BACKGROUND: Neoadjuvant chemoradiotherapy followed by radical surgery has been a common practice for patients with locally advanced rectal cancer, but the response rate varies among patients. This study aimed to develop a ResNet-Vision Transformer based magnetic resonance imaging (MRI)-endoscopy fusion model to precisely predict treatment response and provide personalized treatment. METHODS: In this multicenter study, 366 eligible patients who had undergone neoadjuvant chemoradiotherapy followed by radical surgery at eight Chinese tertiary hospitals between January 2017 and June 2024 were recruited, with 2928 pretreatment colonic endoscopic images and 366 pelvic MRI images. An MRI-endoscopy fusion model was constructed based on the ResNet backbone and Transformer network using pretreatment MRI and endoscopic images. Treatment response was defined as good response or non-good response based on the tumor regression grade. The Delong test and the Hanley-McNeil test were utilized to compare prediction performance among different models and different subgroups, respectively. The predictive performance of the MRI-endoscopy fusion model was comprehensively validated in the test sets and was further compared to that of the single-modal MRI model and single-modal endoscopy model. RESULTS: The MRI-endoscopy fusion model demonstrated favorable prediction performance. In the internal validation set, the area under the curve (AUC) and accuracy were 0.852 (95% confidence interval [CI]: 0.744-0.940) and 0.737 (95% CI: 0.712-0.844), respectively. Moreover, the AUC and accuracy reached 0.769 (95% CI: 0.678-0.861) and 0.729 (95% CI: 0.628-0.821), respectively, in the external test set. In addition, the MRI-endoscopy fusion model outperformed the single-modal MRI model (AUC: 0.692 [95% CI: 0.609-0.783], accuracy: 0.659 [95% CI: 0.565-0.775]) and the single-modal endoscopy model (AUC: 0.720 [95% CI: 0.617-0.823], accuracy: 0.713 [95% CI: 0.612-0.809]) in the external test set. CONCLUSION The MRI-endoscopy fusion model based on ResNet-Vision Transformer achieved favorable performance in predicting treatment response to neoadjuvant chemoradiotherapy and holds tremendous potential for enabling personalized treatment regimens for locally advanced rectal cancer patients.
Humans ; Rectal Neoplasms/diagnostic imaging* ; Magnetic Resonance Imaging/methods* ; Male ; Female ; Middle Aged ; Neoadjuvant Therapy/methods* ; Aged ; Adult ; Chemoradiotherapy/methods* ; Endoscopy/methods* ; Treatment Outcome

The accurate segmentation of breast ultrasound images is an important precondition for the lesion determination. The existing segmentation approaches embrace massive parameters, sluggish inference speed, and huge memory consumption. To tackle this problem, we propose T ²KD Attention U-Net (dual-Teacher Knowledge Distillation Attention U-Net), a lightweight semantic segmentation method combined double-path joint distillation in breast ultrasound images. Primarily, we designed two teacher models to learn the fine-grained features from each class of images according to different feature representation and semantic information of benign and malignant breast lesions. Then we leveraged the joint distillation to train a lightweight student model. Finally, we constructed a novel weight balance loss to focus on the semantic feature of small objection, solving the unbalance problem of tumor and background. Specifically, the extensive experiments conducted on Dataset BUSI and Dataset B demonstrated that the T ²KD Attention U-Net outperformed various knowledge distillation counterparts. Concretely, the accuracy, recall, precision, Dice, and mIoU of proposed method were 95.26%, 86.23%, 85.09%, 83.59%and 77.78% on Dataset BUSI, respectively. And these performance indexes were 97.95%, 92.80%, 88.33%, 88.40% and 82.42% on Dataset B, respectively. Compared with other models, the performance of this model was significantly improved. Meanwhile, compared with the teacher model, the number, size, and complexity of student model were significantly reduced (2.2×10 ⁶ vs. 106.1×10 ⁶, 8.4 MB vs. 414 MB, 16.59 GFLOPs vs. 205.98 GFLOPs, respectively). Indeedy, the proposed model guarantees the performances while greatly decreasing the amount of computation, which provides a new method for the deployment of clinical medical scenarios.
Humans ; Breast Neoplasms/diagnostic imaging* ; Female ; Ultrasonography, Mammary/methods* ; Image Processing, Computer-Assisted/methods* ; Algorithms ; Neural Networks, Computer ; Breast/diagnostic imaging*

Lung cancer is the leading cause of cancer-related deaths worldwide. Radiation pneumonitis is a major complication in lung cancer radiotherapy. Four-dimensional computed tomography (4DCT) imaging provides dynamic ventilation information, which is valuable for lung function assessment and radiation pneumonitis prevention. Many methods have been developed to calculate lung ventilation from 4DCT, but a systematic comparison is lacking. Prediction of radiation pneumonitis using 4DCT-based ventilation is still in an early stage, and no comprehensive review exists. This paper presented the first systematic comparison of functional lung ventilation algorithms based on 4DCT over the past 15 years, highlighting their clinical value and limitations. It then reviewed multimodal approaches combining 4DCT ventilation imaging, dose metrics, and clinical data for radiation pneumonitis prediction. Finally, it summarized current research and future directions of 4DCT in lung cancer radiotherapy, offering insights for clinical practice and further studies.
Humans ; Lung Neoplasms/diagnostic imaging* ; Four-Dimensional Computed Tomography/methods* ; Radiation Pneumonitis/etiology* ; Algorithms ; Lung/radiation effects* ; Pulmonary Ventilation

Colorectal polyps are important early markers of colorectal cancer, and their early detection is crucial for cancer prevention. Although existing polyp segmentation models have achieved certain results, they still face challenges such as diverse polyp morphology, blurred boundaries, and insufficient feature extraction. To address these issues, this study proposes a parallel coordinate fusion network (PCFNet), aiming to improve the accuracy and robustness of polyp segmentation. PCFNet integrates parallel convolutional modules and a coordinate attention mechanism, enabling the preservation of global feature information while precisely capturing detailed features, thereby effectively segmenting polyps with complex boundaries. Experimental results on Kvasir-SEG and CVC-ClinicDB demonstrate the outstanding performance of PCFNet across multiple metrics. Specifically, on the Kvasir-SEG dataset, PCFNet achieved an F1-score of 0.897 4 and a mean intersection over union (mIoU) of 0.835 8; on the CVC-ClinicDB dataset, it attained an F1-score of 0.939 8 and an mIoU of 0.892 3. Compared with other methods, PCFNet shows significant improvements across all performance metrics, particularly in multi-scale feature fusion and spatial information capture, demonstrating its innovativeness. The proposed method provides a more reliable AI-assisted diagnostic tool for early colorectal cancer screening.
Humans ; Colonic Polyps/diagnostic imaging* ; Colorectal Neoplasms/diagnostic imaging* ; Neural Networks, Computer ; Image Processing, Computer-Assisted/methods* ; Algorithms ; Early Detection of Cancer

Endometrial cancer (EC) is one of the most common gynecological malignancies, with an increasing incidence rate worldwide. Accurate segmentation of lesion areas in computed tomography (CT) images is a critical step in assisting clinical diagnosis. In this study, we propose a novel deep learning-based segmentation model, termed spatial choice and weight union network (SCWU-Net), which incorporates two newly designed modules: the spatial selection module (SSM) and the combination weight module (CWM). The SSM enhances the model's ability to capture contextual information through deep convolutional blocks, while the CWM, based on joint attention mechanisms, is employed within the skip connections to further boost segmentation performance. By integrating the strengths of both modules into a U-shaped multi-scale architecture, the model achieves precise segmentation of EC lesion regions. Experimental results on a public dataset demonstrate that SCWU-Net achieves a Dice similarity coefficient (DSC) of 82.98%, an intersection over union (IoU) of 78.63%, a precision of 92.36%, and a recall of 84.10%. Its overall performance is significantly outperforming other state-of-the-art models. This study enhances the accuracy of lesion segmentation in EC CT images and holds potential clinical value for the auxiliary diagnosis of endometrial cancer.
Humans ; Endometrial Neoplasms/diagnostic imaging* ; Female ; Tomography, X-Ray Computed/methods* ; Deep Learning ; Algorithms ; Image Processing, Computer-Assisted/methods* ; Neural Networks, Computer

Thyroid nodules are a common endocrine disorder, and their early detection and accurate diagnosis are crucial for the prevention of thyroid cancer. However, the highly heterogeneous morphology and boundaries of thyroid nodules pose significant challenges to their precise identification and classification. Traditional diagnostic approaches rely heavily on physicians' experience, which increases the risk of misdiagnosis and missed diagnoses. With the rapid advancement of computer-aided diagnosis (CAD) technologies, applying deep learning algorithms to the analysis of thyroid nodule ultrasound images has shown great potential. This paper reviews the latest research progress on deep learning-based CAD methods for thyroid nodules, with a focus on their applications in image preprocessing, segmentation and classification. The advantages and limitations of current techniques are analyzed, and potential future directions are discussed. This review aims to highlight the potential of deep learning in thyroid nodule diagnosis and to provide a foundation for selecting feasible pathways for future clinical applications.
Humans ; Thyroid Nodule/diagnostic imaging* ; Deep Learning ; Ultrasonography/methods* ; Diagnosis, Computer-Assisted/methods* ; Algorithms ; Thyroid Neoplasms/diagnostic imaging* ; Image Processing, Computer-Assisted/methods*

We propose a strategy to reduce unnecessary prostate biopsies in Chinese patients with total prostate-specific antigen (tPSA) >10 ng ml -1 and Prostate Imaging Reporting and Data System (PI-RADS) scores between 1 and 3. Clinical data derived from 517 patients of The First Affiliated Hospital of USTC (Hefei, China) from January 2020 to December 2023 who met the screening criteria for the study were retrospectively collected. Independent predictors were identified via univariate and multivariate logistic regression analysis. The diagnostic capacity of clinical variables was evaluated using the receiver operating characteristic (ROC) curves and area under the curve (AUC). A prostate biopsy strategy was developed via risk stratification. Of the 517 patients, 17/348 (4.9%) with PI-RADS 1-2 were diagnosed with clinically significant prostate cancer (csPCa), and 27/169 (16.0%) patients with PI-RADS 3 were diagnosed with csPCa. The appropriate prostate-specific antigen density (PSAD) cut-off values were 0.45 ng ml -2 for PI-RADS 1-2 patients and 0.3 ng ml -2 for PI-RADS 3 patients. The appropriate prostate volume (PV) cut-off values were 40 ml for PI-RADS 1-2 patients and 50 ml for PI-RADS 3 patients. The prostate biopsy strategy based on PSAD and PV developed in this study can reduce unnecessary prostate biopsies in patients with tPSA >10 ng ml -1 and PI-RADS 1-3. In the study, 66.5% (344/517) patients did not need to undergo prostate biopsy, at the expense of missing only 1.7% (6/344) patients with csPCa.
Humans ; Male ; Prostatic Neoplasms/diagnostic imaging* ; Prostate-Specific Antigen/blood* ; Aged ; Middle Aged ; Retrospective Studies ; Prostate/diagnostic imaging* ; Unnecessary Procedures/statistics & numerical data* ; Biopsy/statistics & numerical data* ; China ; ROC Curve

The choice of biopsy method is critical in diagnosing prostate cancer (PCa). This retrospective cohort study compared systematic biopsy (SB) or cognitive fusion-targeted biopsy combined with SB (CB) in detecting PCa and clinically significant prostate cancer (csPCa). Data from 2572 men who underwent either SB or CB in Fudan University Shanghai Cancer Center (Shanghai, China) between January 2019 and December 2023 were analyzed. Propensity score matching (PSM) was used to balance baseline characteristics, and detection rates were compared before and after PSM. Subgroup analyses based on prostate-specific antigen (PSA) levels and Prostate Imaging-Reporting and Data System (PI-RADS) scores were performed. Primary and secondary outcomes were the detection rates of PCa and csPCa, respectively. Of 2572 men, 1778 were included in the PSM analysis. Before PSM, CB had higher detection rates for both PCa (62.9% vs 52.4%, odds ratio [OR]: 1.54, P < 0.001) and csPCa (54.9% vs 43.3%, OR: 1.60, P < 0.001) compared to SB. After PSM, CB remained superior in detecting PCa (63.1% vs 47.9%, OR: 1.86, P < 0.001) and csPCa (55.0% vs 38.2%, OR: 1.98, P < 0.001). In patients with PSA 4-12 ng ml -1 (>4 ng ml -1 and ≤12 ng ml -1 , which is also applicable to the following text), CB detected more PCa (59.8% vs 40.7%, OR: 2.17, P < 0.001) and csPCa (48.1% vs 27.7%, OR: 2.42, P < 0.001). CB also showed superior csPCa detection in those with PI-RADS 3 lesions (32.1% vs 18.0%, OR: 2.15, P = 0.038). Overall, CB significantly improves PCa and csPCa detection, especially in patients with PSA 4-12 ng ml -1 or PI-RADS 3 lesions.
Humans ; Male ; Prostatic Neoplasms/diagnosis* ; Propensity Score ; Retrospective Studies ; Middle Aged ; Aged ; Image-Guided Biopsy/methods* ; Prostate-Specific Antigen/blood* ; Prostate/diagnostic imaging*

Accurate staging is the fundamental basis for the treatment and prognosis of non-small cell lung cancer (NSCLC), and whether the tumor involves the pleura or chest wall is a critical aspect in assessing the staging of peripheral lung cancer. Imaging techniques such as computed tomography (CT), magnetic resonance imaging (MRI), ultrasound (US) and positron emission tomography (PET) are widely used to determine pleural invasion in NSCLC. There has been an increasing number of studies evaluating whether NSCLC invades the pleura and the extent of such invasion. This article provides a review of the staging and the imaging diagnostic criteria of pleural invasion, aiming to offer references for peers in the precise diagnosis of pleural or chest wall invasion.
.
Humans ; Carcinoma, Non-Small-Cell Lung/diagnosis* ; Lung Neoplasms/diagnosis* ; Thoracic Wall/diagnostic imaging* ; Pleura/diagnostic imaging* ; Neoplasm Invasiveness ; Tomography, X-Ray Computed

Lung cancer is still one of the most common malignant tumors in the world. With the increase of its incidence and the development of medical technology, the overall survival of lung cancer patients has significantly extended compared to before. The incidence of brain and meningeal metastases from lung cancer has also been rising year by year, but patients with brain and meningeal metastases from lung cancer have a poor prognosis and a very high mortality rate, and the diagnosis is mainly based on computed tomography (CT), magnetic resonance imaging (MRI) and other imaging examinations. However, the imaging features are diverse and the specificity is low, which makes it easy to be misdiagnosed and missed. Therefore, accurately identifying brain and meningeal metastases and timely targeted treatment is crucial for improving patient prognosis. This paper analyzed the diagnosis and treatment of a case of lung cancer with no obvious recurrence and metastasis in nearly 7-year long-term follow-up after radical lung cancer surgery, but the patient with abnormal behavior, impaired consciousness and epilepsy in the past 5 months, and multiple punctate calcifications in the brain found by head CT and MRI. This paper consider that the patient's mental and behavioral symptoms were caused by brain and meningeal metastasis of lung cancer after excluding infectious disease and ineffective treatment of autoimmune encephalitis, and further pathological biopsy and genetic detection confirmed the diagnosis of metastatic lung adenocarcinoma with epidermal growth factor receptor (EGFR) L858R gene mutation, and the patient's symptoms were significantly improved after targeted therapy by Osimertinib. This paper also searched the relevant literatures of brain calcifications in databases such as China National Knowledge Infrastructure (CNKI), Wanfang, UpToDate, PubMed, etc., and found that intracerebral calcifications exist in a variety of diseases, including infectious, genetic and neurodegenerative diseases, vascular diseases, metabolic diseases and tumors. However, brain calcification in brain and meningeal metastases are often underestimated, and the consequent risk is misdiagnosis and delayed treatment. Therefore, brain and meningeal metastases manifested as brain calcification should not be ignored in patients with a history of previous tumors.
.
Humans ; Lung Neoplasms/pathology* ; Brain Neoplasms/diagnostic imaging* ; Meningeal Neoplasms/diagnostic imaging* ; Calcinosis/diagnostic imaging* ; Male ; Middle Aged ; Tomography, X-Ray Computed ; Magnetic Resonance Imaging