Clin Transl Gastroenterol. 2024 Sep 26. doi: 10.14309/ctg.0000000000000771. Online ahead of print.

ABSTRACT

Endoscopic ultrasound (EUS) allows characterization and biopsy of pancreatic lesions. Pancreatic cystic neoplasms (PCN) include in mucinous (M-PCN) and non-mucinous lesions (NM-PCN). Pancreatic ductal adenocarcinoma (P-DAC) is the commonest pancreatic solid lesion (PSL), followed by pancreatic neuroendocrine tumor (P-NET). While EUS is preferred for pancreatic lesion evaluation, its diagnostic accuracy is suboptimal. This multicentric study aims to develop a convolutional neural network (CNN) for detecting and distinguishing PCN (namely M-PCN and NM-PCN) and PSL (particularly P-DAC and P-NET). A CNN was developed with 378 EUS exams from 4 international reference centers (Centro Hospitalar Universitário São João, Hospital Universitario Puerta de Hierro Majadahonda, New York University Hospitals, Hospital das Clínicas FMUSP). 126.000 images were obtained - 19.528 M-PCN, 8.175 NM-PCN, 64.286 P-DAC, 29.153 P-NET and 4.858 normal pancreas images. A trinary CNN differentiated normal pancreas tissue from M-PCN and NM-PCN. A binary CNN distinguished P-DAC from P-NET. The total dataset was divided in a training and testing dataset (used for model's evaluation) in a 90/10% ratio. The model was evaluated through its sensitivity, specificity, positive and negative predictive values and accuracy. The CNN had 99.1% accuracy for identifying normal pancreatic tissue, 99.0% and 99.8% for M-PCN and NM-PCN, respectively. P-DAC and P-NET were distinguished with 94.0% accuracy. Our group developed the first worldwide CNN capable of detecting and differentiating the commonest PCN and PSL in EUS images, using exams from 4 centers in two continents, minimizing the impact of the demographic bias. Larger multicentric studies are needed for technology implementation.

PMID:39324610 | DOI:10.14309/ctg.0000000000000771

Pest Manag Sci. 2024 Sep 26. doi: 10.1002/ps.8442. Online ahead of print.

ABSTRACT

BACKGROUND: Forest ecosystems are under constant threat from wood-boring pests such as the Emerald ash borer (EAB), which remain elusive owing to their hidden life cycles within tree trunks. Early detection is vital to mitigate economic and ecological damage. The main current monitoring method is manual detection which is ineffective at early stages of infestation. This study introduces VibroEABNet, a deep learning-based joint recognition network designed to enhance the detection of EAB boring vibration signals, with a novel approach integrating denoising and recognition modules.

RESULTS: The proposed VibroEABNet model demonstrated exceptional performance, achieving an average accuracy of 98.98% across multiple signal-to-noise ratios (SNRs) in test datasets and a remarkable 97.5% accuracy in real forest datasets, surpassing traditional models and other deep learning networks evaluated in this study. These findings were supported by rigorous noise resistance analysis and real dataset evaluation, indicating the model's robustness and reliability in practical applications. Furthermore, the model's efficiency was highlighted by its inference time of 26 ms and a compact model size of 8.43 MB, underscoring its suitability for deployment in resource-limited environments.

CONCLUSION: The development of VibroEABNet marks a significant advancement in pest detection methodologies, offering a scalable, accurate and efficient solution for early monitoring of wood-boring pests. The integration of a denoising module within the network structure addresses the challenge of environmental noise, one of the primary limitations in acoustic monitoring of pests. Currently, this research is limited to a specific pest. Future work will focus on the applicability of this network to other wood-boring pests. © 2024 Society of Chemical Industry.

PMID:39324448 | DOI:10.1002/ps.8442

Huan Jing Ke Xue. 2024 Sep 8;45(9):5179-5187. doi: 10.13227/j.hjkx.202310217.

ABSTRACT

Accurate prediction of PM10 concentration is important for effectively managing PM10 exposure and mitigating health and economic risks posed to humans in dry bulk ports. However, accurately capturing the time-series nonlinear variation characteristics of PM10 concentration is challenging owing to the specific intensity of port operation activities and the influence of meteorological factors. To address such challenges, a novel combined deep learning model （CLAF） was proposed, merging cascaded convolutional neural networks （CNN）, long short-term memory （LSTM）, and an attention mechanism （AM）. This integrated model aimed to forecast hourly PM10 concentration in dry bulk ports. First, using the random forest characteristic importance algorithm, the distinct meteorological factors were identified among the selected five meteorological factors. These selected factors were incorporated into the prediction model along with the PM10 concentration. Subsequently, the CNN layer was employed to extract high-dimensional time-varying features from the input variables, while the LSTM layer captured sequential features and long-term dependencies. In the AM layer, different weights were assigned to the output components of the LSTM layer to amplify the effects of important information. Finally, three evaluation metrics were applied to compare the performance of the CLAF model with three basic models and three commonly used prediction models. Real-case data was collected and used in this study. Comparison results demonstrated that considering the meteorological factors could improve the prediction accuracy and fitting performance of PM10 concentration in ports. The CLAF model reduced the mean absolute error statistic by 13.92%-56.9%, decreased the mean square error statistic by 45.99%-81.02%, and improved the goodness-of-fit statistic by 3.2%-15.5%.

PMID:39323136 | DOI:10.13227/j.hjkx.202310217

J Med Internet Res. 2024 Sep 25. doi: 10.2196/58572. Online ahead of print.

ABSTRACT

BACKGROUND: While speech analysis holds promise for mental health assessment, research often focuses on single symptoms, despite symptom co-occurrences and interactions. In addition, predictive models in mental health do not properly assess speech-based systems' limitations, such as uncertainty, or fairness for a safe clinical deployment.

OBJECTIVE: We investigated the predictive potential of mobile-collected speech data for detecting and estimating depression, anxiety, fatigue, and insomnia, focusing on other factors than mere accuracy, in the general population.

METHODS: We included n=865 healthy adults and recorded their answers regarding their perceived mental and sleep states. We asked how they felt and if they had slept well lately. Clinically validated questionnaires measuring depression, anxiety, insomnia, and fatigue severity were also used. We developed a novel speech and machine learning pipeline involving voice activity detection, feature extraction, and model training. We automatically analyzed participants' speech with a fully ML automatic pipeline to capture speech variability. Then, we modelled speech with pretrained deep learning models that were pre-trained on a large open free database and we selected the best one on the validation set. Based on the best speech modelling approach, we evaluated clinical threshold detection, individual score prediction, model uncertainty estimation, and performance fairness across demographics (age, sex, education). We employed a train-validation-test split for all evaluations: to develop our models, select the best ones and assess the generalizability of held-out data.

RESULTS: The best model was WhisperM with a max pooling, and oversampling method. Our methods achieved good detection performance for all symptoms, depression (PHQ-9 AUC= 0.76F1=0.49, BDI AUC=0.78, F1=0,65), anxiety (GAD-7 F1=0.50, AUC=0.77) insomnia (AIS AUC=0.73, F1=0.62), and fatigue (MFI Total Score F1=0.88, AUC=0.68). These strengths were maintained for depression detection with BDI and Fatigue for abstention rates for uncertain cases (Risk-Coverage AUCs < 0.4). Individual symptom scores were predicted with good accuracy (Correlations were all significant, with Pearson strengths between 0.31 and 0.49). Fairness analysis revealed that models were consistent for sex (average Disparity Ratio (DR) = 0.86), to a lesser extent for education level (average Disparity Ratio (DR) = 0.47) and worse for age groups (average Disparity Ratio (DR) = 0.33).

CONCLUSIONS: This study demonstrates the potential of speech-based systems for multifaceted mental health assessment in the general population, not only for detecting clinical thresholds but also for estimating their severity. Addressing fairness and incorporating uncertainty estimation with selective classification are key contributions that can enhance the clinical utility and responsible implementation of such systems. This approach offers promise for more accurate and nuanced mental health assessments, benefiting both patients and clinicians.

PMID:39324329 | DOI:10.2196/58572

Front Artif Intell. 2024 Sep 11;7:1444136. doi: 10.3389/frai.2024.1444136. eCollection 2024.

ABSTRACT

BACKGROUND: Glaucoma (GLAU), Age-related Macular Degeneration (AMD), Retinal Vein Occlusion (RVO), and Diabetic Retinopathy (DR) are common blinding ophthalmic diseases worldwide.

PURPOSE: This approach is expected to enhance the early detection and treatment of common blinding ophthalmic diseases, contributing to the reduction of individual and economic burdens associated with these conditions.

METHODS: We propose an effective deep-learning pipeline that combine both segmentation model and classification model for diagnosis and grading of four common blinding ophthalmic diseases and normal retinal fundus.

RESULTS: In total, 102,786 fundus images of 75,682 individuals were used for training validation and external validation purposes. We test our model on internal validation data set, the micro Area Under the Receiver Operating Characteristic curve (AUROC) of which reached 0.995. Then, we fine-tuned the diagnosis model to classify each of the four disease into early and late stage, respectively, which achieved AUROCs of 0.597 (GL), 0.877 (AMD), 0.972 (RVO), and 0.961 (DR) respectively. To test the generalization of our model, we conducted two external validation experiments on Neimeng and Guangxi cohort, all of which maintained high accuracy.

CONCLUSION: Our algorithm demonstrates accurate artificial intelligence diagnosis pipeline for common blinding ophthalmic diseases based on Lesion-Focused fundus that overcomes the low-accuracy of the traditional classification method that based on raw retinal images, which has good generalization ability on diverse cases in different regions.

PMID:39324131 | PMC:PMC11422385 | DOI:10.3389/frai.2024.1444136

Indian J Orthop. 2024 Sep 11;58(10):1458-1473. doi: 10.1007/s43465-024-01259-4. eCollection 2024 Oct.

ABSTRACT

INTRODUCTION: Knee osteoarthritis (OA) is a prevalent condition that significantly impacts the quality of life, often leading to the need for knee replacement surgery. Accurate and timely identification of knee degeneration is crucial for effective treatment and management. Traditional methods of diagnosing OA rely heavily on radiological assessments, which can be time-consuming and subjective. This study aims to address these challenges by developing a deep learning-based method to predict the likelihood of knee replacement and the Kellgren-Lawrence (KL) grade of knee OA from X-ray images.

METHODOLOGY: We employed the Osteoarthritis Initiative (OAI) dataset and utilized a transfer learning approach with the Inception V3 architecture to enhance the accuracy of OA detection. Our approach involved training 14 different models-Xception, VGG16, VGG19, ResNet50, ResNet101, ResNet152, ResNet50V2, ResNet101V2, ResNet152V2, Inception V3, Inception, ResNetV2, DenseNet121, DenseNet169, DenseNet201-and comparing their performance.

RESULTS: The study incorporated pixel ratio computation and picture pre-processing, alongside a decision tree model for prediction. Our experiments revealed that the Inception V3 model achieved the highest training accuracy of 91% and testing accuracy of 67%, with notable performance in both training and validation phases. This model effectively identified the presence and severity of OA, correlating with the Kellgren-Lawrence scale and facilitating the assessment of knee replacement needs.

CONCLUSION: By integrating advanced deep learning techniques with radiological diagnostics, our methodology supports radiologists in making more accurate and prompt decisions regarding knee degeneration. The Inception V3 model stands out as the optimal choice for knee X-ray analysis, contributing to more efficient and timely healthcare delivery for patients with knee osteoarthritis.

PMID:39324090 | PMC:PMC11420401 | DOI:10.1007/s43465-024-01259-4

Indian J Orthop. 2024 Jun 2;58(10):1362-1374. doi: 10.1007/s43465-024-01189-1. eCollection 2024 Oct.

ABSTRACT

BACKGROUND AND AIMS: Regenerative orthopedics involves approaches like stem cell therapy, platelet-rich plasma (PRP) therapy, the use of biological scaffold implants, tissue engineering, etc. We aim to present a scoping review of the role of artificial intelligence (AI) in different treatment approaches of regenerative orthopedics.

METHODS: Using the PRISMA guidelines, a search for articles for the last ten years (2013-2024) on PubMed was done, using several keywords. We have discussed the state-of-the-art, strengths/benefits, and limitations of the published research, and provide a useful resource for the way ahead in future for researchers working in this area.

RESULTS: Using the eligibility criteria out of 82 initially screened publications, we included 18 studies for this review. We noticed that the treatment responses to regenerative treatments depend on several factors; hence, to facilitate better comprehensive and patient-specific treatments, AI technology is very useful. Machine learning (ML) and deep learning (DL) are a few of the most frequently used AI techniques. They use a data-driven approach for training models to make human-like decisions. Data are fed to the ML/DL algorithm and the trained model makes classifications or predictions based on its learning.

CONCLUSION: The area of regenerative orthopedics is highly sophisticated and significantly aids in providing cost-effective and non-invasive treatments to patients suffering from orthopedic ailments and injuries. Due to its promising future, the use of AI in regenerative orthopedics is an emerging and promising research field; however, its universal clinical applications are associated with some ethical considerations, which need addressing.

SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s43465-024-01189-1.

PMID:39324081 | PMC:PMC11420425 | DOI:10.1007/s43465-024-01189-1

MethodsX. 2024 Sep 12;13:102946. doi: 10.1016/j.mex.2024.102946. eCollection 2024 Dec.

ABSTRACT

The rapid advancement in Artificial Intelligence (AI) and big data has developed significance in the water sector, particularly in hydrological time-series predictions. Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTM) networks have become research focal points due to their effectiveness in modeling non-linear, time-variant hydrological systems. This review explores the different architectures of RNNs, LSTMs, and Gated Recurrent Units (GRUs) and their efficacy in predicting hydrological time-series data.•RNNs are foundational but face limitations such as vanishing gradients, which impede their ability to model long-term dependencies. LSTMs and GRUs have been developed to overcome these limitations, with LSTMs using memory cells and gating mechanisms, while GRUs provide a more streamlined architecture with similar benefits.•The integration of attention mechanisms and hybrid models that combine RNNs, LSTMs, and GRUs with other Machine learning (ML) and Deep Learning (DL) has improved prediction accuracy by capturing both temporal and spatial dependencies.•Despite their effectiveness, practical implementations of these models in hydrological time series prediction require extensive datasets and substantial computational resources. Future research should develop interpretable architectures, enhance data quality, incorporate domain knowledge, and utilize transfer learning to improve model generalization and scalability across diverse hydrological contexts.

PMID:39324077 | PMC:PMC11422155 | DOI:10.1016/j.mex.2024.102946

Bone Rep. 2024 Sep 16;22:101801. doi: 10.1016/j.bonr.2024.101801. eCollection 2024 Sep.

ABSTRACT

Femur fractures are a significant worldwide public health concern that affects patients as well as their families because of their high frequency, morbidity, and mortality. When employing computer-aided diagnostic (CAD) technologies, promising results have been shown in the efficiency and accuracy of fracture classification, particularly with the growing use of Deep Learning (DL) approaches. Nevertheless, the complexity is further increased by the need to collect enough input data to train these algorithms and the challenge of interpreting the findings. By improving on the results of the most recent deep learning-based Arbeitsgemeinschaft für Osteosynthesefragen and Orthopaedic Trauma Association (AO/OTA) system classification of femur fractures, this study intends to support physicians in making correct and timely decisions regarding patient care. A state-of-the-art architecture, YOLOv8, was used and refined while paying close attention to the interpretability of the model. Furthermore, data augmentation techniques were involved during preprocessing, increasing the dataset samples through image processing alterations. The fine-tuned YOLOv8 model achieved remarkable results, with 0.9 accuracy, 0.85 precision, 0.85 recall, and 0.85 F1-score, computed by averaging the values among all the individual classes for each metric. This study shows the proposed architecture's effectiveness in enhancing the AO/OTA system's classification of femur fractures, assisting physicians in making prompt and accurate diagnoses.

PMID:39324016 | PMC:PMC11422035 | DOI:10.1016/j.bonr.2024.101801

Front Neurosci. 2024 Sep 11;18:1411538. doi: 10.3389/fnins.2024.1411538. eCollection 2024.

ABSTRACT

BACKGROUND: Stroke stands as a prominent cause of mortality and disability worldwide, posing a major public health concern. Recent years have witnessed rapid advancements in artificial intelligence (AI). Studies have explored the utilization of AI in imaging analysis, assistive rehabilitation, treatment, clinical decision-making, and outcome and risk prediction concerning stroke. However, there is still a lack of systematic bibliometric analysis to discern the current research status, hotspots, and possible future development trends of AI applications in stroke.

METHODS: The publications on the application of AI in stroke were retrieved from the Web of Science Core Collection, spanning 2004-2024. Only articles or reviews published in English were included in this study. Subsequently, a manual screening process was employed to eliminate literature not pertinent to the topic. Visualization diagrams for comprehensive and in-depth analysis of the included literature were generated using CiteSpace, VOSviewer, and Charticulator.

RESULTS: This bibliometric analysis included a total of 2,447 papers, and the annual publication volume shows a notable upward trajectory. The most prolific authors, countries, and institutions are Dukelow, Sean P., China, and the University of Calgary, respectively, making significant contributions to the advancement of this field. Notably, stable collaborative networks among authors and institutions have formed. Through clustering and citation burst analysis of keywords and references, the current research hotspots have been identified, including machine learning, deep learning, and AI applications in stroke rehabilitation and imaging for early diagnosis. Moreover, emerging research trends focus on machine learning as well as stroke outcomes and risk prediction.

CONCLUSION: This study provides a comprehensive and in-depth analysis of the literature regarding AI in stroke, facilitating a rapid comprehension of the development status, cooperative networks, and research priorities within the field. Furthermore, our analysis may provide a certain reference and guidance for future research endeavors.

PMID:39323917 | PMC:PMC11422388 | DOI:10.3389/fnins.2024.1411538

Int IEEE EMBS Conf Neural Eng. 2023 Apr;2023. doi: 10.1109/ner52421.2023.10123751. Epub 2023 May 19.

ABSTRACT

Brain-computer interfaces (BCIs) can potentially restore lost function in patients with neurological injury. A promising new application of BCI technology has focused on speech restoration. One approach is to synthesize speech from the neural correlates of a person who cannot speak, as they attempt to do so. However, there is no established gold-standard for quantifying the quality of BCI-synthesized speech. Quantitative metrics, such as applying correlation coefficients between true and decoded speech, are not applicable to anarthric users and fail to capture intelligibility by actual human listeners; by contrast, methods involving people completing forced-choice multiple-choice questionnaires are imprecise, not practical at scale, and cannot be used as cost functions for improving speech decoding algorithms. Here, we present a deep learning-based "AI Listener" that can be used to evaluate BCI speech intelligibility objectively, rapidly, and automatically. We begin by adapting several leading Automatic Speech Recognition (ASR) deep learning models - Deepspeech, Wav2vec 2.0, and Kaldi - to suit our application. We then evaluate the performance of these ASRs on multiple speech datasets with varying levels of intelligibility, including: healthy speech, speech from people with dysarthria, and synthesized BCI speech. Our results demonstrate that the multiple-language ASR model XLSR-Wav2vec 2.0, trained to output phonemes, yields superior performance in terms of speech transcription accuracy. Notably, the AI Listener reports that several previously published BCI output datasets are not intelligible, which is consistent with human listeners.

PMID:39323876 | PMC:PMC11424044 | DOI:10.1109/ner52421.2023.10123751

Cureus. 2024 Aug 26;16(8):e67844. doi: 10.7759/cureus.67844. eCollection 2024 Aug.

ABSTRACT

Diabetic retinopathy (DR) remains a leading cause of vision loss worldwide, with early detection critical for preventing irreversible damage. This review explores the current landscape and future directions of artificial intelligence (AI)-enhanced detection of DR from fundus images. Recent advances in deep learning and computer vision have enabled AI systems to analyze retinal images with expert-level accuracy, potentially transforming DR screening. Key developments include convolutional neural networks achieving high sensitivity and specificity in detecting referable DR, multi-task learning approaches that can simultaneously detect and grade DR severity, and lightweight models enabling deployment on mobile devices. While these AI systems show promise in improving the efficiency and accessibility of DR screening, several challenges remain. These include ensuring generalizability across diverse populations, standardizing image acquisition and quality, addressing the "black box" nature of complex models, and integrating AI seamlessly into clinical workflows. Future directions in the field encompass explainable AI to enhance transparency, federated learning to leverage decentralized datasets, and the integration of AI with electronic health records and other diagnostic modalities. There is also growing potential for AI to contribute to personalized treatment planning and predictive analytics for disease progression. As the technology continues to evolve, maintaining a focus on rigorous clinical validation, ethical considerations, and real-world implementation will be crucial for realizing the full potential of AI-enhanced DR detection in improving global eye health outcomes.

PMID:39323686 | PMC:PMC11424092 | DOI:10.7759/cureus.67844

Postgrad Med J. 2024 Sep 26:qgae122. doi: 10.1093/postmj/qgae122. Online ahead of print.

ABSTRACT

"Medical deserts" are areas with low healthcare service levels, challenging the access, quality, and sustainability of care. This qualitative narrative review examines how artificial intelligence (AI), particularly large language models (LLMs), can address these challenges by integrating with e-Health and the Internet of Medical Things to enhance services in under-resourced areas. It explores AI-driven telehealth platforms that overcome language and cultural barriers, increasing accessibility. The utility of LLMs in providing diagnostic assistance where specialist deficits exist is highlighted, demonstrating AI's role in supplementing medical expertise and improving outcomes. Additionally, the development of AI chatbots offers preliminary medical advice, serving as initial contact points in remote areas. The review also discusses AI's role in enhancing medical education and training, supporting the professional development of healthcare workers in these regions. It assesses AI's strategic use in data analysis for effective resource allocation, identifying healthcare provision gaps. AI, especially LLMs, is seen as a promising solution for bridging healthcare gaps in "medical deserts," improving service accessibility, quality, and distribution. However, continued research and development are essential to fully realize AI's potential in addressing the challenges of medical deserts.

PMID:39323384 | DOI:10.1093/postmj/qgae122

Med Biol Eng Comput. 2024 Sep 26. doi: 10.1007/s11517-024-03205-w. Online ahead of print.

ABSTRACT

With the rapid advancement of modern medical technology, microscopy imaging systems have become one of the key technologies in medical image analysis. However, manual use of microscopes presents issues such as operator dependency, inefficiency, and time consumption. To enhance the efficiency and accuracy of medical image capture and reduce the burden of subsequent quantitative analysis, this paper proposes an improved microscope salient object detection algorithm based on U2-Net, incorporating deep learning technology. The improved algorithm first enhances the network's key information extraction capability by incorporating the Convolutional Block Attention Module (CBAM) into U2-Net. It then optimizes network complexity by constructing a Simple Pyramid Pooling Module (SPPM) and uses Ghost convolution to achieve model lightweighting. Additionally, data augmentation is applied to the slides to improve the algorithm's robustness and generalization. The experimental results show that the size of the improved algorithm model is 72.5 MB, which represents a 56.85% reduction compared to the original U2-Net model size of 168.0 MB. Additionally, the model's prediction accuracy has increased from 92.24 to 97.13%, providing an efficient means for subsequent image processing and analysis tasks in microscopy imaging systems.

PMID:39322859 | DOI:10.1007/s11517-024-03205-w

Mol Syst Biol. 2024 Sep 25. doi: 10.1038/s44320-024-00065-2. Online ahead of print.

ABSTRACT

The 3D genome prediction in cancer is crucial for uncovering the impact of structural variations (SVs) on tumorigenesis, especially when they are present in noncoding regions. We present InfoHiC, a systemic framework for predicting the 3D cancer genome directly from whole-genome sequencing (WGS). InfoHiC utilizes contig-specific copy number encoding on the SV contig assembly, and performs a contig-to-total Hi-C conversion for the cancer Hi-C prediction from multiple SV contigs. We showed that InfoHiC can predict 3D genome folding from all types of SVs using breast cancer cell line data. We applied it to WGS data of patients with breast cancer and pediatric patients with medulloblastoma, and identified neo topologically associating domains. For breast cancer, we discovered super-enhancer hijacking events associated with oncogenic overexpression and poor survival outcomes. For medulloblastoma, we found SVs in noncoding regions that caused super-enhancer hijacking events of medulloblastoma driver genes (GFI1, GFI1B, and PRDM6). In addition, we provide trained models for cancer Hi-C prediction from WGS at https://github.com/dmcb-gist/InfoHiC , uncovering the impacts of SVs in cancer patients and revealing novel therapeutic targets.

PMID:39322849 | DOI:10.1038/s44320-024-00065-2

J Imaging Inform Med. 2024 Sep 25. doi: 10.1007/s10278-024-01252-1. Online ahead of print.

ABSTRACT

Acute leukemia is characterized by the swift proliferation of immature white blood cells (WBC) in the blood and bone marrow. It is categorized into acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML), depending on whether the cell-line origin is lymphoid or myeloid, respectively. Deep learning (DL) and artificial intelligence (AI) are revolutionizing medical sciences by assisting clinicians with rapid illness identification, reducing workload, and enhancing diagnostic accuracy. This paper proposes a DL-based novel BSNEU-net framework to detect acute leukemia. It comprises 4 Union Blocks (UB) and incorporates block feature map distortion (BFMD) with switchable normalization (SN) in each UB. The UB employs union convolution to extract more discriminant features. The BFMD is adapted to acquire more generalized patterns to minimize overfitting, whereas SN layers are appended to improve the model's convergence and generalization capabilities. The uniform utilization of batch normalization across convolution layers is sensitive to the mini-batch dimension changes, which is effectively remedied by incorporating an SN layer. Here, a new dataset comprising 2400 blood smear images of ALL, AML, and healthy cases is proposed, as DL methodologies necessitate a sizeable and well-annotated dataset to combat overfitting issues. Further, a heterogeneous dataset comprising 2700 smear images is created by combining four publicly accessible benchmark datasets of ALL, AML, and healthy cases. The BSNEU-net model achieved excellent performance with 99.37% accuracy on the novel dataset and 99.44% accuracy on the heterogeneous dataset. The comparative analysis signifies the superiority of the proposed methodology with comparing schemes.

PMID:39322814 | DOI:10.1007/s10278-024-01252-1

Commun Med (Lond). 2024 Sep 25;4(1):182. doi: 10.1038/s43856-024-00609-5.

ABSTRACT

BACKGROUND: Integration of speech into healthcare has intensified privacy concerns due to its potential as a non-invasive biomarker containing individual biometric information. In response, speaker anonymization aims to conceal personally identifiable information while retaining crucial linguistic content. However, the application of anonymization techniques to pathological speech, a critical area where privacy is especially vital, has not been extensively examined.

METHODS: This study investigates anonymization's impact on pathological speech across over 2700 speakers from multiple German institutions, focusing on privacy, pathological utility, and demographic fairness. We explore both deep-learning-based and signal processing-based anonymization methods.

RESULTS: We document substantial privacy improvements across disorders-evidenced by equal error rate increases up to 1933%, with minimal overall impact on utility. Specific disorders such as Dysarthria, Dysphonia, and Cleft Lip and Palate experience minimal utility changes, while Dysglossia shows slight improvements. Our findings underscore that the impact of anonymization varies substantially across different disorders. This necessitates disorder-specific anonymization strategies to optimally balance privacy with diagnostic utility. Additionally, our fairness analysis reveals consistent anonymization effects across most of the demographics.

CONCLUSIONS: This study demonstrates the effectiveness of anonymization in pathological speech for enhancing privacy, while also highlighting the importance of customized and disorder-specific approaches to account for inversion attacks.

PMID:39322637 | DOI:10.1038/s43856-024-00609-5

Acad Radiol. 2024 Sep 24:S1076-6332(24)00671-8. doi: 10.1016/j.acra.2024.09.021. Online ahead of print.

ABSTRACT

RATIONALE AND OBJECTIVES: Recent radiomics studies on predicting pathological outcomes of glioma have shown immense potential. However, the predictive ability remains suboptimal due to the tumor intrinsic heterogeneity. We aimed to achieve better pathological prediction outcomes by combining habitat analysis with deep learning.

MATERIALS AND METHODS: 387 cases of primary glioma from three hospitals were collected, along with their T1 contrast-enhanced and T2-weighted MR sequences, pathological reports and clinical histories. The training set consisted of 264 patients, 82 patients composed the test set, and 41 patients were used as the validation set for hyperparameter tuning and optimal model selection. All groups were sourced from different centers. Through radiomics, deep learning, habitat analysis and combined analysis, we extracted imaging features separately and jointly modeled them with clinical features. We identified the optimal models for predicting glioma grades, Ki67 expression levels, P53 mutation and IDH1 mutation.

RESULTS: Using a LightGBM model with DenseNet161 features based on habitat subregions, the best tumor grade prediction model was achieved. A LightGBM model with ResNet50 features based on habitat subregions yielded the best Ki67 expression level prediction model. An SVM model with Radiomics and Inception_v3 features provided the best prediction of P53 mutation. The best model for predicting IDH1 mutation was achieved by an MLP model with Radiomics features based on habitat subregions. Clinical features might be potentially helpful for the prediction with relatively weak evidence.

CONCLUSION: Habitat+Deep Learning feature extraction methods were optimal for predicting grades and Ki67 levels. Deep Learning is optimal for predicting P53 mutation, while the combination of Habitat+ Radiomics models yielded the best prediction for IDH1 mutation.

PMID:39322536 | DOI:10.1016/j.acra.2024.09.021

Clin Radiol. 2024 Aug 30:S0009-9260(24)00499-9. doi: 10.1016/j.crad.2024.08.029. Online ahead of print.

ABSTRACT

BACKGROUND: The emergence of zero echo time (ZTE) imaging has transformed bone imaging, overcoming historical limitations in capturing detailed bone structures. By minimizing the time gap between radiofrequency excitation and data acquisition, ZTE generates CT-like images. While ZTE has shown promise in various applications, its potential in assessing skull base and calvarium lesions remains unexplored. Hence we aim to introduce a novel perspective by investigating the utility of inverted ZTE images (iZTE) and pseudoCT (pCT) images for studying lesions in the skull base and calvarium.

MATERIALS AND METHODS: A total of 35 eligible patients, with an average age of 42 years and a male/female ratio of 1:4, underwent ZTE MRI and images are processed to generate iZTE and pCT images were generated through a series of steps including intensity equalization, thresholding, and deep learning-based pCT generation. These images were then compared to CT scans using a rating scale; inter-rater kappa coefficient evaluated observer consensus while statistical metrics like sensitivity and specificity assessed their performance in capturing bone-related characteristics.

RESULTS: The study revealed excellent interobserver agreement for lesion assessment using both pCT and iZTE imaging modalities, with kappa coefficient of 0.91 (P < 0.0001) and 0.92 respectively (P < 0.0001). Also, pCT and iZTE accurately predicted various lesion characteristics with sensitivity ranging from 84.3% to 95.1% and 82.6%-94.2% (95% CI) with a diagnostic accuracy of 95.56% and 94.44% respectively. Although both of them encountered challenges with ground glassing, hyperostosis, and intralesional bony fragments, they showed good performance in other bony lesion assessments.

CONCLUSIONS: The pilot study suggests strong potential for integrating the ZTE imaging into standard care for skull base and calvarial bony lesions assessment. Additionally, larger-scale studies are needed for comprehensive assessment of its efficacy.

PMID:39322533 | DOI:10.1016/j.crad.2024.08.029

Surgery. 2024 Sep 24:S0039-6060(24)00670-6. doi: 10.1016/j.surg.2024.06.059. Online ahead of print.

ABSTRACT

OBJECTIVE: We sought to identify patients at risk of "futile" surgery for intrahepatic cholangiocarcinoma using an artificial intelligence (AI)-based model based on preoperative variables.

METHODS: Intrahepatic cholangiocarcinoma patients who underwent resection between 1990 and 2020 were identified from a multi-institutional database. Futility was defined either as mortality or recurrence within 12 months of surgery. Various machine learning and deep learning techniques were used to develop prediction models for futile surgery.

RESULTS: Overall, 827 intrahepatic cholangiocarcinoma patients were included. Among 378 patients (45.7%) who had futile surgery, 297 patients (78.6%) developed intrahepatic cholangiocarcinoma recurrence and 81 patients (21.4%) died within 12 months of surgical resection. An ensemble model consisting of multilayer perceptron and gradient boosting classifiers that used 10 preoperative factors demonstrated the highest accuracy, with areas under receiver operating characteristic curves of 0.830 (95% confidence interval 0.798-0.861) and 0.781 (95% confidence interval 0.707-0.853) in the training and testing cohorts, respectively. The model displayed sensitivity and specificity of 64.5% and 80.0%, respectively, with positive and negative predictive values of 73.1% and 72.7%, respectively. Radiologic tumor burden score, serum carbohydrate antigen 19-9, and direct bilirubin levels were the factors most strongly predictive of futile surgery. The artificial intelligence-based model was made available online for ease of use and clinical applicability (https://altaf-pawlik-icc-futilityofsurgery-calculator.streamlit.app/).

CONCLUSION: The artificial intelligence ensemble model demonstrated high accuracy to identify patients preoperatively at high risk of undergoing futile surgery for intrahepatic cholangiocarcinoma. Artificial intelligence-based prediction models can provide clinicians with reliable preoperative guidance and aid in avoiding futile surgical procedures that are unlikely to provide patients long-term benefits.

PMID:39322483 | DOI:10.1016/j.surg.2024.06.059