients at risk of developing more serious xerostomia before radiation therapy is administered. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

13 pages, 1528 KiB

Open AccessArticle

Innovative Tool for Automatic Detection of Arterial Stenosis on Cone Beam Computed Tomography

by Agnese Simoni, Eleonora Barcali, Cosimo Lorenzetto, Eleonora Tiribilli, Vieri Rastrelli, Leonardo Manetti, Cosimo Nardi, Ernesto Iadanza and Leonardo Bocchi

Appl. Sci. 2023, 13(2), 805; https://doi.org/10.3390/app13020805 - 6 Jan 2023

Cited by 2 | Viewed by 1314

Abstract

Arterial stenosis is one of the main vascular diseases that are treated with minimally invasive surgery approaches. The aim of this study was to provide a tool to support the medical doctor in planning endovascular surgery, allowing the rapid detection of stenotic vessels [...] Read more.

Arterial stenosis is one of the main vascular diseases that are treated with minimally invasive surgery approaches. The aim of this study was to provide a tool to support the medical doctor in planning endovascular surgery, allowing the rapid detection of stenotic vessels and the quantification of the stenosis. Skeletonization was used to improve vessels’ visualization. The distance transform was used to obtain a linear representation of the diameter of critical vessels selected by the user. The system also provides an estimate of the exact distance between landmarks on the vascular tree and the occlusion, important information that can be used in the planning of the surgery. The advantage of the proposed tool is to lead the examination on the linear representation of the chosen vessels that are free from tortuous vascular courses and from vessel crossings. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

26 pages, 2079 KiB

Open AccessArticle

Binary Starling Murmuration Optimizer Algorithm to Select Effective Features from Medical Data

by Mohammad H. Nadimi-Shahraki, Zahra Asghari Varzaneh, Hoda Zamani and Seyedali Mirjalili

Appl. Sci. 2023, 13(1), 564; https://doi.org/10.3390/app13010564 - 31 Dec 2022

Cited by 36 | Viewed by 3223

Abstract

Feature selection is an NP-hard problem to remove irrelevant and redundant features with no predictive information to increase the performance of machine learning algorithms. Many wrapper-based methods using metaheuristic algorithms have been proposed to select effective features. However, they achieve differently on medical [...] Read more.

Feature selection is an NP-hard problem to remove irrelevant and redundant features with no predictive information to increase the performance of machine learning algorithms. Many wrapper-based methods using metaheuristic algorithms have been proposed to select effective features. However, they achieve differently on medical data, and most of them cannot find those effective features that may fulfill the required accuracy in diagnosing important diseases such as Diabetes, Heart problems, Hepatitis, and Coronavirus, which are targeted datasets in this study. To tackle this drawback, an algorithm is needed that can strike a balance between local and global search strategies in selecting effective features from medical datasets. In this paper, a new binary optimizer algorithm named BSMO is proposed. It is based on the newly proposed starling murmuration optimizer (SMO) that has a high ability to solve different complex and engineering problems, and it is expected that BSMO can also effectively find an optimal subset of features. Two distinct approaches are utilized by the BSMO algorithm when searching medical datasets to find effective features. Each dimension in a continuous solution generated by SMO is simply mapped to 0 or 1 using a variable threshold in the second approach, whereas in the first, binary versions of BSMO are developed using several S-shaped and V-shaped transfer functions. The performance of the proposed BSMO was evaluated using four targeted medical datasets, and results were compared with well-known binary metaheuristic algorithms in terms of different metrics, including fitness, accuracy, sensitivity, specificity, precision, and error. Finally, the superiority of the proposed BSMO algorithm was statistically analyzed using Friedman non-parametric test. The statistical and experimental tests proved that the proposed BSMO attains better performance in comparison to the competitive algorithms such as ACO, BBA, bGWO, and BWOA for selecting effective features from the medical datasets targeted in this study. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

23 pages, 1153 KiB

Open AccessArticle

Mixed-Sized Biomedical Image Segmentation Based on U-Net Architectures

by Priscilla Benedetti, Mauro Femminella and Gianluca Reali

Appl. Sci. 2023, 13(1), 329; https://doi.org/10.3390/app13010329 - 27 Dec 2022

Cited by 8 | Viewed by 4349

Abstract

Convolutional neural networks (CNNs) are becoming increasingly popular in medical Image Segmentation. Among them, U-Net is a widely used model that can lead to cutting-edge results for 2D biomedical Image Segmentation. However, U-Net performance can be influenced by many factors, such as the [...] Read more.

Convolutional neural networks (CNNs) are becoming increasingly popular in medical Image Segmentation. Among them, U-Net is a widely used model that can lead to cutting-edge results for 2D biomedical Image Segmentation. However, U-Net performance can be influenced by many factors, such as the size of the training dataset, the performance metrics used, the quality of the images and, in particular, the shape and size of the organ to be segmented. This could entail a loss of robustness of the U-Net-based models. In this paper, the performance of the considered networks is determined by using the publicly available images from the 3D-IRCADb-01 dataset. Different organs with different features are considered. Experimental results show that the U-Net-based segmentation performance decreases when organs with sparse binary masks are considered. The solution proposed in this paper, based on automated zooming of the parts of interest, allows improving the performance of the segmentation model by up to 20% in terms of Dice coefficient metric, when very sparse segmentation images are used, without affecting the cost of the learning process. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

16 pages, 1653 KiB

Open AccessArticle

Lightweight Dual Mutual-Feedback Network for Artificial Intelligence in Medical Image Super-Resolution

by Beibei Wang, Binyu Yan, Gwanggil Jeon, Xiaomin Yang, Changjun Liu and Zhuoyue Zhang

Appl. Sci. 2022, 12(24), 12794; https://doi.org/10.3390/app122412794 - 13 Dec 2022

Cited by 4 | Viewed by 1533

Abstract

As a result of hardware resource constraints, it is difficult to obtain medical images with a sufficient resolution to diagnose small lesions. Recently, super-resolution (SR) was introduced into the field of medicine to enhance and restore medical image details so as to help [...] Read more.

As a result of hardware resource constraints, it is difficult to obtain medical images with a sufficient resolution to diagnose small lesions. Recently, super-resolution (SR) was introduced into the field of medicine to enhance and restore medical image details so as to help doctors make more accurate diagnoses of lesions. High-frequency information enhances the accuracy of the image reconstruction, which is demonstrated by deep SR networks. However, deep networks are not applicable to resource-constrained medical devices because they have too many parameters, which requires a lot of memory and higher processor computing power. For this reason, a lightweight SR network that demonstrates good performance is needed to improve the resolution of medical images. A feedback mechanism enables the previous layers to perceive high-frequency information of the latter layers, but no new parameters are introduced, which is rarely used in lightweight networks. Therefore, in this work, a lightweight dual mutual-feedback network (DMFN) is proposed for medical image super-resolution, which contains two back-projection units that operate in a dual mutual-feedback manner. The features generated by the up-projection unit are fed back into the down-projection unit and, simultaneously, the features generated by the down-projection unit are fed back into the up-projection unit. Moreover, a contrast-enhanced residual block (CRB) is proposed as each cell block used in projection units, which enhances the pixel contrast in the channel and spatial dimensions. Finally, we designed a unity feedback to down-sample the SR result as the inverse process of SR. Furthermore, we compared it with the input LR to narrow the solution space of the SR function. The final ablation studies and comparison results show that our DMFN performs well without utilizing a large amount of computing resources. Thus, it can be used in resource-constrained medical devices to obtain medical images with better resolutions. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

16 pages, 477 KiB

Open AccessArticle

Deep 3D Convolutional Neural Network for Facial Micro-Expression Analysis from Video Images

by Kranthi Kumar Talluri, Marc-André Fiedler and Ayoub Al-Hamadi

Appl. Sci. 2022, 12(21), 11078; https://doi.org/10.3390/app122111078 - 1 Nov 2022

Cited by 7 | Viewed by 2414

Abstract

Micro-expression is the involuntary emotion of the human that reflects the genuine feelings that cannot be hidden. Micro-expression is exhibited by facial expressions that last for a short duration and have very low intensity. Because of these reasons, micro-expression recognition is a challenging [...] Read more.

Micro-expression is the involuntary emotion of the human that reflects the genuine feelings that cannot be hidden. Micro-expression is exhibited by facial expressions that last for a short duration and have very low intensity. Because of these reasons, micro-expression recognition is a challenging task. Recent research on the application of 3D convolutional neural networks (CNNs) has gained much popularity for video-based micro-expression analysis. For this purpose, both spatial as well as temporal features are of great importance to achieve high accuracies. The real possibly suppressed emotions of a person are valuable information for a variety of applications, such as in security, psychology, neuroscience, medicine and many other disciplines. This paper proposes a 3D CNN model architecture which is able to extract spatial and temporal features simultaneously. Thereby, the selection of the frame sequence plays a crucial role, since the emotions are only distinctive in a subset of the frames. Thus, we employ a novel pre-processing technique to select the Apex frame sequence from the entire video, where the timestamp of the most pronounced emotion is centered within this sequence. After an extensive evaluation including many experiments, the results show that the train–test split evaluation is biased toward a particular split and cannot be recommended in case of small and imbalanced datasets. Instead, a stratified K-fold evaluation technique is utilized to evaluate the model, which proves to be much more appropriate when using the three benchmark datasets CASME II, SMIC, and SAMM. Moreover, intra-dataset as well as cross-dataset evaluations were conducted in a total of eight different scenarios. For comparison purposes, two networks from the state of the art were reimplemented and compared with the presented architecture. In stratified K-fold evaluation, our proposed model outperforms both reimplemented state-of-the-art methods in seven out of eight evaluation scenarios. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

14 pages, 7297 KiB

Open AccessArticle

Towards Explainable Deep Neural Networks for the Automatic Detection of Diabetic Retinopathy

by Hanan Saleh Alghamdi

Appl. Sci. 2022, 12(19), 9435; https://doi.org/10.3390/app12199435 - 21 Sep 2022

Cited by 15 | Viewed by 2786

Abstract

Diabetic Retinopathy (DR) is a common complication associated with diabetes, causing irreversible vision loss. Early detection of DR can be very helpful for clinical treatment. Ophthalmologists’ manual approach to DR diagnoses is expensive and time-consuming; thus, automatic detection of DR is becoming vital, [...] Read more.

Diabetic Retinopathy (DR) is a common complication associated with diabetes, causing irreversible vision loss. Early detection of DR can be very helpful for clinical treatment. Ophthalmologists’ manual approach to DR diagnoses is expensive and time-consuming; thus, automatic detection of DR is becoming vital, especially with the increasing number of diabetes patients worldwide. Deep learning methods for analyzing medical images have recently become prevalent, achieving state-of-the-art results. Consequently, the need for interpretable deep learning has increased. Although it was demonstrated that the representation depth is beneficial for classification accuracy for DR diagnoses, model explainability is rarely analyzed. In this paper, we evaluated three state-of-the-art deep learning models to accelerate DR detection using the fundus images dataset. We have also proposed a novel explainability metric to leverage domain-based knowledge and validate the reasoning of a deep learning model’s decisions. We conducted two experiments to classify fundus images into normal and abnormal cases and to categorize the images according to the DR severity. The results show the superiority of the VGG-16 model in terms of accuracy, precision, and recall for both binary and DR five-stage classification. Although the achieved accuracy of all evaluated models demonstrates their capability to capture some lesion patterns in the relevant DR cases, the evaluation of the models in terms of their explainability using the Grad-CAM-based color visualization approach shows that the models are not necessarily able to detect DR related lesions to make the classification decision. Thus, more investigations are needed to improve the deep learning model’s explainability for medical diagnosis. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

15 pages, 3005 KiB

Open AccessArticle

Guided Random Mask: Adaptively Regularizing Deep Neural Networks for Medical Image Analysis by Potential Lesions

by Xiaorui Yu, Shuqi Wang and Junjie Hu

Appl. Sci. 2022, 12(18), 9099; https://doi.org/10.3390/app12189099 - 9 Sep 2022

Cited by 2 | Viewed by 2058

Abstract

Data augmentation is a critical regularization method that contributes to numerous state-of-the-art results achieved by deep neural networks (DNNs). The visual interpretation method demonstrates that the DNNs behave like object detectors, focusing on the discriminative regions in the input image. Many studies have [...] Read more.

Data augmentation is a critical regularization method that contributes to numerous state-of-the-art results achieved by deep neural networks (DNNs). The visual interpretation method demonstrates that the DNNs behave like object detectors, focusing on the discriminative regions in the input image. Many studies have also discovered that the DNNs correctly identify the lesions in the input, which has been confirmed in the current work. However, for medical images containing complicated lesions, we observe the DNNs focus on the most prominent abnormalities, neglecting sub-clinical characteristics that may also help diagnosis. We speculate this bias may hamper the generalization ability of DNNs, potentially causing false predicted results. Based on this consideration, a simple yet effective data augmentation method called guided random mask (GRM) is proposed to discover the lesions with different characteristics. Visual interpretation of the inference result is used as guidance to generate random-sized masks, forcing the DNNs to learn both the prominent and subtle lesions. One notable difference between GRM and conventional data augmentation methods is the association with the training phase of DNNs. The parameters in vanilla augmentation methods are independent of the training phase, which may limit their effectiveness when the scale and appearance of region-of-interests vary. Nevertheless, the effectiveness of the proposed GRM method evolves with the training of DNNs, adaptively regularizing the DNNs to alleviate the over-fitting problem. Moreover, the GRM is a parameter-free augmentation method that can be incorporated into DNNs without modifying the architecture. The GRM is empirically verified on multiple datasets with different modalities, including optical coherence tomography, x-ray, and color fundus images. Quantitative experimental results show that the proposed GRM method achieves higher classification accuracy than the commonly used augmentation methods in multiple networks. Visualization analysis also demonstrates that the GRM can better localize lesions than the vanilla network. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

16 pages, 1860 KiB

Open AccessArticle

Augmented Reality in Surgery: A Scoping Review

by Eleonora Barcali, Ernesto Iadanza, Leonardo Manetti, Piergiorgio Francia, Cosimo Nardi and Leonardo Bocchi

Appl. Sci. 2022, 12(14), 6890; https://doi.org/10.3390/app12146890 - 7 Jul 2022

Cited by 39 | Viewed by 5798

Abstract

Augmented reality (AR) is an innovative system that enhances the real world by superimposing virtual objects on reality. The aim of this study was to analyze the application of AR in medicine and which of its technical solutions are the most used. We [...] Read more.

Augmented reality (AR) is an innovative system that enhances the real world by superimposing virtual objects on reality. The aim of this study was to analyze the application of AR in medicine and which of its technical solutions are the most used. We carried out a scoping review of the articles published between 2019 and February 2022. The initial search yielded a total of 2649 articles. After applying filters, removing duplicates and screening, we included 34 articles in our analysis. The analysis of the articles highlighted that AR has been traditionally and mainly used in orthopedics in addition to maxillofacial surgery and oncology. Regarding the display application in AR, the Microsoft HoloLens Optical Viewer is the most used method. Moreover, for the tracking and registration phases, the marker-based method with a rigid registration remains the most used system. Overall, the results of this study suggested that AR is an innovative technology with numerous advantages, finding applications in several new surgery domains. Considering the available data, it is not possible to clearly identify all the fields of application and the best technologies regarding AR. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

24 pages, 5911 KiB

Open AccessArticle

IMNets: Deep Learning Using an Incremental Modular Network Synthesis Approach for Medical Imaging Applications

by Redha Ali, Russell C. Hardie, Barath Narayanan Narayanan and Temesguen M. Kebede

Appl. Sci. 2022, 12(11), 5500; https://doi.org/10.3390/app12115500 - 29 May 2022

Cited by 45 | Viewed by 4539

Abstract

Deep learning approaches play a crucial role in computer-aided diagnosis systems to support clinical decision-making. However, developing such automated solutions is challenging due to the limited availability of annotated medical data. In this study, we proposed a novel and computationally efficient deep learning [...] Read more.

Deep learning approaches play a crucial role in computer-aided diagnosis systems to support clinical decision-making. However, developing such automated solutions is challenging due to the limited availability of annotated medical data. In this study, we proposed a novel and computationally efficient deep learning approach to leverage small data for learning generalizable and domain invariant representations in different medical imaging applications such as malaria, diabetic retinopathy, and tuberculosis. We refer to our approach as Incremental Modular Network Synthesis (IMNS), and the resulting CNNs as Incremental Modular Networks (IMNets). Our IMNS approach is to use small network modules that we call SubNets which are capable of generating salient features for a particular problem. Then, we build up ever larger and more powerful networks by combining these SubNets in different configurations. At each stage, only one new SubNet module undergoes learning updates. This reduces the computational resource requirements for training and aids in network optimization. We compare IMNets against classic and state-of-the-art deep learning architectures such as AlexNet, ResNet-50, Inception v3, DenseNet-201, and NasNet for the various experiments conducted in this study. Our proposed IMNS design leads to high average classification accuracies of 97.0%, 97.9%, and 88.6% for malaria, diabetic retinopathy, and tuberculosis, respectively. Our modular design for deep learning achieves the state-of-the-art performance in the scenarios tested. The IMNets produced here have a relatively low computational complexity compared to traditional deep learning architectures. The largest IMNet tested here has 0.95 M of the learnable parameters and 0.08 G of the floating-point multiply–add (MAdd) operations. The simpler IMNets train faster, have lower memory requirements, and process images faster than the benchmark methods tested. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

Review

Jump to: Editorial, Research, Other

28 pages, 585 KiB

Open AccessReview

Deep Active Learning for Computer Vision Tasks: Methodologies, Applications, and Challenges

by Mingfei Wu, Chen Li and Zehuan Yao

Appl. Sci. 2022, 12(16), 8103; https://doi.org/10.3390/app12168103 - 12 Aug 2022

Cited by 26 | Viewed by 5617

Abstract

Active learning is a label-efficient machine learning method that actively selects the most valuable unlabeled samples to annotate. Active learning focuses on achieving the best possible performance while using as few, high-quality sample annotations as possible. Recently, active learning achieved promotion combined with [...] Read more.

Active learning is a label-efficient machine learning method that actively selects the most valuable unlabeled samples to annotate. Active learning focuses on achieving the best possible performance while using as few, high-quality sample annotations as possible. Recently, active learning achieved promotion combined with deep learning-based methods, which are named deep active learning methods in this paper. Deep active learning plays a crucial role in computer vision tasks, especially in label-insensitive scenarios, such as hard-to-label tasks (medical images analysis) and time-consuming tasks (autonomous driving). However, deep active learning still has some challenges, such as unstable performance and dirty data, which are future research trends. Compared with other reviews on deep active learning, our work introduced the deep active learning from computer vision-related methodologies and corresponding applications. The expected audience of this vision-friendly survey are researchers who are working in computer vision but willing to utilize deep active learning methods to solve vision problems. Specifically, this review systematically focuses on the details of methods, applications, and challenges in vision tasks, and we also introduce the classic theories, strategies, and scenarios of active learning in brief. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

Other

Jump to: Editorial, Research, Review

32 pages, 979 KiB

Open AccessSystematic Review

Applications of Deep Learning to Neurodevelopment in Pediatric Imaging: Achievements and Challenges

by Mengjiao Hu, Cosimo Nardi, Haihong Zhang and Kai-Keng Ang

Appl. Sci. 2023, 13(4), 2302; https://doi.org/10.3390/app13042302 - 10 Feb 2023

Cited by 3 | Viewed by 2377

Abstract

Deep learning has achieved remarkable progress, particularly in neuroimaging analysis. Deep learning applications have also been extended from adult to pediatric medical images, and thus, this paper aims to present a systematic review of this recent research. We first introduce the commonly used [...] Read more.

Deep learning has achieved remarkable progress, particularly in neuroimaging analysis. Deep learning applications have also been extended from adult to pediatric medical images, and thus, this paper aims to present a systematic review of this recent research. We first introduce the commonly used deep learning methods and architectures in neuroimaging, such as convolutional neural networks, auto-encoders, and generative adversarial networks. A non-exhaustive list of commonly used publicly available pediatric neuroimaging datasets and repositories are included, followed by a categorical review of recent works in pediatric MRI-based deep learning studies in the past five years. These works are categorized into recognizing neurodevelopmental disorders, identifying brain and tissue structures, estimating brain age/maturity, predicting neurodevelopment outcomes, and optimizing MRI brain imaging and analysis. Finally, we also discuss the recent achievements and challenges on these applications of deep learning to pediatric neuroimaging. Full article

(This article belongs to the Special Issue Artificial Intelligence in Medical Imaging: The Beginning of a New Era)

► Show Figures

Figure 1

Journal Menu

Journal Browser

Artificial Intelligence in Medical Imaging: The Beginning of a New Era

Share This Special Issue

Special Issue Editor

Special Issue Information

Keywords

Benefits of Publishing in a Special Issue

Published Papers (20 papers)

Editorial

Research

Review

Other

Further Information

Guidelines

MDPI Initiatives

Follow MDPI