Efficacy of fusion imaging for immediate post-ablation assessment of malignant liver neoplasms: A systematic review

Glo-net: A dual task branch based neural network for multi-class glomeruli segmentation

Accurate segmentation and classification of glomeruli are fundamental to histopathology slide analysis in renal pathology, which helps to characterize individual kidney disease. Accurate segmentation of glomeruli of different types faces two main challenges compared to traditional primitives segmentation in computational image analysis. Limited by small kernel size, traditional convolutional neural networks could hardly understand the complete context information of different glomeruli. Moreover, typical semantic segmentation networks lack adequate attention to difficult glomerular samples during the training process due to serious class imbalance between different glomeruli types. We propose a new deep learning approach, Glo-Net, which accurately segments and classifies glomeruli based on digitized pathology slides. Specifically, Glo-Net divides the traditional semantic segmentation network into two branches, i.e., segmentation and classification. While the segmentation branch specifically aims at localizing and delineating the boundary of individual glomerulus, the classification branch could focus on differentiating the glomerular types based on segmented pixels. In addition, an innovative loss function is added to the classification task to compensate for the class imbalance and minor types of glomeruli. The proposed network's average accuracy and F-score in classification tasks on the multi-institution datasets (including an external validation set) are 0.858 and 0.704, respectively. The average intersection over union (IoU) in segmentation tasks is 0.866. The Glo-Net demonstrates a 5 % improvement in classification accuracy, with up to 14 % increases for minor classes and an average 6 % IoU increase for segmentation tasks. Quantitative results show that our network achieves overall higher accuracy for segmentation and classification among nine subtypes of glomeruli compared to previous work with improved robustness and generalizability.

Interpretable prediction of drug-drug interactions via text embedding in biomedical literature

Polypharmacy is a promising approach for treating diseases, especially those with complex symptoms. However, it can lead to unexpected drug-drug interactions (DDIs), potentially reducing efficacy and triggering adverse drug reactions (ADRs). Predicting the risk of DDIs is crucial for ensuring safe drug use, particularly by identifying the types of DDIs and the mechanisms involved. Therefore, this study used biomedical literature to proposed hierarchical attention-based deep learning models to predict DDIs and their types. The proposed model consists of two components: drug embedding and DDI prediction. The drug embedding module extracts representation vectors that effectively capture drug properties using sentence and sequence embedding methods. For sentence embedding, a pre-trained biomedical language model is used to map drug-related sentences into vector space. For sequence embedding, sentence embedding vectors are sequentially fed into bidirectional long short-term memory with a hierarchical attention network, enabling the analysis of sentences relevant to DDI prediction while accounting for the order of the sentences. Finally, DDI prediction is performed using a deep neural network based on the sequence embedding vectors of a drug pair. Our model achieved high performances in the accuracy (0.85-0.90), AUROC (0.98-0.99), and AUPR (0.63-0.95) performance across 164 DDI types. Additionally, the proposed model showed improvements in up to 11 % in AUROC, and 8 % in AUPR. Furthermore, model interprets predictions by leveraging attention mechanisms and drug similarity. The results indicated that the model considered various factors beyond similarity to predict DDIs. These findings may help prevent unforeseen medical accidents and reduce healthcare costs by predicting detailed drug interaction types.

Automatic image generation and stage prediction of breast cancer immunobiological through a proposed IHC-GAN model

Invasive breast cancer diagnosis and treatment planning require an accurate assessment of human epidermal growth factor receptor 2 (HER2) expression levels. While immunohistochemical techniques (IHC) are the gold standard for HER2 evaluation, their implementation can be resource-intensive and costly. To reduce these obstacles and expedite the procedure, we present an efficient deep-learning model that generates high-quality IHC-stained images directly from Hematoxylin and Eosin (H&E) stained images. We propose a new IHC-GAN that enhances the Pix2PixHD model into a dual generator module, improving its performance and simplifying its structure. Furthermore, to strengthen feature extraction for HE-stained image classification, we integrate MobileNetV3 as the backbone network. The extracted features are then merged with those generated by the generator to improve overall performance. Moreover, the decoder's performance is enhanced by providing the related features from the classified labels by incorporating the adaptive instance normalization technique. The proposed IHC-GAN was trained and validated on a comprehensive dataset comprising 4,870 registered image pairs, encompassing a spectrum of HER2 expression levels. Our findings demonstrate promising results in translating H&E images to IHC-equivalent representations, offering a potential solution to reduce the costs associated with traditional HER2 assessment methods. We extensively validate our model and the current dataset. We compare it with state-of-the-art techniques, achieving high performance using different evaluation metrics, showing 0.0927 FID, 22.87 PSNR, and 0.3735 SSIM. The proposed approach exhibits significant enhancements over current GAN models, including an 88% reduction in Frechet Inception Distance (FID), a 4% enhancement in Learned Perceptual Image Patch Similarity (LPIPS), a 10% increase in Peak Signal-to-Noise Ratio (PSNR), and a 45% reduction in Mean Squared Error (MSE). This advancement holds significant potential for enhancing efficiency, reducing manpower requirements, and facilitating timely treatment decisions in breast cancer care.

Microwave ablation for colorectal liver metastases with ultrasound fusion imaging assistance: a stratified analysis study based on tumor size and location

PURPOSE

To investigate the efficacy of ultrasound fusion imaging-assisted microwave ablation (MWA) for patients with colorectal liver metastases (CRLM) based on stratified analysis of tumor size and location.

METHODS

Patients with CRLM who underwent ultrasound fusion imaging-assisted MWA in our hospital between February 2020 and February 2023 were enrolled into this retrospective study. Ultrasound fusion imaging was used for detection, guidance, monitoring and immediate evaluation throughout the MWA procedures. Technical success, technique efficacy, local tumor progression (LTP), intrahepatic progression and overall survival (OS) were recorded and analyzed. The subgroup analysis of intrahepatic progression of MWA for CRLM was performed according to tumor size and location.

RESULTS

A total of 51 patients with 122 nodules were enrolled. Both technical success and technique efficacy were acquired in all nodules. In a median follow-up period of 19 months, 2.5% of the nodules (3/122) were observed LTP. The 1-year and 2-year cumulative intrahepatic progression rates were 38.7% and 52.1% respectively. Patients were divided into subgroups according to tumor size (≥ 30 mm, n = 13; < 30 mm, n = 38) and tumor location (perivascular, n = 20; non-perivascular, n = 31 and subcapsular, n = 36; non-subcapsular, n = 15). The cumulative intrahepatic progression rates were similar between the subgroups regarding tumor size and perivascular location, while significantly higher in the subcapsular group than in the non-subcapsular group (p = 0.021).

CONCLUSION

Ultrasound fusion imaging-assisted MWA exhibited satisfactory local efficacy for CRLM, especially for non-subcapsular tumors.

GeoCrack: A High-Resolution Dataset For Segmentation of Fracture Edges in Geological Outcrops

GeoCrack is the first large-scale open source annotated dataset of fracture traces from geological outcrops, enabling deep learning-based fracture segmentation, setting a new standard for natural fracture characterization datasets. GeoCrack contains images from photogrammetric surveys of fractured rock exposures across 11 sites in Europe and the Middle East, capturing diverse lithologies and tectonic settings. Each image was cleaned, normalized, and manually segmented, followed by a recursive annotation vetting process to ensure the quality and accuracy of the digitized fracture edges. The processed images and corresponding binary masks were divided into 224 × 224 patches, yielding 12,158 pairs. GeoCrack captures representive real-world challenges in fracture edge annotation, such as contrast variations between fracture traces and the host medium due to geological and geomorphological factors like aperture dilation, host rock composition, outcrop weathering, and groundwater staining. Physical occlusions like shadows and vegetation are also considered to minimize false positives. GeoCrack was validated using a U-Net implementation for fracture segmentation, achieving satisfactory IoU of 85%. GeoCrack holds strong potential to advance deep fracture segmentation in geological applications, effectively tackling the diverse challenges of real-world fracture identification.

Precise liver tumor ablation: the clinical potential of US-US overlay fusion guidance

Image-guided thermal ablation is a minimally invasive option for patients with early-stage hepatocellular carcinoma (HCC). However, the risk of local recurrence remains substantial because ultrasound (US) artifacts have a negative impact on the assessment of ablative margins during and immediately after ablation. Precise, real-time assessment of the ablation zone is key to reducing the risk of local tumor progression. With the advent of US image fusion technology, ablative margins can now be assessed three-dimensionally with greater accuracy. Therefore, US-US overlay fusion guidance has the potential to improve the local controllability of ablation in patients with HCC. This review discusses the US-US fusion guidance technique and its current clinical applications for hepatic interventions, with descriptions of its concept, methodology, and efficacy.

Artificial intelligence in predicting recurrence after first-line treatment of liver cancer: a systematic review and meta-analysis

BACKGROUND

The aim of this study was to conduct a systematic review and meta-analysis to comprehensively evaluate the performance and methodological quality of artificial intelligence (AI) in predicting recurrence after single first-line treatment for liver cancer.

METHODS

A rigorous and systematic evaluation was conducted on the AI studies related to recurrence after single first-line treatment for liver cancer, retrieved from the PubMed, Embase, Web of Science, Cochrane Library, and CNKI databases. The area under the curve (AUC), sensitivity (SENC), and specificity (SPEC) of each study were extracted for meta-analysis.

RESULTS

Six percutaneous ablation (PA) studies, 16 surgical resection (SR) studies, and 5 transarterial chemoembolization (TACE) studies were included in the meta-analysis for predicting recurrence after hepatocellular carcinoma (HCC) treatment, respectively. Four SR studies and 2 PA studies were included in the meta-analysis for recurrence after intrahepatic cholangiocarcinoma (ICC) and colorectal cancer liver metastasis (CRLM) treatment. The pooled SENC, SEPC, and AUC of AI in predicting recurrence after primary HCC treatment via PA, SR, and TACE were 0.78, 0.90, and 0.92; 0.81, 0.77, and 0.86; and 0.73, 0.79, and 0.79, respectively. The values for ICC treated with SR and CRLM treated with PA were 0.85, 0.71, 0.86 and 0.69, 0.63,0.74, respectively.

CONCLUSION

This systematic review and meta-analysis demonstrates the comprehensive application value of AI in predicting recurrence after a single first-line treatment of liver cancer, with satisfactory results, indicating the clinical translation potential of AI in predicting recurrence after liver cancer treatment.

The value evaluation of Nomogram prediction model based on CTA imaging features for selecting treatment methods for isolated superior mesenteric artery dissection

OBJECTIVE

To evaluate value of Nomogram prediction model based on CTA imaging features for selecting treatment methods for isolated superior mesenteric artery dissection (ISMAD).

METHODS

Symptomatic ISMAD patients were randomly divided into a training set and a validation set in a 7:3 ratio. In the training set, relevant risk factors for conservative treatment failure in ISMAD patients were analyzed, and a Nomogram prediction model for treatment outcome of ISMAD was constructed with risk factors. The predictive value of the model was evaluated.

RESULTS

Low true lumen residual ratio (TLRR), long dissection length, and large arterial angle (superior mesenteric artery [SMA]/abdominal aorta [AA]) were identified as independent high-risk factors for conservative treatment failure (P < 0.05). The receiver operating characteristic curve (ROC) results showed that the area under curve (AUC) of Nomogram prediction model was 0.826 (95% CI: 0.740-0.912), indicating good discrimination. The Hosmer-Lemeshow goodness-of-fit test showed good consistency between the predicted curve and the ideal curve of the Nomogram prediction model. The decision curve analysis (DCA) analysis results showed that when probability threshold for the occurrence of conservative treatment failure predicted was 0.05-0.98, patients could obtain more net benefits. Similar results were obtained for the predictive value in the validation set.

CONCLUSION

Low TLRR, long dissection length, and large arterial angle (SMA/AA) are independent high-risk factors for conservative treatment failure in ISMAD. The Nomogram model constructed with independent high-risk factors has good clinical effectiveness in predicting the failure.

MRI T2WI-based radiomics combined with KRAS gene mutation constructed models for predicting liver metastasis in rectal cancer

BACKGROUND

The study aimed to identify the optimal model for predicting rectal cancer liver metastasis (RCLM). This involved constructing various prediction models to aid clinicians in early diagnosis and precise decision-making.

METHODS

A retrospective analysis was conducted on 193 patients diagnosed with rectal adenocarcinoma were randomly divided into training set (n = 136) and validation set (n = 57) at a ratio of 7:3. The predictive performance of three models was internally validated by 10-fold cross-validation in the training set. Delineation of the tumor region of interest (ROI) was performed, followed by the extraction of radiomics features from the ROI. The least absolute shrinkage and selection operator (LASSO) regression algorithm and multivariate Cox analysis were employed to reduce the dimensionality of radiomics features and identify significant features. Logistic regression was employed to construct three prediction models: clinical, radiomics, and combined models (radiomics + clinical). The predictive performance of each model was assessed and compared.

RESULTS

KRAS mutation emerged as an independent predictor of liver metastasis, yielding an odds ratio (OR) of 8.296 (95%CI: 3.471-19.830; p < 0.001). 5 radiomics features will be used to construct radiomics model. The combined model was built by integrating radiomics model with clinical model. In both the training set (AUC:0.842, 95%CI: 0.778-0.907) and the validation set (AUC: 0.805; 95%CI: 0.692-0.918), the AUCs for the combined model surpassed those of the radiomics and clinical models.

CONCLUSIONS

Our study reveals that KRAS mutation stands as an independent predictor of RCLM. The radiomics features based on MR play a crucial role in the evaluation of RCLM. The combined model exhibits superior performance in the prediction of liver metastasis.

CLINICAL TRIAL NUMBER

Not applicable.

The Paradox of Modern Technology in Standardizing Thermal Liver Ablation: Fostering Uniformity or Diversity?

PURPOSE

Currently, significant medical practice variation exists in thermal ablation (TA) of malignant liver tumors with associated differences in outcomes. The IMaging and Advanced Guidance for workflow optimization in Interventional Oncology (IMAGIO) consortium aims to integrate interventional oncology into the standard clinical pathway for cancer treatment in Europe by 2030, by development of a standardized low-complex-high-precision workflow for TA of malignant liver tumors. This study was conducted at the start of the IMAGIO project with the aim to explore the current state and future role of modern technology in TA of malignant liver tumors.

MATERIALS AND METHODS

A cross-sectional questionnaire was conducted followed by an expert focus group discussion with core members and collaborating partners of the consortium.

RESULTS

Of the 13 participants, 10 respondents filled in the questionnaire. During the focus group discussion, there was consensus on the need for international standardization in TA and several aspects of the procedure, such as planning based on cross-sectional images, the adoption of different techniques for needle placement and the importance of needle position- and post-ablative margin confirmation scans. Yet, also considerable heterogeneity was reported in the adoption of modern technology, particularly in navigational systems and computer-assisted margin assessment.

CONCLUSION

This study mirrored the current diversity in workflow of thermal liver ablation. To obtain comparable outcomes worldwide, standardization is needed. While advancements in tools and software hold the potential to homogenize outcome measurement and minimize operator-dependent variability, the rapid increase in availability also contributes to enhanced workflow variation.

Outcomes and Complications of Image-Guided Percutaneous Tumour Ablation for Hepatocellular Carcinoma at the Irish National Liver Transplant Centre

Background: Image-guided tumour ablation is a minimally invasive treatment for early stage hepatocellular carcinoma (HCC). Our study reviews the complications and long term outcomes in patients treated at a tertiary referral centre. Methods: Retrospective study. All patients with HCC who underwent microwave ablation (MWA) or radiofrequency ablation (RFA) from 1st January 2014 to 31st December 2022 were identified. Treatment response of target lesion, complications, and survival were recorded. Results: One hundred seventy ablations were performed in 118 patients; 70% MWA, 30% RFA. Median radiological follow-up 21 months (range 3-107). Follow-up imaging was reported using LI-RADS and mRECIST. At first follow-up imaging, 94 patients had complete response (primary efficacy rate 80.3%) while 19.7% (n = 23) had residual disease. Fifteen of these had repeat ablation; 10 had complete response (secondary efficacy rate 85.6%). By end of study duration, 70.5% (n = 79) achieved sustained local complete response from single ablation without documented recurrence. 14.3% (n = 16) required more than one ablation of target lesion. Overall, 84.8% (n = 95) demonstrated long term local complete response to ablation. Complication occurred in 5.9% (n = 10); 40.0% Grade I, 40.0% Grade II, 10.0% Grade III, 10.0% Grade IV as per the CIRSE Classification. 1-, 3-, and 5-year overall survival (OS) rate was 97%, 68%, and 61% respectively. Mean OS was 5.3 years (median 4.7). No difference in OS (P = .7) or local progression free survival (P = .5) between patients treated with MWA versus RFA. Conclusion: This study demonstrates excellent long-term response to TA, with acceptable complication profile. No difference in survival between RFA versus MWA.

Accuracy of deep learning in the differential diagnosis of coronary artery stenosis: a systematic review and meta-analysis

BACKGROUND

In recent years, as deep learning has received widespread attention in the field of heart disease, some studies have explored the potential of deep learning based on coronary angiography (CAG) or coronary CT angiography (CCTA) images in detecting the extent of coronary artery stenosis. However, there is still a lack of a systematic understanding of its diagnostic accuracy, impeding the advancement of intelligent diagnosis of coronary artery stenosis. Therefore, we conducted this study to review the accuracy of image-based deep learning in detecting coronary artery stenosis.

METHODS

We retrieved PubMed, Cochrane, Embase, and Web of Science until April 11, 2023. The risk of bias in the included studies was appraised using the QUADAS-2 tool. We extracted the accuracy of deep learning in the test set and performed subgroup analyses by binary and multiclass classification scenarios. We performed a subgroup analysis based on different degrees of stenosis and applied a double arcsine transformation to process the data. The analysis was done by using R.

RESULTS

Our systematic review finally included 18 studies, involving 3568 patients and 13,362 images. In the included studies, deep learning models were constructed based on CAG and CCTA. In binary classification tasks, the accuracy for detecting > 25%, > 50% and > 70% degrees of stenosis at the vessel level were 0.81 (95% CI: 0.71-0.85), 0.73 (95% CI: 0.58-0.88) and 0.61 (95% CI: 0.56-0.65), respectively. In multiclass classification tasks, the accuracy for detecting 0-25%, 25-50%, 50-70%, and 70-100% degrees of stenosis at the vessel level were 0.78 (95% CI: 0.73-0.84), 0.86 (95% CI: 0.78-0.93), 0.83 (95% CI: 0.70-0.97), and 0.70 (95% CI: 0.42-0.98), respectively.

CONCLUSIONS

Our study shows that deep learning models based on CAG and CCTA appear to be relatively accurate in diagnosing different degrees of coronary artery stenosis. However, for various degrees of stenosis, their accuracy still needs to be further improved.

Predicting invasion in early-stage ground-glass opacity pulmonary adenocarcinoma: a radiomics-based machine learning approach

BACKGROUND

To design a pulmonary ground-glass nodules (GGN) classification method based on computed tomography (CT) radiomics and machine learning for prediction of invasion in early-stage ground-glass opacity (GGO) pulmonary adenocarcinoma.

METHODS

This retrospective study included pulmonary GGN patients who were histologically confirmed to have adenocarcinoma in situ (AIS), minimally invasive adenocarcinoma (MIA), or invasive adenocarcinoma cancer (IAC) from 2020 to 2023. CT images of all patients were automatically segmented and 107 radiomic features were obtained for each patient. Classification models were developed using random forest (RF) and cross-validation, including three one-versus-others models and one three-class model. For each model, features were ranked by normalized Gini importance, and a minimal subset was selected with a cumulative importance exceeding 0.9. These selected features were then used to train the final models. The models' performance metrics, including area under the curve (AUC), accuracy, sensitivity, and specificity, were computed. AUC and accuracy were compared to determine the final optimal method.

RESULTS

The study comprised 193 patients (mean age 54 ± 11 years, 65 men), including 65 AIS, 54 MIA, and 74 IAC, divided into one training cohort (N = 154) and one test cohort (N = 39). The final three-class RF model outperformed three individual one-versus-others models in distinguishing each class from the other two. For the multiclass classification model, the AUC, accuracy, sensitivity, and specificity were 0.87, 0.79, 0.62, and 0.88 for AIS; 0.90, 0.79, 0.54, and 0.89 for MIA; and 0.87, 0.69, 0.73, and 0.67 for IAC, respectively.

CONCLUSIONS

A radiomics-based multiclass RF model could effectively differentiate three types of pulmonary GGN, which enabled early diagnosis of GGO pulmonary adenocarcinoma.

nnU-Net based segmentation and 3D reconstruction of uterine fibroids with MRI images for HIFU surgery planning

High-Intensity Focused Ultrasound (HIFU) ablation represents a rapidly advancing non-invasive treatment modality that has achieved considerable success in addressing uterine fibroids, which constitute over 50% of benign gynecological tumors. Preoperative Magnetic Resonance Imaging (MRI) plays a pivotal role in the planning and guidance of HIFU surgery for uterine fibroids, wherein the segmentation of tumors holds critical significance. The segmentation process was previously manually executed by medical experts, entailing a time-consuming and labor-intensive procedure heavily reliant on clinical expertise. This study introduced deep learning-based nnU-Net models, offering a cost-effective approach for their application in the segmentation of uterine fibroids utilizing preoperative MRI images. Furthermore, 3D reconstruction of the segmented targets was implemented to guide HIFU surgery. The evaluation of segmentation and 3D reconstruction performance was conducted with a focus on enhancing the safety and effectiveness of HIFU surgery. Results demonstrated the nnU-Net's commendable performance in the segmentation of uterine fibroids and their surrounding organs. Specifically, 3D nnU-Net achieved Dice Similarity Coefficients (DSC) of 92.55% for the uterus, 95.63% for fibroids, 92.69% for the spine, 89.63% for the endometrium, 97.75% for the bladder, and 90.45% for the urethral orifice. Compared to other state-of-the-art methods such as HIFUNet, U-Net, R2U-Net, ConvUNeXt and 2D nnU-Net, 3D nnU-Net demonstrated significantly higher DSC values, highlighting its superior accuracy and robustness. In conclusion, the efficacy of the 3D nnU-Net model for automated segmentation of the uterus and its surrounding organs was robustly validated. When integrated with intra-operative ultrasound imaging, this segmentation method and 3D reconstruction hold substantial potential to enhance the safety and efficiency of HIFU surgery in the clinical treatment of uterine fibroids.

Fractional differentiation based image enhancement for automatic detection of malignant melanoma

Recent improvements in artificial intelligence and computer vision make it possible to automatically detect abnormalities in medical images. Skin lesions are one broad class of them. There are types of lesions that cause skin cancer, again with several types. Melanoma is one of the deadliest types of skin cancer. Its early diagnosis is at utmost importance. The treatments are greatly aided with artificial intelligence by the quick and precise diagnosis of these conditions. The identification and delineation of boundaries inside skin lesions have shown promise when using the basic image processing approaches for edge detection. Further enhancements regarding edge detections are possible. In this paper, the use of fractional differentiation for improved edge detection is explored on the application of skin lesion detection. A framework based on fractional differential filters for edge detection in skin lesion images is proposed that can improve automatic detection rate of malignant melanoma. The derived images are used to enhance the input images. Obtained images then undergo a classification process based on deep learning. A well-studied dataset of HAM10000 is used in the experiments. The system achieves 81.04% accuracy with EfficientNet model using the proposed fractional derivative based enhancements whereas accuracies are around 77.94% when using original images. In almost all the experiments, the enhanced images improved the accuracy. The results show that the proposed method improves the recognition performance.

Dual-energy computed tomography with new virtual monoenergetic image reconstruction enhances prostate lesion image quality and improves the diagnostic efficacy for prostate cancer

BACKGROUND

Prostate cancer is one of the most common malignant tumors in middle-aged and elderly men and carries significant prognostic implications, and recent studies suggest that dual-energy computed tomography (DECT) utilizing new virtual monoenergetic images can enhance cancer detection rates. This study aimed to assess the impact of virtual monoenergetic images reconstructed from DECT arterial phase scans on the image quality of prostate lesions and their diagnostic performance for prostate cancer.

METHODS

We conducted a retrospective analysis of 83 patients with prostate cancer or prostatic hyperplasia who underwent DECT scans at Meizhou People's Hospital between July 2019 and December 2023. The variables analyzed included age, tumor diameter and serum prostate-specific antigen (PSA) levels, among others. We also compared CT values, signal-to-noise ratio (SNR), subjective image quality ratings, and contrast-to-noise ratio (CNR) between virtual monoenergetic images (40-100 keV) and conventional linear blending images. Receiver operating characteristic (ROC) curve analyses were performed to evaluate the diagnostic efficacy of virtual monoenergetic images (40 keV and 50 keV) compared to conventional images.

RESULTS

Virtual monoenergetic images at 40 keV showed significantly higher CT values (168.19 ± 57.14) compared to conventional linear blending images (66.66 ± 15.5) for prostate cancer (P < 0.001). The 50 keV images also demonstrated elevated CT values (121.73 ± 39.21) compared to conventional images (P < 0.001). CNR values for the 40 keV (3.81 ± 2.13) and 50 keV (2.95 ± 1.50) groups were significantly higher than the conventional blending group (P < 0.001). Subjective evaluations indicated markedly better image quality scores for 40 keV (median score of 5) and 50 keV (median score of 5) images compared to conventional images (P < 0.05). ROC curve analysis revealed superior diagnostic accuracy for 40 keV (AUC: 0.910) and 50 keV (AUC: 0.910) images based on CT values compared to conventional images (AUC: 0.849).

CONCLUSIONS

Virtual monoenergetic images reconstructed at 40 keV and 50 keV from DECT arterial phase scans substantially enhance the image quality of prostate lesions and improve diagnostic efficacy for prostate cancer.

An effective no-reference image quality index prediction with a hybrid Artificial Intelligence approach for denoised MRI images

As the quantity and significance of digital pictures in the medical industry continue to increase, Image Quality Assessment (IQA) has recently become a prevalent subject in the research community. Due to the wide range of distortions that Magnetic Resonance Images (MRI) can experience and the wide variety of information they contain, No-Reference Image Quality Assessment (NR-IQA) has always been a challenging study issue. In an attempt to address this issue, a novel hybrid Artificial Intelligence (AI) is proposed to analyze NR-IQ in massive MRI data. First, the features from the denoised MRI images are extracted using the gray level run length matrix (GLRLM) and EfficientNet B7 algorithm. Next, the Multi-Objective Reptile Search Algorithm (MRSA) was proposed for optimal feature vector selection. Then, the Self-evolving Deep Belief Fuzzy Neural network (SDBFN) algorithm was proposed for the effective NR-IQ analysis. The implementation of this research is executed using MATLAB software. The simulation results are compared with the various conventional methods in terms of correlation coefficient (PLCC), Root Mean Square Error (RMSE), Spearman Rank Order Correlation Coefficient (SROCC) and Kendall Rank Order Correlation Coefficient (KROCC), and Mean Absolute Error (MAE). In addition, our proposed approach yielded a quality number approximately we achieved significant 20% improvement than existing methods, with the PLCC parameter showing a notable increase compared to current techniques. Moreover, the RMSE number decreased by 12% when compared to existing methods. Graphical representations indicated mean MAE values of 0.02 for MRI knee dataset, 0.09 for MRI brain dataset, and 0.098 for MRI breast dataset, showcasing significantly lower MAE values compared to the baseline models.

Retinex theory-based nonlinear luminance enhancement and denoising for low-light endoscopic images

BACKGROUND

The quality of low-light endoscopic images involves applications in medical disciplines such as physiology and anatomy for the identification and judgement of tissue structures. Due to the use of point light sources and the constraints of narrow physiological structures, medical endoscopic images display uneven brightness, low contrast, and a lack of texture information, presenting diagnostic challenges for physicians.

METHODS

In this paper, a nonlinear brightness enhancement and denoising network based on Retinex theory is designed to improve the brightness and details of low-light endoscopic images. The nonlinear luminance enhancement module uses higher-order curvilinear functions to improve overall brightness; the dual-attention denoising module captures detailed features of anatomical structures; and the color loss function mitigates color distortion.

RESULTS

Experimental results on the Endo4IE dataset demonstrate that the proposed method outperforms existing state-of-the-art methods in terms of Peak Signal-to-Noise Ratio (PSNR), Structural Similarity (SSIM), and Learned Perceptual Image Patch Similarity (LPIPS). The PSNR is 27.2202, SSIM is 0.8342, and the LPIPS is 0.1492. It provides a method to enhance image quality in clinical diagnosis and treatment.

CONCLUSIONS

It offers an efficient method to enhance images captured by endoscopes and offers valuable insights into intricate human physiological structures, which can effectively assist clinical diagnosis and treatment.

Metal implant segmentation in CT images based on diffusion model

BACKGROUND

Computed tomography (CT) is widely in clinics and is affected by metal implants. Metal segmentation is crucial for metal artifact correction, and the common threshold method often fails to accurately segment metals.

PURPOSE

This study aims to segment metal implants in CT images using a diffusion model and further validate it with clinical artifact images and phantom images of known size.

METHODS

A retrospective study was conducted on 100 patients who received radiation therapy without metal artifacts, and simulated artifact data were generated using publicly available mask data. The study utilized 11,280 slices for training and verification, and 2,820 slices for testing. Metal mask segmentation was performed using DiffSeg, a diffusion model incorporating conditional dynamic coding and a global frequency parser (GFParser). Conditional dynamic coding fuses the current segmentation mask and prior images at multiple scales, while GFParser helps eliminate high-frequency noise in the mask. Clinical artifact images and phantom images are also used for model validation.

RESULTS

Compared with the ground truth, the accuracy of DiffSeg for metal segmentation of simulated data was 97.89% and that of DSC was 95.45%. The mask shape obtained by threshold segmentation covered the ground truth and DSCs were 82.92% and 84.19% for threshold segmentation based on 2500 HU and 3000 HU. Evaluation metrics and visualization results show that DiffSeg performs better than other classical deep learning networks, especially for clinical CT, artifact data, and phantom data.

CONCLUSION

DiffSeg efficiently and robustly segments metal masks in artifact data with conditional dynamic coding and GFParser. Future work will involve embedding the metal segmentation model in metal artifact reduction to improve the reduction effect.

Development of a flexible liver phantom for hepatocellular carcinoma treatment planning: a useful tool for training & education

PURPOSE

Hepatocellular carcinoma (HCC) is one of the most common types of liver cancer that could potentially be surrounded by healthy arteries or veins that a surgeon would have to avoid during treatment. A realistic 3D liver model is an unmet need for HCC preoperative planning.

METHODS

This paper presents a method to create a soft phantom model of the human liver with the help of a 3D-printed mold, silicone, ballistic gel, and a blender.

RESULTS

For silicone, the elastic modulus of seven different ratios of base silicone and silicone hardener are tested; while for ballistic gel, a model using 20% gelatin and 10% gelatin is created for the tumor and the rest of the liver, respectively. It is found that the silicone modulus of elasticity matches with the real liver modulus of elasticity. It is also found that the 10% gelatin part of the ballistic gel model is an excellent emulation of a healthy human liver.

CONCLUSION

The 3D flexible liver phantom made from a 10% gelatin-to-water mixture demonstrates decent fidelity to real liver tissue in terms of texture and elasticity. It holds significant potential for improving medical training, preoperative planning, and surgical research. We believe that continued development and validation of such models could further enhance their utility and impact in the field of hepatobiliary treatment planning and education.

STC-UNet: renal tumor segmentation based on enhanced feature extraction at different network levels

Renal tumors are one of the common diseases of urology, and precise segmentation of these tumors plays a crucial role in aiding physicians to improve diagnostic accuracy and treatment effectiveness. Nevertheless, inherent challenges associated with renal tumors, such as indistinct boundaries, morphological variations, and uncertainties in size and location, segmenting renal tumors accurately remains a significant challenge in the field of medical image segmentation. With the development of deep learning, substantial achievements have been made in the domain of medical image segmentation. However, existing models lack specificity in extracting features of renal tumors across different network hierarchies, which results in insufficient extraction of renal tumor features and subsequently affects the accuracy of renal tumor segmentation. To address this issue, we propose the Selective Kernel, Vision Transformer, and Coordinate Attention Enhanced U-Net (STC-UNet). This model aims to enhance feature extraction, adapting to the distinctive characteristics of renal tumors across various network levels. Specifically, the Selective Kernel modules are introduced in the shallow layers of the U-Net, where detailed features are more abundant. By selectively employing convolutional kernels of different scales, the model enhances its capability to extract detailed features of renal tumors across multiple scales. Subsequently, in the deeper layers of the network, where feature maps are smaller yet contain rich semantic information, the Vision Transformer modules are integrated in a non-patch manner. These assist the model in capturing long-range contextual information globally. Their non-patch implementation facilitates the capture of fine-grained features, thereby achieving collaborative enhancement of global-local information and ultimately strengthening the model's extraction of semantic features of renal tumors. Finally, in the decoder segment, the Coordinate Attention modules embedding positional information are proposed aiming to enhance the model's feature recovery and tumor region localization capabilities. Our model is validated on the KiTS19 dataset, and experimental results indicate that compared to the baseline model, STC-UNet shows improvements of 1.60%, 2.02%, 2.27%, 1.18%, 1.52%, and 1.35% in IoU, Dice, Accuracy, Precision, Recall, and F1-score, respectively. Furthermore, the experimental results demonstrate that the proposed STC-UNet method surpasses other advanced algorithms in both visual effectiveness and objective evaluation metrics.

Past, present, and future perspectives of ultrasound-guided ablation of liver tumors: Where could artificial intelligence lead interventional oncology?

The first ablation procedures for small hepatocellular carcinomas were percutaneous ethanol injection under ultrasound (US) guidance. Later, radiofrequency ablation was shown to achieve larger coagulation areas than percutaneous ethanol injection and became the most used ablation technique worldwide. In the past decade, microwave ablation systems have achieved larger ablation areas than radiofrequency ablation, suggesting that the 3-cm barrier could be broken in the treatment of liver tumors. Likewise, US techniques to guide percutaneous ablation have seen important progress. Contrast-enhanced US (CEUS) can define and target the tumor better than US and can assess the size of the ablation area after the procedure, which allows immediate retreatment of the residual tumor foci. Furthermore, fusion imaging fuses real-time US images with computed tomography or magnetic resonance imaging with significant improvements in detecting and targeting lesions with low conspicuity on CEUS. Recently, software powered by artificial intelligence has been developed to allow three-dimensional segmentation and reconstruction of the anatomical structures, aiding in procedure planning, assessing ablation completeness, and targeting the residual viable foci with greater precision than CEUS. Hopefully, this could lead to the ablation of tumors up to 5-7 cm in size.

Enhancing ECG-based heart age: impact of acquisition parameters and generalization strategies for varying signal morphologies and corruptions

Electrocardiogram (ECG) is a non-invasive approach to capture the overall electrical activity produced by the contraction and relaxation of the cardiac muscles. It has been established in the literature that the difference between ECG-derived age and chronological age represents a general measure of cardiovascular health. Elevated ECG-derived age strongly correlates with cardiovascular conditions (e.g., atherosclerotic cardiovascular disease). However, the neural networks for ECG age estimation are yet to be thoroughly evaluated from the perspective of ECG acquisition parameters. Additionally, deep learning systems for ECG analysis encounter challenges in generalizing across diverse ECG morphologies in various ethnic groups and are susceptible to errors with signals that exhibit random or systematic distortions To address these challenges, we perform a comprehensive empirical study to determine the threshold for the sampling rate and duration of ECG signals while considering their impact on the computational cost of the neural networks. To tackle the concern of ECG waveform variability in different populations, we evaluate the feasibility of utilizing pre-trained and fine-tuned networks to estimate ECG age in different ethnic groups. Additionally, we empirically demonstrate that finetuning is an environmentally sustainable way to train neural networks, and it significantly decreases the ECG instances required (by more than 100 × ) for attaining performance similar to the networks trained from random weight initialization on a complete dataset. Finally, we systematically evaluate augmentation schemes for ECG signals in the context of age estimation and introduce a random cropping scheme that provides best-in-class performance while using shorter-duration ECG signals. The results also show that random cropping enables the networks to perform well with systematic and random ECG signal corruptions.

Ensemble learning enhances the precision of preliminary detection of primary hepatocellular carcinoma based on serological and demographic indices

Primary hepatocellular carcinoma (PHC) is associated with high rates of morbidity and malignancy in China and throughout the world. In clinical practice, a combination of ultrasound and alpha-fetoprotein (AFP) measurement is frequently employed for initial screening. However, the accuracy of this approach often falls short of the desired standard. Consequently, this study aimed to investigate the enhancement of precision of preliminary detection of PHC by ensemble learning techniques. To achieve this, 712 patients with PHC and 1887 healthy controls were enrolled for the assessment of four ensemble learning methods, namely, Random Forest (RF), LightGBM, Xgboost, and Catboost. A total of eleven characteristics, comprising nine serological indices and two demographic indices, were selected from the participants for use in detecting PHC. The findings identified an optimal feature subset consisting of eight features, namely AFP, albumin (ALB), alanine aminotransferase (ALT), platelets (PLT), age, alkaline phosphatase (ALP), hemoglobin (Hb), and body mass index (BMI), that achieved the highest classification accuracy of 96.62%. This emphasizes the importance of the collective use of these features in PHC diagnosis. In conclusion, the results provide evidence that the integration of serological and demographic indices together with ensemble learning models, can contribute to the precision of preliminary diagnosis of PHC.

Preoperative single-dose camrelizumab and/or microwave ablation in women with early-stage breast cancer: A window-of-opportunity trial

BACKGROUND

Immune checkpoint blockade has shown low response rates for advanced breast cancer, and combination strategies are needed. Microwave ablation (MWA) may be a trigger of antitumor immunity. This window-of-opportunity trial (ClinicalTrials.gov: NCT04805736) was conducted to determine the safety and feasibility of preoperative camrelizumab (an anti-PD-1 antibody) combined with MWA in the treatment of early-stage breast cancer.

METHODS

Sixty participants were randomized to preoperatively receive single-dose camrelizumab alone (n = 20), MWA alone (n = 20), or camrelizumab+MWA (n = 20). A random number table was used to allocate interventions. The primary outcome was the safety and feasibility of MWA combined with camrelizumab.

FINDINGS

Camrelizumab and MWA were well tolerated alone and in combination without delays in prescheduled surgery. No treatment-related grade III/IV adverse events were observed. Different from in the single-dose camrelizumab or MWA group, participants showed stable counts of blood cells after combination therapy. After combination therapy, peripheral CD8+ T cells showed enhanced cytotoxic and effect-memory functions. Clonal expansional CD8+ T cells showed higher cytotoxic activity and effector memory- and tumor-specific signatures than emergent clones after combination therapy. Enhanced interactions between clonal expansional CD8+ T cells and monocytes were observed, suggesting that monocytes contributed to the enhanced functions of clonal expansional CD8+ T cells. Major histocompatibility complex (MHC) class I-related pathways and interferon signaling pathways were activated in monocytes by combination therapy.

CONCLUSIONS

Camrelizumab combined with MWA was feasible for early-stage breast cancer. Peripheral CD8+ T cells were activated after combination therapy, dependent on monocytes with activated MHC class I pathways.

FUNDING

This study was supported by the Natural Science Foundation of Jiangsu Province (BK20230017).

Unveiling the future of breast cancer assessment: a critical review on generative adversarial networks in elastography ultrasound

Elastography Ultrasound provides elasticity information of the tissues, which is crucial for understanding the density and texture, allowing for the diagnosis of different medical conditions such as fibrosis and cancer. In the current medical imaging scenario, elastograms for B-mode Ultrasound are restricted to well-equipped hospitals, making the modality unavailable for pocket ultrasound. To highlight the recent progress in elastogram synthesis, this article performs a critical review of generative adversarial network (GAN) methodology for elastogram generation from B-mode Ultrasound images. Along with a brief overview of cutting-edge medical image synthesis, the article highlights the contribution of the GAN framework in light of its impact and thoroughly analyzes the results to validate whether the existing challenges have been effectively addressed. Specifically, This article highlights that GANs can successfully generate accurate elastograms for deep-seated breast tumors (without having artifacts) and improve diagnostic effectiveness for pocket US. Furthermore, the results of the GAN framework are thoroughly analyzed by considering the quantitative metrics, visual evaluations, and cancer diagnostic accuracy. Finally, essential unaddressed challenges that lie at the intersection of elastography and GANs are presented, and a few future directions are shared for the elastogram synthesis research.

Estimating age and gender from electrocardiogram signals: A comprehensive review of the past decade

Twelve lead electrocardiogram signals capture unique fingerprints about the body's biological processes and electrical activity of heart muscles. Machine learning and deep learning-based models can learn the embedded patterns in the electrocardiogram to estimate complex metrics such as age and gender that depend on multiple aspects of human physiology. ECG estimated age with respect to the chronological age reflects the overall well-being of the cardiovascular system, with significant positive deviations indicating an aged cardiovascular system and a higher likelihood of cardiovascular mortality. Several conventional, machine learning, and deep learning-based methods have been proposed to estimate age from electronic health records, health surveys, and ECG data. This manuscript comprehensively reviews the methodologies proposed for ECG-based age and gender estimation over the last decade. Specifically, the review highlights that elevated ECG age is associated with atherosclerotic cardiovascular disease, abnormal peripheral endothelial dysfunction, and high mortality, among many other cardiovascular disorders. Furthermore, the survey presents overarching observations and insights across methods for age and gender estimation. This paper also presents several essential methodological improvements and clinical applications of ECG-estimated age and gender to encourage further improvements of the state-of-the-art methodologies.