Clinical trials

This study aims to develop a multimodal deep learning model integrating MRI, ultrasound, digital pathology and clinical information based on multicenter retrospective data. To externally validate the model in an independent prospective cohort, and evaluate its accuracy in predicting pathological complete response (pCR), 3-year and 5-year disease-free survival (DFS). To establish visual tools such as nomograms, assisting clinicians in identifying patients with chemoresistance and facilitating individualized de-escalation or escalation treatment strategies.

We consecutively enrolled patients presenting with breast space-occupying lesions at our hospital. Medical imaging data from simultaneously acquired ultrafast DCE-MRI, time-dependent diffusion MRI (TDD-MRI), magnetic resonance elastography (MRE), andvirtual magnetic resonance elastography(vMRE), along with clinical and pathological data, were collected. The study aims to evaluate and explore the clinical application value of these multimodal imaging techniques in the precision diagnosis, treatment, and prognosis of breast cancer. The findings are intended to provide a robust imaging basis for formulating and adjusting more precise, individualized treatment plans and prognostic assessments for patients. Ultimately, this research seeks to improve patient quality of life, optimize the allocation of medical resources, and advance the development of precision medicine.

This study aims to construct a multi-task deep learning model system to mine deep features in DBT images, so as to achieve accurate detection of breast lesions, differential diagnosis of benign and malignant (especially for the challenging BI-RADS 4A category), prediction of molecular subtypes, and evaluation of neoadjuvant chemotherapy (NAC) efficacy, providing an imaging basis for precision medicine.

Breast malignant tumors are a serious threat to women's health, and the current treatment for breast malignant tumors is still dominated by surgery, but the problems faced by patients after surgery such as edema and tumor recurrence are still relatively common, and the high recurrence rate and the occurrence of postoperative complications are closely related to the removal of metastatic lymph nodes during surgery. Therefore, accurate assessment of sentinel lymph node (SLN) metastases is essential to determine the extent of axillary lymph node dissection (ALND) and minimize complications. In this study, we developed a new technique that can rapidly distinguish between metastatic lymph nodes and normal sentinel lymph nodes (SLNs) in breast cancer patients. Briefly, fresh tissue is incubated with the probe and imaged immediately after intraoperative sentinel lymph node resection to identify the metastatic status of SLN. The accuracy of fluorescence imaging is confirmed by pathological diagnosis.

For early breast cancer patients, breast-conserving surgery and sentinel lymph node biopsy have emerged as the cornerstone of precision oncology, harmonizing oncologic efficacy with the preservation of breast cosmetics and axillary function. However, conventional margin evaluation relies on subjective experience, leading to positive margin rates of 15%-50% Additionally, sentinel lymph node biopsy faces inherent technical limitations, including prolonged procedural time and significant false-negative rates in detecting micro-metastases. To overcome these challenges, the investigators will introduce a stable "sandwich" EpCAM-targeted surface-enhanced Raman spectroscopy (SERS) detection system to provide rapid evaluation of surgical margins and sentinel lymph node (SLN) metastasis status. In brief, a portion of resected tissue will be immediately homogenized by grinding, and the tissue homogenate will then be analyzed using this detection system to determine its histological characteristics. Pathological confirmation will further validate the method. This approach aims to improve the accuracy and efficiency of intraoperative assessment of breast-conserving surgical margins and SLN status, thereby effectively reducing local recurrence rates and complications.

Bladder cancer ranks as the fourth most common malignancy among males the United States . Approximately 75% of patients present with non-muscle-invasive bladder cancer (NMIBC). For the diagnosis and treatment of NMIBC, current guidelines widely recommend white light cystoscopy (WLC) and transurethral resection of the bladder tumor (TURBT). Unfortunately, up to 70% of patients with NMIBC experience intravesical recurrence within five years of their initial treatment. The high recurrence rates necessitate long-term surveillance for most NMIBC patients, making it one of the most costly malignancies to manage. In fact, a higher risk of disease recurrence is also associated with the now widely used WLC and TURBT, which cause false-negative investigations with an inadequate resection or residual tumor, especially when urothelial tumors present as carcinoma in situ or multiple. Indocyanine green (ICG)-based the second near-infrared (NIR-II) fluorescence imaging offers real-time visualization during surgery, potentially reducing residual tumor. Herein, the investigators will introduce a novel NIR-II probe, TTP-ICG, based on a Trop2-targeting peptide (TTP) and an approach which enables differentiation between cancer and para-cancer . In brief, tissues will be soaked in TTP-ICG after resection, and their histological characterization will be determined under NIR-II fluorescence imaging. Pathological confirmation will further validate our approach. To explore the conditions for the future in vivo real-time identification of NMIBC during NIR-II fluorescence -guided surgery.

Spectral CT was used to prospectively collect medical images and clinical data related to colorectal cancer, evaluate the effect of image quality in the diagnosis of colorectal cancer, and evaluate the application value in the accurate staging of colorectal cancer, so as to provide a more accurate clinical basis for diagnosis, promote the development of individualized treatment, and ultimately improve the prognosis of patients.

Spectral CT was used to prospectively collect medical images and clinical data related to liver cancer, evaluate the effect of image quality in the diagnosis of liver cancer, and evaluate the application value in the accurate staging of liver cancer, so as to provide a more accurate clinical basis for diagnosis, promote the development of individualized treatment, and ultimately improve the prognosis of patients.

Spectral CT was used to prospectively collect medical images and clinical data related to gastric cancer, evaluate the effect of image quality in the diagnosis of gastric cancer, and evaluate the application value in the accurate staging of gastric cancer, so as to provide a more accurate clinical basis for diagnosis, promote the development of individualized treatment, and ultimately improve the prognosis of patients.

Spectral CT was used to prospectively collect medical images and clinical data related to breast cancer, evaluate the effect of image quality in the diagnosis of breast cancer, and evaluate the application value in the accurate staging of breast cancer, so as to provide a more accurate clinical basis for diagnosis, promote the development of individualized treatment, and ultimately improve the prognosis of patients.

To prospectively collect medical images and clinical data related to esophageal cancer using 4D Free Breathing DCE-MRI technology, to evaluate the image quality in the diagnosis of esophageal cancer, and to predict the efficacy of preoperative neoadjuvant chemoradiotherapy for esophageal cancer.

This study aims to address the challenge of predicting the efficacy of targeted and immunotherapy in lung cancer patients. By utilizing two DCE-MRI images-one before treatment and one after one treatment cycle-along with the patients' clinical and pathological information, we aim to achieve accurate prediction of the therapeutic efficacy of targeted and immunotherapy in lung cancer patients.

Accurate staging is critical for effective gastric cancer treatment planning. Conventional DWI (C-DWI) has limitations in image quality due to magnetic field inhomogeneity, which hampers precise T staging. This study aims to compare the image quality and T staging accuracy of C-DWI, TSE-DWI, and ZOOMit-DWI (Z-DWI) sequences in MRI examinations of gastric cancer patients. A prospective study plan to enroll 72 gastric cancer patients undergoing preoperative MRI with C-DWI, TSE-DWI, and Z-DWI sequences using a 3.0 T scanner. Quantitative metrics (ADC, SNR, CNR) and image quality were evaluated. T staging accuracy was assessed by comparing MRI results with postoperative pathological staging using ROC analysis.

The purpose of this study is to quantitatively assess the changes in liver fat content in cancer patients before and after treatment. The main questions it aims to answer are:How does the liver fat fraction change before and after chemotherapy? In this study, patients undergoing mDixon Quant scanning are subjected to fully automated segmentation and measurement of liver fat content using artificial intelligence.

Accurate evaluation of tumor boundaries in breast-conserving surgery is closely associated with reducing the second operation of patients. Meanwhile, accurately assessing sentinel lymph node (SLN) metastasis is crucial for determining the extent of axillary lymph node dissection (ALND) and minimizing complications. Near-infrared-II (NIR-II) fluorescence imaging using molecular agents has shown promise for in situ imaging during resection. However, very effective probes can be applied to clinical trials up to now, which limits the clinical application of fluorescence imaging. Here we developed a new technology that can quickly differentiation between cancer and para-cancer tissue as well as metastatic and normal sentinel lymph nodes(SLN). In brief, the fresh tissues were incubated with the probe immediately after intraoperative resection and imaged to identify the tumor area and distinguish the metastatic status of SLN. The accuracy of fluorescence imaging was confirmed by pathological diagnosis.

The goal of this observational study is to develop an artificial intelligence model to transform unstained lymph node tissue slice images directly into stained images. The main questions it aims to answer are: Can the virtual staining model generate hematoxylin and eosin (H\&E) and immunohistochemistry (IHC) images suitable for clinical diagnosis from unstained paraffin-embedded lymph node slice images, including those from breast axillary lymph nodes and other tumor lymph nodes? Can the virtual staining model generate H\&E and IHC images suitable for clinical diagnosis from unstained frozen sentinel lymph node slice images from breast cancer patients? Researchers will retrospectively collect paraffin-embedded lymph node slices from tumor patients and prospectively collect frozen sentinel lymph node slices from breast cancer patients.

The goal of this observational study is to develop an artificial intelligence model based on pathology and magnetic resonance imaging (MRI) images to predict the metastatic status of non-sentinel lymph nodes in patients with breast cancer sentinel lymph node metastasis. The main questions it aims to answer are: Can an artificial intelligence model based on MRI images of breast cancer patients predict the non-sentinel lymph node metastatic status in patients with breast cancer sentinel lymph node metastasis? Can an artificial intelligence model based on intraoperative frozen section images of sentinel lymph nodes in breast cancer patients predict the non-sentinel lymph node metastasis status in patients with sentinel lymph node metastasis from breast cancer? Can artificial intelligence models based on preoperative MRI and intraoperative frozen section images of sentinel lymph nodes in breast cancer patients predict the non-sentinel lymph node metastatic status in patients with sentinel lymph node metastasis from breast cancer? Researchers will retrospectively and prospectively collect preoperative MRI and intraoperative sentinel lymph node section images from breast cancer patients.

The goal of this clinical trial is to evaluate the value of rapid on-site evaluation（ROSE） in lymph node biopsy of breast cancer. The main question it aims to answer is:The ROSE technique was applied to breast cancer lymph node biopsy to improve the diagnostic efficiency and accuracy. Participants underwent routine lymph node biopsy, and the test personnel conducted routine disease examination and rose technique evaluation on the removed ly

This study is a single arm, single center, phase II, prospective clinical study aimed at exploring the effectiveness and safety of watch and wait strategy guided by dynamic minimal disease residual (MRD) monitoring to achieve clinical complete response after neoadjuvant immunotherapy for locally advanced colorectal cancer with deficient mismatch repair/ microsatellite instability-high (dMMR/MSI-H).