JAMA. 2024 Aug 1. doi: 10.1001/jama.2024.10770. Online ahead of print.

ABSTRACT

IMPORTANCE: Accurate assessment of gestational age (GA) is essential to good pregnancy care but often requires ultrasonography, which may not be available in low-resource settings. This study developed a deep learning artificial intelligence (AI) model to estimate GA from blind ultrasonography sweeps and incorporated it into the software of a low-cost, battery-powered device.

OBJECTIVE: To evaluate GA estimation accuracy of an AI-enabled ultrasonography tool when used by novice users with no prior training in sonography.

DESIGN, SETTING, AND PARTICIPANTS: This prospective diagnostic accuracy study enrolled 400 individuals with viable, single, nonanomalous, first-trimester pregnancies in Lusaka, Zambia, and Chapel Hill, North Carolina. Credentialed sonographers established the "ground truth" GA via transvaginal crown-rump length measurement. At random follow-up visits throughout gestation, including a primary evaluation window from 14 0/7 weeks' to 27 6/7 weeks' gestation, novice users obtained blind sweeps of the maternal abdomen using the AI-enabled device (index test) and credentialed sonographers performed fetal biometry with a high-specification machine (study standard).

MAIN OUTCOMES AND MEASURES: The primary outcome was the mean absolute error (MAE) of the index test and study standard, which was calculated by comparing each method's estimate to the previously established GA and considered equivalent if the difference fell within a prespecified margin of ±2 days.

RESULTS: In the primary evaluation window, the AI-enabled device met criteria for equivalence to the study standard, with an MAE (SE) of 3.2 (0.1) days vs 3.0 (0.1) days (difference, 0.2 days [95% CI, -0.1 to 0.5]). Additionally, the percentage of assessments within 7 days of the ground truth GA was comparable (90.7% for the index test vs 92.5% for the study standard). Performance was consistent in prespecified subgroups, including the Zambia and North Carolina cohorts and those with high body mass index.

CONCLUSIONS AND RELEVANCE: Between 14 and 27 weeks' gestation, novice users with no prior training in ultrasonography estimated GA as accurately with the low-cost, point-of-care AI tool as credentialed sonographers performing standard biometry on high-specification machines. These findings have immediate implications for obstetrical care in low-resource settings, advancing the World Health Organization goal of ultrasonography estimation of GA for all pregnant people.

TRIAL REGISTRATION: ClinicalTrials.gov Identifier: NCT05433519.

PMID:39088200 | DOI:10.1001/jama.2024.10770

Neurosurg Rev. 2024 Aug 1;47(1):391. doi: 10.1007/s10143-024-02636-1.

ABSTRACT

Cerebral aneurysms, affecting 2-5% of the global population, are often asymptomatic and commonly located within the Circle of Willis. A recent study in Neurosurgical Review highlights a significant reduction in the annual rupture rates of unruptured cerebral aneurysms (UCAs) in Japan from 2003 to 2018. By analyzing age-adjusted mortality rates of subarachnoid hemorrhage (SAH) and the number of treated ruptured cerebral aneurysms (RCAs), researchers found a substantial decrease in rupture rates-from 1.44 to 0.87% and from 0.92 to 0.76%, respectively (p < 0.001). This 88% reduction was largely attributed to improved hypertension management. Recent advancements in artificial intelligence (AI) and machine learning (ML) further support these findings. The RAPID Aneurysm software demonstrated high accuracy in detecting cerebral aneurysms on CT Angiography (CTA), while ML algorithms showed promise in predicting aneurysm rupture risk. A meta-analysis indicated that ML models could achieve 83% sensitivity and specificity in rupture prediction. Additionally, deep learning techniques, such as the PointNet + + architecture, achieved an AUC of 0.85 in rupture risk prediction. These technological advancements in AI and ML are poised to enhance early detection and risk management, potentially contributing to the observed reduction in UCA rupture rates and improving patient outcomes.

PMID:39088154 | DOI:10.1007/s10143-024-02636-1

Med Mol Morphol. 2024 Aug 1. doi: 10.1007/s00795-024-00399-8. Online ahead of print.

ABSTRACT

The aim of this study is to establish a deep learning (DL) model to predict the pathological type of gastric adenocarcinoma cancer based on whole-slide images(WSIs). We downloaded 356 histopathological images of gastric adenocarcinoma (STAD) patients from The Cancer Genome Atlas database and randomly divided them into the training set, validation set and test set (8:1:1). Additionally, 80 H&E-stained WSIs of STAD were collected for external validation. The CLAM tool was used to cut the WSIs and further construct the model by DL algorithm, achieving an accuracy of over 90% in identifying and predicting histopathological subtypes. External validation results demonstrated the model had a certain generalization ability. Moreover, DL features were extracted from the model to further investigate the differences in immune infiltration and patient prognosis between the two subtypes. The DL model can accurately predict the pathological classification of STAD patients, and provide certain reference value for clinical diagnosis. The nomogram combining DL-signature, gene-signature and clinical features can be used as a prognostic classifier for clinical decision-making and treatment.

PMID:39088070 | DOI:10.1007/s00795-024-00399-8

Appl Neuropsychol Adult. 2024 Aug 1:1-15. doi: 10.1080/23279095.2024.2382823. Online ahead of print.

ABSTRACT

The cognitive impairment known as dementia affects millions of individuals throughout the globe. The use of machine learning (ML) and deep learning (DL) algorithms has shown great promise as a means of early identification and treatment of dementia. Dementias such as Alzheimer's Dementia, frontotemporal dementia, Lewy body dementia, and vascular dementia are all discussed in this article, along with a literature review on using ML algorithms in their diagnosis. Different ML algorithms, such as support vector machines, artificial neural networks, decision trees, and random forests, are compared and contrasted, along with their benefits and drawbacks. As discussed in this article, accurate ML models may be achieved by carefully considering feature selection and data preparation. We also discuss how ML algorithms can predict disease progression and patient responses to therapy. However, overreliance on ML and DL technologies should be avoided without further proof. It's important to note that these technologies are meant to assist in diagnosis but should not be used as the sole criteria for a final diagnosis. The research implies that ML algorithms may help increase the precision with which dementia is diagnosed, especially in its early stages. The efficacy of ML and DL algorithms in clinical contexts must be verified, and ethical issues around the use of personal data must be addressed, but this requires more study.

PMID:39087520 | DOI:10.1080/23279095.2024.2382823

J Clin Pediatr Dent. 2024 Jul;48(4):191-199. doi: 10.22514/jocpd.2024.093. Epub 2024 Jul 3.

ABSTRACT

Bone age determination in individuals is important for the diagnosis and treatment of growing children. This study aimed to develop a deep-learning model for bone age estimation using lateral cephalometric radiographs (LCRs) and regions of interest (ROIs) in growing children and evaluate its performance. This retrospective study included 1050 patients aged 4-18 years who underwent LCR and hand-wrist radiography on the same day at Pusan National University Dental Hospital and Ulsan University Hospital between January 2014 and June 2023. Two pretrained convolutional neural networks, InceptionResNet-v2 and NasNet-Large, were employed to develop a deep-learning model for bone age estimation. The LCRs and ROIs, which were designated as the cervical vertebrae areas, were labeled according to the patient's bone age. Bone age was collected from the same patient's hand-wrist radiograph. Deep-learning models trained with five-fold cross-validation were tested using internal and external validations. The LCR-trained model outperformed the ROI-trained models. In addition, visualization of each deep learning model using the gradient-weighted regression activation mapping technique revealed a difference in focus in bone age estimation. The findings of this comparative study are significant because they demonstrate the feasibility of bone age estimation via deep learning with craniofacial bones and dentition, in addition to the cervical vertebrae on the LCR of growing children.

PMID:39087230 | DOI:10.22514/jocpd.2024.093

J Med Imaging (Bellingham). 2024 Jul;11(4):047501. doi: 10.1117/1.JMI.11.4.047501. Epub 2024 Jul 30.

ABSTRACT

PURPOSE: Endometrial cancer (EC) is one of the most common types of cancer affecting women. While the hematoxylin-and-eosin (H&E) staining remains the standard for histological analysis, the immunohistochemistry (IHC) method provides molecular-level visualizations. Our study proposes a digital staining method to generate the hematoxylin-3,3'-diaminobenzidine (H-DAB) IHC stain of Ki-67 for the whole slide image of the EC tumor from its H&E stain counterpart.

APPROACH: We employed a color unmixing technique to yield stain density maps from the optical density (OD) of the stains and utilized the U-Net for end-to-end inference. The effectiveness of the proposed method was evaluated using the Pearson correlation between the digital and physical stain's labeling index (LI), a key metric indicating tumor proliferation. Two different cross-validation schemes were designed in our study: intraslide validation and cross-case validation (CCV). In the widely used intraslide scheme, the training and validation sets might include different regions from the same slide. The rigorous CCV validation scheme strictly prohibited any validation slide from contributing to training.

RESULTS: The proposed method yielded a high-resolution digital stain with preserved histological features, indicating a reliable correlation with the physical stain in terms of the Ki-67 LI. In the intraslide scheme, using intraslide patches resulted in a biased accuracy (e.g., R = 0.98 ) significantly higher than that of CCV. The CCV scheme retained a fair correlation (e.g., R = 0.66 ) between the LIs calculated from the digital stain and its physical IHC counterpart. Inferring the OD of the IHC stain from that of the H&E stain enhanced the correlation metric, outperforming that of the baseline model using the RGB space.

CONCLUSIONS: Our study revealed that molecule-level insights could be obtained from H&E images using deep learning. Furthermore, the improvement brought via OD inference indicated a possible method for creating more generalizable models for digital staining via per-stain analysis.

PMID:39087085 | PMC:PMC11287056 | DOI:10.1117/1.JMI.11.4.047501

J Med Imaging (Bellingham). 2024 Jul;11(4):044504. doi: 10.1117/1.JMI.11.4.044504. Epub 2024 Jul 30.

ABSTRACT

PURPOSE: Analyzing the anatomy of the aorta and left ventricular outflow tract (LVOT) is crucial for risk assessment and planning of transcatheter aortic valve implantation (TAVI). A comprehensive analysis of the aortic root and LVOT requires the extraction of the patient-individual anatomy via segmentation. Deep learning has shown good performance on various segmentation tasks. If this is formulated as a supervised problem, large amounts of annotated data are required for training. Therefore, minimizing the annotation complexity is desirable.

APPROACH: We propose two-dimensional (2D) cross-sectional annotation and point cloud-based surface reconstruction to train a fully automatic 3D segmentation network for the aortic root and the LVOT. Our sparse annotation scheme enables easy and fast training data generation for tubular structures such as the aortic root. From the segmentation results, we derive clinically relevant parameters for TAVI planning.

RESULTS: The proposed 2D cross-sectional annotation results in high inter-observer agreement [Dice similarity coefficient (DSC): 0.94]. The segmentation model achieves a DSC of 0.90 and an average surface distance of 0.96 mm. Our approach achieves an aortic annulus maximum diameter difference between prediction and annotation of 0.45 mm (inter-observer variance: 0.25 mm).

CONCLUSIONS: The presented approach facilitates reproducible annotations. The annotations allow for training accurate segmentation models of the aortic root and LVOT. The segmentation results facilitate reproducible and quantifiable measurements for TAVI planning.

PMID:39087084 | PMC:PMC11287057 | DOI:10.1117/1.JMI.11.4.044504

R Soc Open Sci. 2024 Jul 31;11(7):240228. doi: 10.1098/rsos.240228. eCollection 2024 Jul.

ABSTRACT

Finding the optimal treatment strategy to accelerate wound healing is of utmost importance, but it presents a formidable challenge owing to the intrinsic nonlinear nature of the process. We propose an adaptive closed-loop control framework that incorporates deep learning, optimal control and reinforcement learning to accelerate wound healing. By adaptively learning a linear representation of nonlinear wound healing dynamics using deep learning and interactively training a deep reinforcement learning agent for tracking the optimal signal derived from this representation without the need for intricate mathematical modelling, our approach has not only successfully reduced the wound healing time by 45.56% compared to the one without any treatment, but also demonstrates the advantages of offering a safer and more economical treatment strategy. The proposed methodology showcases a significant potential for expediting wound healing by effectively integrating perception, predictive modelling and optimal adaptive control, eliminating the need for intricate mathematical models.

PMID:39086835 | PMC:PMC11289634 | DOI:10.1098/rsos.240228

Front Vet Sci. 2024 Jul 17;11:1436795. doi: 10.3389/fvets.2024.1436795. eCollection 2024.

ABSTRACT

Facial expressions are essential for communication and emotional expression across species. Despite the improvements brought by tools like the Horse Grimace Scale (HGS) in pain recognition in horses, their reliance on human identification of characteristic traits presents drawbacks such as subjectivity, training requirements, costs, and potential bias. Despite these challenges, the development of facial expression pain scales for animals has been making strides. To address these limitations, Automated Pain Recognition (APR) powered by Artificial Intelligence (AI) offers a promising advancement. Notably, computer vision and machine learning have revolutionized our approach to identifying and addressing pain in non-verbal patients, including animals, with profound implications for both veterinary medicine and animal welfare. By leveraging the capabilities of AI algorithms, we can construct sophisticated models capable of analyzing diverse data inputs, encompassing not only facial expressions but also body language, vocalizations, and physiological signals, to provide precise and objective evaluations of an animal's pain levels. While the advancement of APR holds great promise for improving animal welfare by enabling better pain management, it also brings forth the need to overcome data limitations, ensure ethical practices, and develop robust ground truth measures. This narrative review aimed to provide a comprehensive overview, tracing the journey from the initial application of facial expression recognition for the development of pain scales in animals to the recent application, evolution, and limitations of APR, thereby contributing to understanding this rapidly evolving field.

PMID:39086767 | PMC:PMC11288915 | DOI:10.3389/fvets.2024.1436795

Arch Med Sci Atheroscler Dis. 2024 Jun 25;9:e122-e128. doi: 10.5114/amsad/183420. eCollection 2024.

ABSTRACT

Artificial intelligence is growing quickly, and its application in the global diabetes pandemic has the potential to completely change the way this chronic illness is identified and treated. Machine learning methods have been used to construct algorithms supporting predictive models for the risk of getting diabetes or its complications. Social media and Internet forums also increase patient participation in diabetes care. Diabetes resource usage optimisation has benefited from technological improvements. As a lifestyle therapy intervention, digital therapies have made a name for themselves in the treatment of diabetes. Artificial intelligence will cause a paradigm shift in diabetes care, moving away from current methods and toward the creation of focused, data-driven precision treatment.

PMID:39086621 | PMC:PMC11289240 | DOI:10.5114/amsad/183420

Appl Sci (Basel). 2023 Jul 2;13(14):8555. doi: 10.3390/app13148555. Epub 2023 Jul 24.

ABSTRACT

Stargardt disease is the most common form of juvenile-onset macular dystrophy. Spectral-domain optical coherence tomography (SD-OCT) imaging provides an opportunity to directly measure changes to retinal layers due to Stargardt atrophy. Generally, atrophy segmentation and prediction can be conducted using mean intensity feature maps generated from the relevant retinal layers. In this paper, we report an approach using advanced OCT-derived features to augment and enhance data beyond the commonly used mean intensity features for enhanced prediction of Stargardt atrophy with an ensemble deep learning neural network. With all the relevant retinal layers, this neural network architecture achieves a median Dice coefficient of 0.830 for six-month predictions and 0.828 for twelve-month predictions, showing a significant improvement over a neural network using only mean intensity, which achieved Dice coefficients of 0.744 and 0.762 for six-month and twelve-month predictions, respectively. When using feature maps generated from different layers of the retina, significant differences in performance were observed. This study shows promising results for using multiple OCT-derived features beyond intensity for assessing the prognosis of Stargardt disease and quantifying the rate of progression.

PMID:39086558 | PMC:PMC11288976 | DOI:10.3390/app13148555

Front Toxicol. 2024 Jul 17;6:1401036. doi: 10.3389/ftox.2024.1401036. eCollection 2024.

ABSTRACT

The cell painting (CP) assay has emerged as a potent imaging-based high-throughput phenotypic profiling (HTPP) tool that provides comprehensive input data for in silico prediction of compound activities and potential hazards in drug discovery and toxicology. CP enables the rapid, multiplexed investigation of various molecular mechanisms for thousands of compounds at the single-cell level. The resulting large volumes of image data provide great opportunities but also pose challenges to image and data analysis routines as well as property prediction models. This review addresses the integration of CP-based phenotypic data together with or in substitute of structural information from compounds into machine (ML) and deep learning (DL) models to predict compound activities for various human-relevant disease endpoints and to identify the underlying modes-of-action (MoA) while avoiding unnecessary animal testing. The successful application of CP in combination with powerful ML/DL models promises further advances in understanding compound responses of cells guiding therapeutic development and risk assessment. Therefore, this review highlights the importance of unlocking the potential of CP assays when combined with molecular fingerprints for compound evaluation and discusses the current challenges that are associated with this approach.

PMID:39086553 | PMC:PMC11288911 | DOI:10.3389/ftox.2024.1401036

Magn Reson Med. 2024 Jul 31. doi: 10.1002/mrm.30241. Online ahead of print.

ABSTRACT

PURPOSE: To evaluate the influence of the confounding factors, direct water saturation (DWS), and magnetization transfer contrast (MTC) effects on measured Z-spectra and amide proton transfer (APT) contrast in brain tumors.

METHODS: High-grade glioma patients were scanned using an RF saturation-encoded 3D MR fingerprinting (MRF) sequence at 3 T. For MRF reconstruction, a recurrent neural network was designed to learn free water and semisolid macromolecule parameter mappings of the underlying multiple tissue properties from saturation-transfer MRF signals. The DWS spectra and MTC spectra were synthesized by solving Bloch-McConnell equations and evaluated in brain tumors.

RESULTS: The dominant contribution to the saturation effect at 3.5 ppm was from DWS and MTC effects, but 25%-33% of the saturated signal in the gadolinium-enhancing tumor (13%-20% for normal tissue) was due to the APT effect. The APT# signal of the gadolinium-enhancing tumor was significantly higher than that of the normal-appearing white matter (10.1% vs. 8.3% at 1 μT and 11.2% vs. 7.8% at 1.5 μT).

CONCLUSION: The RF saturation-encoded MRF allowed us to separate contributions to the saturation signal at 3.5 ppm in the Z-spectrum. Although free water and semisolid MTC are the main contributors, significant APT contrast between tumor and normal tissues was observed.

PMID:39086185 | DOI:10.1002/mrm.30241

Psychiatry Investig. 2024 Aug 2. doi: 10.30773/pi.2024.0157. Online ahead of print.

ABSTRACT

OBJECTIVE: This study aimed to use deep learning (DL) to develop a cost-effective and accessible screening tool to improve the detection of cognitive decline, a precursor of Alzheimer's disease (AD). This study integrating a comprehensive battery of neuropsychological tests adjusted for individual demographic variables such as age, sex, and education level.

METHODS: A total of 2,863 subjects with subjective cognitive complaints who underwent a comprehensive neuropsychological assessment were included. A random forest classifier was used to discern the most predictive test combinations to distinguish between dementia and nondementia cases. The model was trained and validated on this dataset, focusing on feature importance to determine the cognitive tests that were most indicative of decline.

RESULTS: Subjects had a mean age of 72.68 years and an average education level of 7.62 years. The DL model achieved an accuracy of 82.42% and an area under the curve of 0.816, effectively classifying dementia. Feature importance analysis identified significant tests across cognitive domains: attention was gauged by the Trail Making Test Part B, language by the Boston Naming Test, memory by the Rey Complex Figure Test delayed recall, visuospatial skills by the Rey Complex Figure Test copy score, and frontal function by the Stroop Test Word reading time.

CONCLUSION: This study showed the potential of DL to improve AD diagnostics, suggesting that a wide range of cognitive assessments could yield a more accurate diagnosis than traditional methods. This research establishes a foundation for future broader studies, which could substantiate the approach and further refine the screening tool.

PMID:39086161 | DOI:10.30773/pi.2024.0157

Eye Vis (Lond). 2024 Aug 1;11(1):27. doi: 10.1186/s40662-024-00396-z.

ABSTRACT

BACKGROUND: Acute retinal necrosis (ARN) is a relatively rare but highly damaging and potentially sight-threatening type of uveitis caused by infection with the human herpesvirus. Without timely diagnosis and appropriate treatment, ARN can lead to severe vision loss. We aimed to develop a deep learning framework to distinguish ARN from other types of intermediate, posterior, and panuveitis using ultra-widefield color fundus photography (UWFCFP).

METHODS: We conducted a two-center retrospective discovery and validation study to develop and validate a deep learning model called DeepDrARN for automatic uveitis detection and differentiation of ARN from other uveitis types using 11,508 UWFCFPs from 1,112 participants. Model performance was evaluated with the area under the receiver operating characteristic curve (AUROC), the area under the precision and recall curves (AUPR), sensitivity and specificity, and compared with seven ophthalmologists.

RESULTS: DeepDrARN for uveitis screening achieved an AUROC of 0.996 (95% CI: 0.994-0.999) in the internal validation cohort and demonstrated good generalizability with an AUROC of 0.973 (95% CI: 0.956-0.990) in the external validation cohort. DeepDrARN also demonstrated excellent predictive ability in distinguishing ARN from other types of uveitis with AUROCs of 0.960 (95% CI: 0.943-0.977) and 0.971 (95% CI: 0.956-0.986) in the internal and external validation cohorts. DeepDrARN was also tested in the differentiation of ARN, non-ARN uveitis (NAU) and normal subjects, with sensitivities of 88.9% and 78.7% and specificities of 93.8% and 89.1% in the internal and external validation cohorts, respectively. The performance of DeepDrARN is comparable to that of ophthalmologists and even exceeds the average accuracy of seven ophthalmologists, showing an improvement of 6.57% in uveitis screening and 11.14% in ARN identification.

CONCLUSIONS: Our study demonstrates the feasibility of deep learning algorithms in enabling early detection, reducing treatment delays, and improving outcomes for ARN patients.

PMID:39085922 | DOI:10.1186/s40662-024-00396-z

Biomed Eng Online. 2024 Jul 31;23(1):76. doi: 10.1186/s12938-024-01265-5.

ABSTRACT

BACKGROUND: Transcranial sonography (TCS) plays a crucial role in diagnosing Parkinson's disease. However, the intricate nature of TCS pathological features, the lack of consistent diagnostic criteria, and the dependence on physicians' expertise can hinder accurate diagnosis. Current TCS-based diagnostic methods, which rely on machine learning, often involve complex feature engineering and may struggle to capture deep image features. While deep learning offers advantages in image processing, it has not been tailored to address specific TCS and movement disorder considerations. Consequently, there is a scarcity of research on deep learning algorithms for TCS-based PD diagnosis.

METHODS: This study introduces a deep learning residual network model, augmented with attention mechanisms and multi-scale feature extraction, termed AMSNet, to assist in accurate diagnosis. Initially, a multi-scale feature extraction module is implemented to robustly handle the irregular morphological features and significant area information present in TCS images. This module effectively mitigates the effects of artifacts and noise. When combined with a convolutional attention module, it enhances the model's ability to learn features of lesion areas. Subsequently, a residual network architecture, integrated with channel attention, is utilized to capture hierarchical and detailed textures within the images, further enhancing the model's feature representation capabilities.

RESULTS: The study compiled TCS images and personal data from 1109 participants. Experiments conducted on this dataset demonstrated that AMSNet achieved remarkable classification accuracy (92.79%), precision (95.42%), and specificity (93.1%). It surpassed the performance of previously employed machine learning algorithms in this domain, as well as current general-purpose deep learning models.

CONCLUSION: The AMSNet proposed in this study deviates from traditional machine learning approaches that necessitate intricate feature engineering. It is capable of automatically extracting and learning deep pathological features, and has the capacity to comprehend and articulate complex data. This underscores the substantial potential of deep learning methods in the application of TCS images for the diagnosis of movement disorders.

PMID:39085884 | DOI:10.1186/s12938-024-01265-5

Radiat Oncol. 2024 Jul 31;19(1):98. doi: 10.1186/s13014-024-02496-5.

ABSTRACT

BACKGROUND: The purpose of this study was to improve the deep learning (DL) model performance in predicting and classifying IMRT gamma passing rate (GPR) by using input features related to machine parameters and a class balancing technique.

METHODS: A total of 2348 fields from 204 IMRT plans for patients with nasopharyngeal carcinoma were retrospectively collected to form a dataset. Input feature maps, including fluence, leaf gap, leaf speed of both banks, and corresponding errors, were constructed from the dynamic log files. The SHAP framework was employed to compute the impact of each feature on the model output for recursive feature elimination. A series of UNet++ based models were trained on the obtained eight feature sets with three fine-tuning methods including the standard mean squared error (MSE) loss, a re-sampling technique, and a proposed weighted MSE loss (WMSE). Differences in mean absolute error, area under the receiver operating characteristic curve (AUC), sensitivity, and specificity were compared between the different models.

RESULTS: The models trained with feature sets including leaf speed and leaf gap features predicted GPR for failed fields more accurately than the other models (F(7, 147) = 5.378, p < 0.001). The WMSE loss had the highest accuracy in predicting GPR for failed fields among the three fine-tuning methods (F(2, 42) = 14.149, p < 0.001), while an opposite trend was observed in predicting GPR for passed fields (F(2, 730) = 9.907, p < 0.001). The WMSE_FS5 model achieved a superior AUC (0.92) and more balanced sensitivity (0.77) and specificity (0.89) compared to the other models.

CONCLUSIONS: Machine parameters can provide discriminative input features for GPR prediction in DL. The novel weighted loss function demonstrates the ability to balance the prediction and classification accuracy between the passed and failed fields. The proposed approach is able to improve the DL model performance in predicting and classifying GPR, and can potentially be integrated into the plan optimization process to generate higher deliverability plans.

TRIAL REGISTRATION: This clinical trial was registered in the Chinese Clinical Trial Registry on March 26th, 2020 (registration number: ChiCTR2000031276). https://clinicaltrials.gov/ct2/show/ChiCTR2000031276.

PMID:39085872 | DOI:10.1186/s13014-024-02496-5

Adv Ther. 2024 Jul 31. doi: 10.1007/s12325-024-02944-4. Online ahead of print.

ABSTRACT

INTRODUCTION: Automated bone age assessment (BAA) is of growing interest because of its accuracy and time efficiency in daily practice. In this study, we validated the clinical applicability of a commercially available artificial intelligence (AI)-powered X-ray bone age analyzer equipped with a deep learning-based automated BAA system and compared its performance with that of the Tanner-Whitehouse 3 (TW-3) method.

METHODS: Radiographs prospectively collected from 30 centers across various regions in China, including 900 Chinese children and adolescents, were assessed independently by six doctors (three experts and three residents) and an AI analyzer for TW3 radius, ulna, and short bones (RUS) and TW3 carpal bone age. The experts' mean estimates were accepted as the gold standard. The performance of the AI analyzer was compared with that of each resident.

RESULTS: For the estimation of TW3-RUS, the AI analyzer had a mean absolute error (MAE) of 0.48 ± 0.42. The percentage of patients with an absolute error of < 1.0 years was 86.78%. The MAE was significantly lower than that of rater 1 (0.54 ± 0.49, P = 0.0068); however, it was not significant for rater 2 (0.48 ± 0.48) or rater 3 (0.49 ± 0.46). For TW3 carpal, the AI analyzer had an MAE of 0.48 ± 0.65. The percentage of patients with an absolute error of < 1.0 years was 88.78%. The MAE was significantly lower than that of rater 2 (0.58 ± 0.67, P = 0.0018) and numerically lower for rater 1 (0.54 ± 0.64) and rater 3 (0.50 ± 0.53). These results were consistent for the subgroups according to sex, and differences between the age groups were observed.

CONCLUSION: In this comprehensive validation study conducted in China, an AI-powered X-ray bone age analyzer showed accuracies that matched or exceeded those of doctor raters. This method may improve the efficiency of clinical routines by reducing reading time without compromising accuracy.

PMID:39085749 | DOI:10.1007/s12325-024-02944-4

Int J Comput Assist Radiol Surg. 2024 Jul 31. doi: 10.1007/s11548-024-03240-w. Online ahead of print.

ABSTRACT

PURPOSE: This study addressed the challenge of detecting and classifying the severity of ductopenia in parotid glands, a structural abnormality characterized by a reduced number of salivary ducts, previously shown to be associated with salivary gland impairment. The aim of the study was to develop an automatic algorithm designed to improve diagnostic accuracy and efficiency in analyzing ductopenic parotid glands using sialo cone-beam CT (sialo-CBCT) images.

METHODS: We developed an end-to-end automatic pipeline consisting of three main steps: (1) region of interest (ROI) computation, (2) parotid gland segmentation using the Frangi filter, and (3) ductopenia case classification with a residual neural network (RNN) augmented by multidirectional maximum intensity projection (MIP) images. To explore the impact of the first two steps, the RNN was trained on three datasets: (1) original MIP images, (2) MIP images with predefined ROIs, and (3) MIP images after segmentation.

RESULTS: Evaluation was conducted on 126 parotid sialo-CBCT scans of normal, moderate, and severe ductopenic cases, yielding a high performance of 100% for the ROI computation and 89% for the gland segmentation. Improvements in accuracy and F1 score were noted among the original MIP images (accuracy: 0.73, F1 score: 0.53), ROI-predefined images (accuracy: 0.78, F1 score: 0.56), and segmented images (accuracy: 0.95, F1 score: 0.90). Notably, ductopenic detection sensitivity was 0.99 in the segmented dataset, highlighting the capabilities of the algorithm in detecting ductopenic cases.

CONCLUSIONS: Our method, which combines classical image processing and deep learning techniques, offers a promising solution for automatic detection of parotid glands ductopenia in sialo-CBCT scans. This may be used for further research aimed at understanding the role of presence and severity of ductopenia in salivary gland dysfunction.

PMID:39085681 | DOI:10.1007/s11548-024-03240-w

Discov Oncol. 2024 Jul 31;15(1):323. doi: 10.1007/s12672-024-01177-9.

ABSTRACT

PURPOSE OBJECTIVE(S): Manual contouring of the prostate region in planning computed tomography (CT) images is a challenging task due to factors such as low contrast in soft tissues, inter- and intra-observer variability, and variations in organ size and shape. Consequently, the use of automated contouring methods can offer significant advantages. In this study, we aimed to investigate automated male pelvic multi-organ contouring in multi-center planning CT images using a hybrid convolutional neural network-vision transformer (CNN-ViT) that combines convolutional and ViT techniques.

MATERIALS/METHODS: We used retrospective data from 104 localized prostate cancer patients, with delineations of the clinical target volume (CTV) and critical organs at risk (OAR) for external beam radiotherapy. We introduced a novel attention-based fusion module that merges detailed features extracted through convolution with the global features obtained through the ViT.

RESULTS: The average dice similarity coefficients (DSCs) achieved by VGG16-UNet-ViT for the prostate, bladder, rectum, right femoral head (RFH), and left femoral head (LFH) were 91.75%, 95.32%, 87.00%, 96.30%, and 96.34%, respectively. Experiments conducted on multi-center planning CT images indicate that combining the ViT structure with the CNN network resulted in superior performance for all organs compared to pure CNN and transformer architectures. Furthermore, the proposed method achieves more precise contours compared to state-of-the-art techniques.

CONCLUSION: Results demonstrate that integrating ViT into CNN architectures significantly improves segmentation performance. These results show promise as a reliable and efficient tool to facilitate prostate radiotherapy treatment planning.

PMID:39085488 | DOI:10.1007/s12672-024-01177-9