PLoS One. 2024 Jun 24;19(6):e0299623. doi: 10.1371/journal.pone.0299623. eCollection 2024.

ABSTRACT

BACKGROUND: In medical imaging, the integration of deep-learning-based semantic segmentation algorithms with preprocessing techniques can reduce the need for human annotation and advance disease classification. Among established preprocessing techniques, Contrast Limited Adaptive Histogram Equalization (CLAHE) has demonstrated efficacy in improving segmentation algorithms across various modalities, such as X-rays and CT. However, there remains a demand for improved contrast enhancement methods considering the heterogeneity of datasets and the various contrasts across different anatomic structures.

METHOD: This study proposes a novel preprocessing technique, ps-KDE, to investigate its impact on deep learning algorithms to segment major organs in posterior-anterior chest X-rays. Ps-KDE augments image contrast by substituting pixel values based on their normalized frequency across all images. We evaluate our approach on a U-Net architecture with ResNet34 backbone pre-trained on ImageNet. Five separate models are trained to segment the heart, left lung, right lung, left clavicle, and right clavicle.

RESULTS: The model trained to segment the left lung using ps-KDE achieved a Dice score of 0.780 (SD = 0.13), while that of trained on CLAHE achieved a Dice score of 0.717 (SD = 0.19), p<0.01. ps-KDE also appears to be more robust as CLAHE-based models misclassified right lungs in select test images for the left lung model. The algorithm for performing ps-KDE is available at https://github.com/wyc79/ps-KDE.

DISCUSSION: Our results suggest that ps-KDE offers advantages over current preprocessing techniques when segmenting certain lung regions. This could be beneficial in subsequent analyses such as disease classification and risk stratification.

PMID:38913621 | DOI:10.1371/journal.pone.0299623

IEEE Trans Ultrason Ferroelectr Freq Control. 2024 Jun 24;PP. doi: 10.1109/TUFFC.2024.3418030. Online ahead of print.

ABSTRACT

Left ventricle (LV) segmentation of 2D echocardiography images is an essential step in the analysis of cardiac morphology and function and - more generally - diagnosis of cardiovascular diseases. Several deep learning (DL) algorithms have recently been proposed for the automatic segmentation of the LV, showing significant performance improvement over the traditional segmentation algorithms. However, unlike the traditional methods, prior information about the segmentation problem, e.g. anatomical shape information, is not usually incorporated for training the DL algorithms. This can degrade the generalization performance of the DL models on unseen images if their characteristics are somewhat different from those of the training images, e.g. low-quality testing images. In this study, a new shape-constrained deep convolutional neural network (CNN) - called BEAS-Net - is introduced for automatic LV segmentation. The BEAS-Net learns how to associate the image features, encoded by its convolutional layers, with anatomical shape-prior information derived by the B-spline explicit active surface (BEAS) algorithm to generate physiologically meaningful segmentation contours when dealing with artifactual or low-quality images. The performance of the proposed network was evaluated using three different in-vivo datasets and was compared a deep segmentation algorithm based on the U-Net model. Both networks yielded comparable results when tested on images of acceptable quality, but the BEAS-Net outperformed the benchmark DL model on artifactual and low-quality images.

PMID:38913532 | DOI:10.1109/TUFFC.2024.3418030

IEEE Trans Neural Netw Learn Syst. 2024 Jun 24;PP. doi: 10.1109/TNNLS.2024.3411020. Online ahead of print.

ABSTRACT

Gradient-descent-based optimizers are prone to slowdowns in training deep learning models, as stationary points are ubiquitous in the loss landscape of most neural networks. We present an intuitive concept of bypassing the stationary points and realize the concept into a novel method designed to actively rescue optimizers from slowdowns encountered in neural network training. The method, bypass pipeline, revitalizes the optimizer by extending the model space and later contracts the model back to its original space with function-preserving algebraic constraints. We implement the method into the bypass algorithm, verify that the algorithm shows theoretically expected behaviors of bypassing, and demonstrate its empirical benefit in regression and classification benchmarks. Bypass algorithm is highly practical, as it is computationally efficient and compatible with other improvements of first-order optimizers. In addition, bypassing for neural networks leads to new theoretical research such as model-specific bypassing and neural architecture search (NAS).

PMID:38913523 | DOI:10.1109/TNNLS.2024.3411020

Ophthalmol Ther. 2024 Jun 24. doi: 10.1007/s40123-024-00981-4. Online ahead of print.

ABSTRACT

We conducted a systematic review of research in artificial intelligence (AI) for retinal fundus photographic images. We highlighted the use of various AI algorithms, including deep learning (DL) models, for application in ophthalmic and non-ophthalmic (i.e., systemic) disorders. We found that the use of AI algorithms for the interpretation of retinal images, compared to clinical data and physician experts, represents an innovative solution with demonstrated superior accuracy in identifying many ophthalmic (e.g., diabetic retinopathy (DR), age-related macular degeneration (AMD), optic nerve disorders), and non-ophthalmic disorders (e.g., dementia, cardiovascular disease). There has been a significant amount of clinical and imaging data for this research, leading to the potential incorporation of AI and DL for automated analysis. AI has the potential to transform healthcare by improving accuracy, speed, and workflow, lowering cost, increasing access, reducing mistakes, and transforming healthcare worker education and training.

PMID:38913289 | DOI:10.1007/s40123-024-00981-4

Eur Radiol. 2024 Jun 24. doi: 10.1007/s00330-024-10851-z. Online ahead of print.

ABSTRACT

OBJECTIVES: To train the machine and deep learning models to automate the justification analysis of radiology referrals in accordance with iGuide categorisation, and to determine if prediction models can generalise across multiple clinical sites and outperform human experts.

METHODS: Adult brain computed tomography (CT) referrals from scans performed in three CT centres in Ireland in 2020 and 2021 were retrospectively collected. Two radiographers analysed the justification of 3000 randomly selected referrals using iGuide, with two consultant radiologists analysing the referrals with disagreement. Insufficient or duplicate referrals were discarded. The inter-rater agreement among radiographers and consultants was computed. A random split (4:1) was performed to apply machine learning (ML) and deep learning (DL) techniques to unstructured clinical indications to automate retrospective justification auditing with multi-class classification. The accuracy and macro-averaged F1 score of the best-performing classifier of each type on the training set were computed on the test set.

RESULTS: 42 referrals were ignored. 1909 (64.5%) referrals were justified, 811 (27.4%) were potentially justified, and 238 (8.1%) were unjustified. The agreement between radiographers (κ = 0.268) was lower than radiologists (κ = 0.460). The best-performing ML model was the bag-of-words-based gradient-boosting classifier achieving a 94.4% accuracy and a macro F1 of 0.94. DL models were inferior, with bi-directional long short-term memory achieving 92.3% accuracy, a macro F1 of 0.92, and outperforming multilayer perceptrons.

CONCLUSION: Interpreting unstructured clinical indications is challenging necessitating clinical decision support. ML and DL can generalise across multiple clinical sites, outperform human experts, and be used as an artificial intelligence-based iGuide interpreter when retrospectively vetting radiology referrals.

CLINICAL RELEVANCE STATEMENT: Healthcare vendors and clinical sites should consider developing and utilising artificial intelligence-enabled systems for justifying medical exposures. This would enable better implementation of imaging referral guidelines in clinical practices and reduce population dose burden, CT waiting lists, and wasteful use of resources.

KEY POINTS: Significant variations exist among human experts in interpreting unstructured clinical indications/patient presentations. Machine and deep learning can automate the justification analysis of radiology referrals according to iGuide categorisation. Machine and deep learning can improve retrospective and prospective justification auditing for better implementation of imaging referral guidelines.

PMID:38913244 | DOI:10.1007/s00330-024-10851-z

Insights Imaging. 2024 Jun 24;15(1):161. doi: 10.1186/s13244-024-01737-1.

ABSTRACT

OBJECTIVES: The clinical decision-making regarding choosing surgery alone (SA) or surgery followed by postoperative adjuvant chemotherapy (SPOCT) in esophageal squamous cell carcinoma (ESCC) remains controversial. We aim to propose a pre-therapy PET/CT image-based deep learning approach to improve the survival benefit and clinical management of ESCC patients.

METHODS: This retrospective multicenter study included 837 ESCC patients from three institutions. Prognostic biomarkers integrating six networks were developed to build an ESCC prognosis (ESCCPro) model and predict the survival probability of ESCC patients treated with SA and SPOCT. Patients who did not undergo surgical resection were in a control group. Overall survival (OS) was the primary end-point event. The expected improvement in survival prognosis with the application of ESCCPro to assign treatment protocols was estimated by comparing the survival of patients in each subgroup. Seven clinicians with varying experience evaluated how ESCCPro performed in assisting clinical decision-making.

RESULTS: In this retrospective multicenter study, patients receiving SA had a median OS 9.2 months longer than controls. No significant differences in survival were found between SA patients with predicted poor outcomes and the controls (p > 0.05). It was estimated that if ESCCPro was used to determine SA and SPOCT eligibility, the median OS in the ESCCPro-recommended SA group and SPOCT group would have been 15.3 months and 24.9 months longer, respectively. In addition, ESCCPro also significantly improved prognosis accuracy, certainty, and the efficiency of clinical experts.

CONCLUSION: ESCCPro assistance improved the survival benefit of ESCC patients and the clinical decision-making among the two treatment approaches.

CRITICAL RELEVANCE STATEMENT: The ESCCPro model for treatment decision-making is promising to improve overall survival in ESCC patients undergoing surgical resection and patients undergoing surgery followed by postoperative adjuvant chemotherapy.

KEY POINTS: ESCC is associated with a poor prognosis and unclear ideal treatments. ESCCPro predicts the survival of patients with ESCC and the expected benefit from SA. ESCCPro improves clinicians' stratification of patients' prognoses.

PMID:38913225 | DOI:10.1186/s13244-024-01737-1

Radiologie (Heidelb). 2024 Jun 24. doi: 10.1007/s00117-024-01335-8. Online ahead of print.

ABSTRACT

BACKGROUND: Artificial intelligence (AI) has the potential to fundamentally change radiology workflow.

OBJECTIVES: This review article provides an overview of AI applications in cardiovascular radiology with a focus on image acquisition, image reconstruction, and workflow optimization.

MATERIALS AND METHODS: First, established applications of AI are presented for cardiovascular computed tomography (CT) and magnetic resonance imaging (MRI). Building on this, we describe the range of applications that are currently being developed and evaluated. The practical benefits, opportunities, and potential risks of artificial intelligence in cardiovascular imaging are critically discussed. The presentation is based on the relevant specialist literature and our own clinical and scientific experience.

RESULTS: AI-based techniques for image reconstruction are already commercially available and enable dose reduction in cardiovascular CT and accelerated image acquisition in cardiac MRI. Postprocessing of cardiovascular CT and MRI examinations can already be considerably simplified using established AI-based segmentation algorithms. In contrast, the practical benefits of many AI applications aimed at the diagnosis of cardiovascular diseases are less evident. Potential risks such as automation bias and considerations regarding cost efficiency should also be taken into account.

CONCLUSIONS: In a market characterized by great expectations and rapid technical development, it is important to realistically assess the practical benefits of AI applications for your own hospital or practice.

PMID:38913176 | DOI:10.1007/s00117-024-01335-8

Int Urogynecol J. 2024 Jun 24. doi: 10.1007/s00192-024-05841-0. Online ahead of print.

ABSTRACT

INTRODUCTION AND HYPOTHESIS: The objective was to create and validate the usefulness of a convolutional neural network (CNN) for identifying different organs of the pelvic floor in the midsagittal plane via dynamic ultrasound.

METHODS: This observational and prospective study included 110 patients. Transperineal ultrasound scans were performed by an expert sonographer of the pelvic floor. A video of each patient was made that captured the midsagittal plane of the pelvic floor at rest and the change in the pelvic structures during the Valsalva maneuver. After saving the captured videos, we manually labeled the different organs in each video. Three different architectures were tested-UNet, FPN, and LinkNet-to determine which CNN model best recognized anatomical structures. The best model was trained with the 86 cases for the number of epochs determined by the stop criterion via cross-validation. The Dice Similarity Index (DSI) was used for CNN validation.

RESULTS: Eighty-six patients were included to train the CNN and 24 to test the CNN. After applying the trained CNN to the 24 test videos, we did not observe any failed segmentation. In fact, we obtained a DSI of 0.79 (95% CI: 0.73 - 0.82) as the median of the 24 test videos. When we studied the organs independently, we observed differences in the DSI of each organ. The poorest DSIs were obtained in the bladder (0.71 [95% CI: 0.70 - 0.73]) and uterus (0.70 [95% CI: 0.68 - 0.74]), whereas the highest DSIs were obtained in the anus (0.81 [95% CI: 0.80 - 0.86]) and levator ani muscle (0.83 [95% CI: 0.82 - 0.83]).

CONCLUSIONS: Our results show that it is possible to apply deep learning using a trained CNN to identify different pelvic floor organs in the midsagittal plane via dynamic ultrasound.

PMID:38913129 | DOI:10.1007/s00192-024-05841-0

Brief Funct Genomics. 2024 Jun 23:elae027. doi: 10.1093/bfgp/elae027. Online ahead of print.

ABSTRACT

RNA interference (RNAi) technology is widely used in the biological prevention and control of terrestrial insects. One of the main factors with the application of RNAi in insects is the difference in RNAi efficiency, which may vary not only in different insects, but also in different genes of the same insect, and even in different double-stranded RNAs (dsRNAs) of the same gene. This work focuses on the last question and establishes a bioinformatics software that can help researchers screen for the most efficient dsRNA targeting target genes. Among insects, the red flour beetle (Tribolium castaneum) is known to be one of the most sensitive to RNAi. From iBeetle-Base, we extracted 12 027 efficient dsRNA sequences with a lethality rate of ≥20% or with experimentation-induced phenotypic changes and processed these data to correspond to specific silence efficiency. Based on the first complied novel benchmark dataset, we specifically designed a deep neural network to identify and characterize efficient dsRNA for RNAi in insects. The dna2vec word embedding model was trained to extract distributed feature representations, and three powerful modules, namely convolutional neural network, bidirectional long short-term memory network, and self-attention mechanism, were integrated to form our predictor model to characterize the extracted dsRNAs and their silencing efficiencies for T. castaneum. Our model dsRNAPredictor showed reliable performance in multiple independent tests based on different species, including both T. castaneum and Aedes aegypti. This indicates that dsRNAPredictor can facilitate prescreening for designing high-efficiency dsRNA targeting target genes of insects in advance.

PMID:38912767 | DOI:10.1093/bfgp/elae027

Heliyon. 2024 May 29;10(11):e32077. doi: 10.1016/j.heliyon.2024.e32077. eCollection 2024 Jun 15.

ABSTRACT

Oral cancer early diagnosis is a critical task in the field of medical science, and one of the most necessary things is to develop sound and effective strategies for early detection. The current research investigates a new strategy to diagnose an oral cancer based upon combination of effective learning and medical imaging. The current research investigates a new strategy to diagnose an oral cancer using Gated Recurrent Unit (GRU) networks optimized by an improved model of the NGO (Northern Goshawk Optimization) algorithm. The proposed approach has several advantages over existing methods, including its ability to analyze large and complex datasets, its high accuracy, as well as its capacity to detect oral cancer at the very beginning stage. The improved NGO algorithm is utilized to improve the GRU network that helps to improve the performance of the network and increase the accuracy of the diagnosis. The paper describes the proposed approach and evaluates its performance using a dataset of oral cancer patients. The findings of the study demonstrate the efficiency of the suggested approach in accurately diagnosing oral cancer.

PMID:38912510 | PMC:PMC11190545 | DOI:10.1016/j.heliyon.2024.e32077

Heliyon. 2024 Jun 5;10(11):e32472. doi: 10.1016/j.heliyon.2024.e32472. eCollection 2024 Jun 15.

ABSTRACT

Unmanned aerial vehicles (UAVs) have garnered attention for their potential to improve wireless communication networks by establishing line-of-sight (LoS) connections. However, urban environments pose challenges such as tall buildings and trees, impacting communication pathways. Intelligent reflection surfaces (IRSs) offer a solution by creating virtual LoS routes through signal reflection, enhancing reliability and coverage. This paper presents a three-dimensional dynamic channel model for UAV-assisted communication systems with IRSs. Additionally, it proposes a novel channel-tracking approach using deep learning and artificial intelligence techniques, comprising preliminary estimation with a deep neural network and continuous monitoring with a Stacked Bidirectional Long and Short-Term Memory (Bi-LSTM) model. Simulation results demonstrate faster convergence and superior performance compared to benchmarks, highlighting the effectiveness of integrating IRSs into UAV-enabled communication for enhanced reliability and efficiency.

PMID:38912507 | PMC:PMC11193030 | DOI:10.1016/j.heliyon.2024.e32472

Heliyon. 2024 May 30;10(11):e32071. doi: 10.1016/j.heliyon.2024.e32071. eCollection 2024 Jun 15.

ABSTRACT

Efficiently handling huge data amounts and enabling processing-intensive applications to run in faraway areas simultaneously is the ultimate objective of 5G networks. Currently, in order to distribute computing tasks, ongoing studies are exploring the incorporation of fog-cloud servers onto satellites, presenting a promising solution to enhance connectivity in remote areas. Nevertheless, analyzing the copious amounts of data produced by scattered sensors remains a challenging endeavor. The conventional strategy of transmitting this data to a central server for analysis can be costly. In contrast to centralized learning methods, distributed machine learning (ML) provides an alternative approach, albeit with notable drawbacks. This paper addresses the comparative learning expenses of centralized and distributed learning systems to tackle these challenges directly. It proposes the creation of an integrated system that harmoniously merges cloud servers with satellite network structures, leveraging the strengths of each system. This integration could represent a major breakthrough in satellite-based networking technology by streamlining data processing from remote nodes and cutting down on expenses. The core of this approach lies in the adaptive tailoring of learning techniques for individual entities based on their specific contextual nuances. The experimental findings underscore the prowess of the innovative lightweight strategy, LMAED2L (Enhanced Deep Learning for Earth Data Analysis), across a spectrum of machine learning assignments, showcasing remarkable and consistent performance under diverse operational conditions. Through a strategic fusion of centralized and distributed learning frameworks, the LMAED2L method emerges as a dynamic and effective remedy for the intricate data analysis challenges encountered within satellite networks interfaced with cloud servers. The empirical findings reveal a significant performance boost of our novel approach over traditional methods, with an average increase in reward (4.1 %), task completion rate (3.9 %), and delivered packets (3.4 %). This report suggests that these advancements will catalyze the integration of cutting-edge machine learning algorithms within future networks, elevating responsiveness, efficiency, and resource utilization to new heights.

PMID:38912450 | PMC:PMC11190546 | DOI:10.1016/j.heliyon.2024.e32071

Heliyon. 2024 May 31;10(11):e32279. doi: 10.1016/j.heliyon.2024.e32279. eCollection 2024 Jun 15.

ABSTRACT

Early cancer detection and treatment depend on the discovery of specific genes that cause cancer. The classification of genetic mutations was initially done manually. However, this process relies on pathologists and can be a time-consuming task. Therefore, to improve the precision of clinical interpretation, researchers have developed computational algorithms that leverage next-generation sequencing technologies for automated mutation analysis. This paper utilized four deep learning classification models with training collections of biomedical texts. These models comprise bidirectional encoder representations from transformers for Biomedical text mining (BioBERT), a specialized language model implemented for biological contexts. Impressive results in multiple tasks, including text classification, language inference, and question answering, can be obtained by simply adding an extra layer to the BioBERT model. Moreover, bidirectional encoder representations from transformers (BERT), long short-term memory (LSTM), and bidirectional LSTM (BiLSTM) have been leveraged to produce very good results in categorizing genetic mutations based on textual evidence. The dataset used in the work was created by Memorial Sloan Kettering Cancer Center (MSKCC), which contains several mutations. Furthermore, this dataset poses a major classification challenge in the Kaggle research prediction competitions. In carrying out the work, three challenges were identified: enormous text length, biased representation of the data, and repeated data instances. Based on the commonly used evaluation metrics, the experimental results show that the BioBERT model outperforms other models with an F1 score of 0.87 and 0.850 MCC, which can be considered as improved performance compared to similar results in the literature that have an F1 score of 0.70 achieved with the BERT model.

PMID:38912449 | PMC:PMC11190593 | DOI:10.1016/j.heliyon.2024.e32279

Precis Clin Med. 2024 May 29;7(2):pbae012. doi: 10.1093/pcmedi/pbae012. eCollection 2024 Jun.

ABSTRACT

BACKGROUND: The prognosis of breast cancer is often unfavorable, emphasizing the need for early metastasis risk detection and accurate treatment predictions. This study aimed to develop a novel multi-modal deep learning model using preoperative data to predict disease-free survival (DFS).

METHODS: We retrospectively collected pathology imaging, molecular and clinical data from The Cancer Genome Atlas and one independent institution in China. We developed a novel Deep Learning Clinical Medicine Based Pathological Gene Multi-modal (DeepClinMed-PGM) model for DFS prediction, integrating clinicopathological data with molecular insights. The patients included the training cohort (n = 741), internal validation cohort (n = 184), and external testing cohort (n = 95).

RESULT: Integrating multi-modal data into the DeepClinMed-PGM model significantly improved area under the receiver operating characteristic curve (AUC) values. In the training cohort, AUC values for 1-, 3-, and 5-year DFS predictions increased to 0.979, 0.957, and 0.871, while in the external testing cohort, the values reached 0.851, 0.878, and 0.938 for 1-, 2-, and 3-year DFS predictions, respectively. The DeepClinMed-PGM's robust discriminative capabilities were consistently evident across various cohorts, including the training cohort [hazard ratio (HR) 0.027, 95% confidence interval (CI) 0.0016-0.046, P < 0.0001], the internal validation cohort (HR 0.117, 95% CI 0.041-0.334, P < 0.0001), and the external cohort (HR 0.061, 95% CI 0.017-0.218, P < 0.0001). Additionally, the DeepClinMed-PGM model demonstrated C-index values of 0.925, 0.823, and 0.864 within the three cohorts, respectively.

CONCLUSION: This study introduces an approach to breast cancer prognosis, integrating imaging and molecular and clinical data for enhanced predictive accuracy, offering promise for personalized treatment strategies.

PMID:38912415 | PMC:PMC11190375 | DOI:10.1093/pcmedi/pbae012

Front Med (Lausanne). 2024 Jun 3;11:1409314. doi: 10.3389/fmed.2024.1409314. eCollection 2024.

ABSTRACT

The rapid spread of COVID-19 pandemic across the world has not only disturbed the global economy but also raised the demand for accurate disease detection models. Although many studies have proposed effective solutions for the early detection and prediction of COVID-19 with Machine Learning (ML) and Deep learning (DL) based techniques, but these models remain vulnerable to data privacy and security breaches. To overcome the challenges of existing systems, we introduced Adaptive Differential Privacy-based Federated Learning (DPFL) model for predicting COVID-19 disease from chest X-ray images which introduces an innovative adaptive mechanism that dynamically adjusts privacy levels based on real-time data sensitivity analysis, improving the practical applicability of Federated Learning (FL) in diverse healthcare environments. We compared and analyzed the performance of this distributed learning model with a traditional centralized model. Moreover, we enhance the model by integrating a FL approach with an early stopping mechanism to achieve efficient COVID-19 prediction with minimal communication overhead. To ensure privacy without compromising model utility and accuracy, we evaluated the proposed model under various noise scales. Finally, we discussed strategies for increasing the model's accuracy while maintaining robustness as well as privacy.

PMID:38912338 | PMC:PMC11193384 | DOI:10.3389/fmed.2024.1409314

Indian J Radiol Imaging. 2023 Oct 10;34(3):469-487. doi: 10.1055/s-0043-1775737. eCollection 2024 Jul.

ABSTRACT

Background Although abundant literature is currently available on the use of deep learning for breast cancer detection in mammography, the quality of such literature is widely variable. Purpose To evaluate published literature on breast cancer detection in mammography for reproducibility and to ascertain best practices for model design. Methods The PubMed and Scopus databases were searched to identify records that described the use of deep learning to detect lesions or classify images into cancer or noncancer. A modification of Quality Assessment of Diagnostic Accuracy Studies (mQUADAS-2) tool was developed for this review and was applied to the included studies. Results of reported studies (area under curve [AUC] of receiver operator curve [ROC] curve, sensitivity, specificity) were recorded. Results A total of 12,123 records were screened, of which 107 fit the inclusion criteria. Training and test datasets, key idea behind model architecture, and results were recorded for these studies. Based on mQUADAS-2 assessment, 103 studies had high risk of bias due to nonrepresentative patient selection. Four studies were of adequate quality, of which three trained their own model, and one used a commercial network. Ensemble models were used in two of these. Common strategies used for model training included patch classifiers, image classification networks (ResNet in 67%), and object detection networks (RetinaNet in 67%). The highest reported AUC was 0.927 ± 0.008 on a screening dataset, while it reached 0.945 (0.919-0.968) on an enriched subset. Higher values of AUC (0.955) and specificity (98.5%) were reached when combined radiologist and Artificial Intelligence readings were used than either of them alone. None of the studies provided explainability beyond localization accuracy. None of the studies have studied interaction between AI and radiologist in a real world setting. Conclusion While deep learning holds much promise in mammography interpretation, evaluation in a reproducible clinical setting and explainable networks are the need of the hour.

PMID:38912238 | PMC:PMC11188703 | DOI:10.1055/s-0043-1775737

J Biomed Opt. 2024 Jul;29(7):076501. doi: 10.1117/1.JBO.29.7.076501. Epub 2024 Jun 18.

ABSTRACT

SIGNIFICANCE: Information about the spatial organization of fibers within a nerve is crucial to our understanding of nerve anatomy and its response to neuromodulation therapies. A serial block-face microscopy method [three-dimensional microscopy with ultraviolet surface excitation (3D-MUSE)] has been developed to image nerves over extended depths ex vivo. To routinely visualize and track nerve fibers in these datasets, a dedicated and customizable software tool is required.

AIM: Our objective was to develop custom software that includes image processing and visualization methods to perform microscopic tractography along the length of a peripheral nerve sample.

APPROACH: We modified common computer vision algorithms (optic flow and structure tensor) to track groups of peripheral nerve fibers along the length of the nerve. Interactive streamline visualization and manual editing tools are provided. Optionally, deep learning segmentation of fascicles (fiber bundles) can be applied to constrain the tracts from inadvertently crossing into the epineurium. As an example, we performed tractography on vagus and tibial nerve datasets and assessed accuracy by comparing the resulting nerve tracts with segmentations of fascicles as they split and merge with each other in the nerve sample stack.

RESULTS: We found that a normalized Dice overlap ( Dice norm ) metric had a mean value above 0.75 across several millimeters along the nerve. We also found that the tractograms were robust to changes in certain image properties (e.g., downsampling in-plane and out-of-plane), which resulted in only a 2% to 9% change to the mean Dice norm values. In a vagus nerve sample, tractography allowed us to readily identify that subsets of fibers from four distinct fascicles merge into a single fascicle as we move ∼ 5 mm along the nerve's length.

CONCLUSIONS: Overall, we demonstrated the feasibility of performing automated microscopic tractography on 3D-MUSE datasets of peripheral nerves. The software should be applicable to other imaging approaches. The code is available at https://github.com/ckolluru/NerveTracker.

PMID:38912214 | PMC:PMC11188586 | DOI:10.1117/1.JBO.29.7.076501

J Biomed Opt. 2024 Jul;29(7):076001. doi: 10.1117/1.JBO.29.7.076001. Epub 2024 Jun 18.

ABSTRACT

SIGNIFICANCE: Retinopathy of prematurity (ROP) poses a significant global threat to childhood vision, necessitating effective screening strategies. This study addresses the impact of color channels in fundus imaging on ROP diagnosis, emphasizing the efficacy and safety of utilizing longer wavelengths, such as red or green for enhanced depth information and improved diagnostic capabilities.

AIM: This study aims to assess the spectral effectiveness in color fundus photography for the deep learning classification of ROP.

APPROACH: A convolutional neural network end-to-end classifier was utilized for deep learning classification of normal, stage 1, stage 2, and stage 3 ROP fundus images. The classification performances with individual-color-channel inputs, i.e., red, green, and blue, and multi-color-channel fusion architectures, including early-fusion, intermediate-fusion, and late-fusion, were quantitatively compared.

RESULTS: For individual-color-channel inputs, similar performance was observed for green channel (88.00% accuracy, 76.00% sensitivity, and 92.00% specificity) and red channel (87.25% accuracy, 74.50% sensitivity, and 91.50% specificity), which is substantially outperforming the blue channel (78.25% accuracy, 56.50% sensitivity, and 85.50% specificity). For multi-color-channel fusion options, the early-fusion and intermediate-fusion architecture showed almost the same performance when compared to the green/red channel input, and they outperformed the late-fusion architecture.

CONCLUSIONS: This study reveals that the classification of ROP stages can be effectively achieved using either the green or red image alone. This finding enables the exclusion of blue images, acknowledged for their increased susceptibility to light toxicity.

PMID:38912212 | PMC:PMC11188587 | DOI:10.1117/1.JBO.29.7.076001

Front Cell Infect Microbiol. 2024 Jun 7;14:1397316. doi: 10.3389/fcimb.2024.1397316. eCollection 2024.

ABSTRACT

While the world struggles to recover from the devastation wrought by the widespread spread of COVID-19, monkeypox virus has emerged as a new global pandemic threat. In this paper, a high precision and lightweight classification network MpoxNet based on ConvNext is proposed to meet the need of fast and safe detection of monkeypox classification. In this method, a two-branch depth-separable convolution residual Squeeze and Excitation module is designed. This design aims to extract more feature information with two branches, and greatly reduces the number of parameters in the model by using depth-separable convolution. In addition, our method introduces a convolutional attention module to enhance the extraction of key features within the receptive field. The experimental results show that MpoxNet has achieved remarkable results in monkeypox disease classification, the accuracy rate is 95.28%, the precision rate is 96.40%, the recall rate is 93.00%, and the F1-Score is 95.80%. This is significantly better than the current mainstream classification model. It is worth noting that the FLOPS and the number of parameters of MpoxNet are only 30.68% and 31.87% of those of ConvNext-Tiny, indicating that the model has a small computational burden and model complexity while efficient performance.

PMID:38912211 | PMC:PMC11190078 | DOI:10.3389/fcimb.2024.1397316

Hand Surg Rehabil. 2024 Jun 21:101742. doi: 10.1016/j.hansur.2024.101742. Online ahead of print.

ABSTRACT

This study proposes a deep-learning algorithm to automatically detect perilunate dislocation on anteroposterior wrist radiographs. A total of 374 annotated radiographs, 345 normal and 29 pathological, were used to train, validate and test two YOLO v8 deep neural models. The first model was used for detecting the carpal region, and the second for segmenting a region between Gilula's second and third arcs. The output of the segmentation model, trained multiple times with varying random initial parameter values and augmentations, was then assigned a probability of being normal or pathological through ensemble averaging. In this dataset, the algorithm achieved an overall F1-score of 0.880: 0.928 in the normal subgroup, with 1.0 precision, and 0.833 in the pathological subgroup with 1.0 recall (or sensitivity), demonstrating that the diagnosis of perilunate dislocation can be improved through automatic analysis of anteroposterior radiographs. Level of evidence: III.

PMID:38909690 | DOI:10.1016/j.hansur.2024.101742