Clinical Research Directory

An Automated Virtual Reality Intervention to Enhance Firearm Safety Counseling in Pediatrics

Clinicians will complete a screen-based automated virtual reality training intervention to learn how to counsel caregivers of pediatric patients on secure firearm storage.

AI-Enhanced Analysis of Breast Density and Background Parenchymal Enhancement (BPE)

This study expands upon previous research investigating the correlation between breast density, Background Parenchymal Enhancement (BPE), and age in contrast-enhanced mammography (CEM). By integrating Artificial Intelligence (AI) methodologies, including Artificial Neural Networks (ANNs) and deep learning models, the study aims to optimize the accuracy of predictions and validate prior findings obtained through multiple linear regression.

AI Ambient Scribe in Psychiatry

The goal of this clinical trial is to learn if an artificial intelligence (AI) tool that listens to appointments and drafts clinical notes can help clinicians in psychiatric care. Researchers also want to know if the tool changes how patients experience their care. The study includes mental health care providers (physicians, psychologists/psychotherapists, nurses, and social workers) at Psychiatrie St. Gallen and the adult patients they see in their appointments. The main questions it aims to answer are: * Does the AI tool lower the time clinicians spend writing notes? * Are notes written with the AI tool as good as or better than notes written without it? * Do patients notice a difference in their relationship with their clinician when the tool is used? Researchers will compare clinicians who use the AI tool to clinicians who keep writing notes their usual way, before and during the intervention period, to see if writing time, note quality, and the patient-rated therapeutic relationship differ between the two groups. Clinicians will: Join one of two groups. One group starts using the AI tool on August 17, 2026 (7 clinicians whose devices were delivered late start on November 9, 2026). The other group keeps writing notes the usual way. Both groups will be observed until the end of November 2026 (the 7 late-starting clinicians and additional control data collection until February 22, 2027), with note quality evaluation in February and March 2027. Clinicians who cannot contribute to all measurements may, with their consent, remain in the study for automated measurement of documentation time only. Patients will: \- Be invited to fill out a short, anonymous survey about their experience with their clinician.

Prospective User Study and Clinical Validation of Multimodal Medical Imaging Large Models

Following model development and locking, the fixed model is evaluated in prospectively collected CT cohorts from a single center. The study is observational and does not affect clinical care. A subset of cases is used in a randomized crossover reader study.

Safety, Feasibility, and Acceptability of the MINERVA Voice Conversational Agent for Emergency Department History-Taking

This study evaluates whether an artificial intelligence (AI) conversational agent called MINERVA (Medical Interface for Navigating the Emergency Room with Voice Assistance) can safely and feasibly gather medical histories from emergency department (ED) patients and produce accurate summaries. The main questions it aims to answer are: Can an AI conversational agent gather and summarize patient histories in the emergency department? Are the resulting summaries accurate? Do they capture essential clinical information? Can an AI agent overseeing these conversations independently identify red-flag symptoms that may need time-sensitive clinical interventions? Adult ED patients interact with MINERVA via either voice chat or typing on a keyboard. MINERVA asks them questions about their symptoms and generates a summary which the patients review for accuracy. MINERVA reviews the conversation for clinical red flags that suggest urgent evaluation is required. Participants fill out a survey providing feedback on their experience. The summary is not saved in the medical record and clinical care is not affected. Physician reviewers will evaluate the safety, accuracy, and clinical value of the MINERVA interview transcripts and summaries using standardized rubrics, including whether MINERVA was correct in identifying red-flag symptoms that may need time-sensitive clinical interventions.

AI-Generated Video Distraction Via Virtual Reality for Fear and Pain During Pediatric Circumcision

This randomized controlled trial evaluated the effect of artificial intelligence-generated video distraction delivered via virtual reality on procedural fear and pain in children undergoing circumcision. A total of 80 children aged 4-12 years were assigned to an intervention group (n=40) or a control group (n=40). Children in the intervention group received AI-generated video distraction via virtual reality, whereas children in the control group received routine care. Pain and emotional responses were assessed from the perspectives of the child, parent, and nurse. Physiological parameters were also measured before and after the procedure. The study aimed to determine whether AI-generated video distraction delivered via virtual reality reduces procedural fear and pain during pediatric circumcision.

A Pragmatic Trial of AI-Assisted Discharge Summary Generation

The goal of this study is to learn whether an artificial intelligence (AI) tool built into the electronic health record helps hospital physicians write discharge summaries. A discharge summary is the document that tells a patient's family doctor and other providers what happened during a hospital stay. The main questions this study aims to answer are: Does access to the AI tool reduce the time physicians spend writing discharge summaries? Does it change the quality of discharge summaries or how soon they are completed? How does it affect physician workload? Physicians will be randomly assigned to either have access to the AI tool or complete discharge summaries as they usually do. Physicians with access can choose whether to use the tool, and they review and edit any AI-generated text before it becomes part of the discharge summary.

Development of a Multimodal AI System for GIST Management

Background: Gastrointestinal Stromal Tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal tract. Accurate pre-operative diagnosis, risk stratification, and genotyping are critical for determining the appropriate surgical approach and targeted therapy (such as Imatinib). However, current methods often rely on invasive postoperative pathology and expensive genetic testing. Study Objective: The purpose of this study is to develop and validate a multimodal Artificial Intelligence (AI) model that integrates clinical data, CT radiomics (imaging features), and pathomics (digital pathology features) to improve the precision of GIST management. Study Design: This is a prospective, observational study. The researchers will recruit patients with suspected gastric submucosal tumors who are scheduled for surgery or biopsy at The Fourth Hospital of Hebei Medical University. Core Tasks: The AI model will be trained to perform three specific tasks: Diagnosis: Distinguish GISTs from other non-GIST mesenchymal tumors (e.g., leiomyomas, schwannomas). Risk Assessment: Stratify GISTs into risk categories (e.g., Low vs. High risk) to predict malignant potential. Genotyping: Predict specific gene mutations (e.g., KIT or PDGFRA mutations) to guide immunotherapy or targeted therapy. Methodology: Patient data (CT scans, pathology slides, and clinical history) will be collected and analyzed by the AI system. The AI's predictions will be compared against the "Gold Standard" results derived from postoperative pathological examination and Next-Generation Sequencing (NGS). This study is non-interventional; the AI results will not affect the standard of care received by the patients.

Large Linguistic Model for Clinical Reaoning of Physical Therapy Students

Clinical reasoning is a fundamental skill for physical therapy students, enabling them to collect and interpret patient information to make accurate diagnoses and treatment decisions. Traditional training methods often limit students' exposure to a diverse range of clinical cases, which can restrict the development of these skills. The integration of Large Language Models (LLMs), such as ChatGPT, into physical therapy education offers a novel approach to enhance clinical reasoning by simulating interactive and realistic patient scenarios. This randomized controlled trial aims to evaluate the effectiveness of an LLM-based educational intervention in improving clinical reasoning skills in physical therapy students. The study will recruit a total of 200 third-year physiotherapy students from multiple university institutions. Participants will be randomly assigned to one of two groups: 1. Experimental Group - Students will receive LLM-based training, engaging with a conversational artificial intelligence model to solve clinical cases over an 4-week period. The model will provide real-time responses to their questions, allowing them to refine their diagnostic and treatment reasoning. 2. Control Group - Students will follow the standard curriculum, participating in conventional case-based learning and supervised clinical reasoning exercises without AI-based assistance. The primary outcome of the study is the improvement in clinical reasoning skills, assessed through standardized written case evaluations and structured practical examinations. Secondary outcomes include changes in digital competence, student engagement levels, overall satisfaction with the educational approach, and cost-effectiveness of the intervention. By assessing the impact of LLMs on clinical reasoning training, this study seeks to determine whether AI-driven educational tools can effectively complement traditional physiotherapy education and improve student preparedness for real-world clinical practice.

Development and Demonstration of a Precision Prevention and Treatment System for Colorectal Tubular Adenomas Lesions

This study is a prospective,multi-center and observational clinical study.Investigators would like to innovatively construct a "trinity" database of colorectal tubular adenomas based on white light - magnifying chromo - pathological images.It simulates the decision - making logic of doctors, and based on the multimodal endoscopic LAFEQ method previously proposed, develop a multimodal deep - learning diagnostic model for colon adenomas and an interpretable risk prediction model for intestinal adenomas. While achieving high - precision auxiliary treatment decisions, clearly present the decision - making basis, and break through the limitation of poor interpretability of previous medical imaging AI models.

Effects of Agent-assisted, LLM-assisted and Traditional Workflows on Diagnosis and Management Planning at Admission

The goal of this clinical trial is to evaluate whether AI-assisted workflows improve physicians' admission diagnosis and management planning performance on standardized simulated inpatient cases, among practicing internal medicine and surgery physicians across all seniority levels and across three tiers of the Chinese healthcare system. The main questions it aims to answer are: * Does the Agent-assisted workflow yield better structured admission diagnosis and management planning scores than standalone LLM assistance? * Does the Agent-assisted workflow outperform the traditional workflow without AI tools? Researchers will compare three parallel groups (traditional workflow group, LLM-assisted group, Agent-assisted group) to determine whether the Agent tool can improve diagnostic accuracy and efficiency. The goal of this clinical trial is to evaluate whether AI-assisted workflows improve physicians' admission diagnosis and management planning performance on standardized simulated inpatient cases, among practicing internal medicine and surgery physicians across all seniority levels and across three tiers of the Chinese healthcare system. The main questions it aims to answer are: * Does the Agent-assisted workflow yield better diagnosis and management planning scores than the LLM-assisted workflow? * Does the Agent-assisted workflow yield better diagnosis and management planning scores than the conventional workflow without AI assistance? Researchers will compare three parallel groups (traditional workflow group, LLM-assisted group, Agent-assisted group) to determine whether the agent tool can improve diagnostic and management planning performance and efficiency. Participants will: * Be recruited from 15 hospitals in China and participate remotely under video proctoring * Be randomly assigned to one of the three fixed workflows, with randomization stratified by hospital tier, specialty and seniority * Work through 6 anonymized simulated HIS admission cases and complete as many as they can. * Submit structured answers for each case covering principal diagnosis, secondary diagnoses, differential diagnoses, diagnostic justification, next diagnostic or therapeutic steps, consultation and referral decisions, and diagnostic confidence. * Have their operation logs and time consumption recorded automatically by the study platform. * Complete a post-session questionnaire after submitting all cases.

Impact of Artificial Intelligence on Trainee Polyp Miss Rates

Based on prior studies, trainee and practicing gastroenterologists miss pre-cancerous polyps (adenomas and serrated polyps) during colonoscopy. The use of computer-aided detection (CADe) systems, a form of artificial intelligence (AI) has been shown to help identify colorectal lesions for practicing gastroenterologists. However, less is known how AI impacts polyp detection for trainees. The investigators are conducting a tandem colonoscopy study wherein a portion of the colon is examined first by the trainee and then the attending physician. For each procedure, randomization will occur which will determine whether or not the trainee will utilize AI for their examination of the colon. At the end of the study, the investigators will determine whether AI helps trainees miss fewer polyps during colonoscopy. The investigators will also conduct interviews with trainees to understand how AI impacts colonoscopy training.

Trust in AI and Digitalization in Healthcare Across Generational Groups

This study aims to investigate differences in perception of barriers and facilitators of digitalization and Artificial Intelligence (AI) usage in healthcare across different generational groups (youth, working-age adults, and seniors). The results will help create practical recommendations for public health projects and consultants to support fair and inclusive use of new digital tools in healthcare. A cross-sectional online survey will be conducted among students at HAW, patients and employees in the rehabilitation center in Oldenburg, and seniors participating in the "Digital im Alter"(DIA) project.

Teaching Doctors in Training to Reason With Artificial Intelligence

The purpose of the TEACH-AI study is to assess whether a brief, structured workshop on artificial intelligence can improve the performance of medicine doctors in training (i.e. residents) in their diagnostic and management reasoning. In this multi-site randomized controlled trial, internal medicine and family medicine residents are assigned either to receive an in-person workshop on safe, effective LLM use before a standardized AI-assisted assessment, or to complete the same assessment before receiving the workshop. Residents will review clinical cases that are fully synthetic, no protected health information is used, using a password protected LLM interface.

Effect of AI-Supported Case Analysis on Nursing Students

The aim of this study is to determine the effect of AI-supported oncology case analysis on nursing students' knowledge, level of learning satisfaction, and clinical decision-making skills. This study is planned to be conducted using a single-blind randomized controlled trial design for the quantitative research component and an interview design for the qualitative research component. The students will be divided into two groups: an intervention group (artificial intelligence) and a control group (traditional instruction).

LLM-Generated Medical Summaries During Hospitalization to Improve Patient Activation in High-Risk Pregnancies

The goal of this clinical trial is to learn if AI-generated medical summaries help hospitalized pregnant participants better understand and actively participate in their own care. The main questions it aims to answer are: Do AI-generated summaries increase participant activation (knowledge, skills, and confidence in managing their health)? How accurate are the AI-generated summaries, as assessed by physicians? Do AI-generated summaries affect depressive symptoms during hospitalization? Researchers will compare participants who receive periodic AI-generated summaries of their medical care with participants who receive standard care to determine whether the summaries improve patient activation during hospitalization. Participants will: Be randomly assigned to receive AI-generated medical summaries or standard care. Complete the Patient Activation Measure (PAM-13) at baseline before the intervention and again after the intervention. Complete the Edinburgh Postnatal Depression Scale (EPDS) at baseline and again after the intervention. Receive up to 3 summaries during their hospitalization (intervention group only).

A Study of the Correlation Between the Severity of Substance Use Disorder and the Intensity of Dependence on Generative Artificial Intelligence

This bicentric, cross-sectional observational study conducted in France evaluates the relationship between substance use disorder (SUD) severity and generative artificial intelligence dependency among outpatients treated in specialized addiction care centers (CSAPA). While conversational generative artificial intelligence tools have seen rapid widespread adoption, potential problematic usage and cognitive dependency remain poorly documented in clinical addictology. Outpatients followed for substance use disorders present shared cognitive, reward-processing, and behavioral vulnerabilities that may heighten their susceptibility to emerging digital dependencies. Eligible adult patients complete a single 15-minute evaluation comprising the Generative Artificial Intelligence Dependency Scale (GAIDS; 11 items rated on a 5-point Likert scale from 1 to 5, total score range: 11 to 55) and the DSM-5 diagnostic criteria checklist for their primary substance of abuse, alongside sociodemographic characteristics. Clinical data, including documented psychiatric comorbidities, are extracted in parallel from electronic health records. Following questionnaire completion, participants receive a dedicated debriefing and clinical restitution interview with an investigator. The primary objective is to evaluate the linear correlation between SUD severity (number of validated DSM-5 criteria, from 0 to 11) and generative artificial intelligence dependency intensity (total raw GAIDS score). Secondary objectives aim to describe generative artificial intelligence dependency levels across specific primary substance classes (alcohol, tobacco, cannabis, cocaine, opioids, etc.), documented comorbid psychiatric disorders (e.g., mood disorders, ADHD, anxiety, personality disorders), and sociodemographic subgroups (age brackets, sex, education, and occupational status).

Identification of Depressive and Anxiety Symptoms Among a Sample of Emergency Department Patients Using Artificial Intelligence (AI) Technology

Behavioral health problems, such as depression and anxiety, are common yet often are not identified by emergency department doctors and nurses. These mental health conditions can be due to medical issues or can worsen medical problems. One way investigators hope to do a better job of learning about mental health is by training Artificial Intelligence (AI) software to detect anxiety and depression by analyzing facial expression and tone of voice. Participants are invited to participate in a study which may help improve emergency department care. An audio and video recording of the participant's responses to some simple, non-psychological questions will be analyzed by a computer to determine whether investigators can assess mood and anxiety by analyzing speech and visual patterns. The audio and video will not be listened to nor watched by study personnel, only analyzed by a computer. The investigator's hope is that it will help others in the future by aiding in the assessment of psychological state. This study is being conducted at CMC ED only.

AI-Assisted Smart Interactive Healthcare Robot

To address workforce shortages and increasing workloads in nursing, technological solutions and AI-powered robots for ward navigation have been introduced. However, limitations remain in their application to clinical care. This study aims to develop and test a programming framework for an AI-assisted nursing care robot ("E-Nursing Assistant") to reduce nurses' workload and improve the efficiency and quality of care.

A Multicenter Cohort Study of AI-Based Methods for Pulmonary Nodule Diagnosis and Follow-up

This is an observational, multicentre study. The primary objective of this study was to evaluate the diagnostic performance and clinical applicability of artificial intelligence models for pulmonary nodule segmentation, benign-malignant risk stratification, and follow-up management. We will collect CT images and medical data from participants at several hospitals. Participants will not receive any drugs or medical interventions.

EUS-Based Artificial Intelligence to Predict Outcomes in Pancreatic Cancer

The EUS-AI-R study is an observational, single-center, two-phase (retrospective-prospective) study designed to develop and validate artificial intelligence (AI) models for predicting chemotherapy response and oncological outcomes in patients with pancreatic ductal adenocarcinoma (PDAC). Patients who underwent endoscopic ultrasound (EUS) with tissue acquisition (EUS-FNA/FNB) for suspected pancreatic lesions at IRCCS San Raffaele Hospital between January 1st, 2019 and January 2026 will be retrospectively included. All patients have a histologically confirmed diagnosis of PDAC and a minimum follow-up of six months. These data derive from an IRB-approved institutional study (BIOPANCREAS; NCT06552078). Retrospective multimodal data, including EUS imaging (B-mode, elastography, contrast-enhanced EUS), clinical and laboratory variables, CT/MRI imaging, digital pathology, and molecular data when available, will be used to develop and internally validate multiple AI models. In the prospective phase, the best-performing AI model will be applied to an independent cohort of patients undergoing EUS at the same institution to evaluate feasibility, calibration, and real-world performance. No additional procedures beyond standard clinical practice will be performed.

The Effect of AI Training on Arthroscopy Outcomes

This study is planned to determine the effects of AI-assisted regional training on surgical anxiety, fear, postoperative pain, and comfort levels in patients undergoing arthroscopy. It will be designed as a prospective, observational, and intervention-based clinical trial. Necessary ethical and institutional permissions will be obtained before the start of the research. Ethical Approvals and Permissions were obtained from the Niğde Ömer Halisdemir University Ethics Committee, Niğde Provincial Health Directorate, and Ömer Halisdemir Training and Research Hospital. The participants are adult patients undergoing arthroscopy surgery. Informed consent will be obtained for participation in the study. Intervention: AI-assisted regional training will be applied before arthroscopy. Data Collection Tools: Pre-Assessment Form, Personal Information Form, Surgical Fear Scale, Surgical Anxiety Scale for Adult Patients, Postpartum Comfort Scale, and Perianesthesia Comfort Scale will be used. Data Analysis: Data will be evaluated using the IBM SPSS Statistics 24 software package. Appropriate statistical analysis methods will be used.

Federated Learning for Point-of-Care Cardiac Ultrasound

This prospective, multicenter study will evaluate a federated machine-learning system designed to analyze focused cardiac point-of-care ultrasound examinations. Federated learning allows participating clinical sites to contribute to model development while keeping raw ultrasound images and directly identifiable patient information within each site's controlled computing environment. Encrypted model updates, rather than patient images, will be transmitted for secure aggregation. The prospective validation cohort will include approximately 3,000 adults undergoing clinically indicated focused cardiac ultrasound. Model performance will be compared with an expert interpretation of a comprehensive transthoracic echocardiogram performed within 24 hours. The primary objective is to determine how accurately the model identifies reduced left ventricular systolic function, defined as a left ventricular ejection fraction below 40%. During the initial validation period, the investigational software will operate in silent mode. Its results will not be displayed to treating clinicians and will not be used to diagnose participants, select treatment, or replace standard clinical interpretation. The study will also evaluate image-quality classification, cardiac-view recognition, performance across clinical sites and ultrasound systems, model calibration, processing time, cybersecurity, privacy resilience, and performance across demographic and clinical subgroups. Long-term monitoring will assess whether model performance changes as clinical populations, ultrasound equipment, acquisition practices, and software environments evolve during the 2026-2037 study period.

Detection of Systemic Diseases Such as Hepatobiliary Diseases From Ocular Images Via Deep Learning

Oculomics is an emerging interdisciplinary field that deciphers multi-dimensional, high-throughput ocular data to predict, diagnose, and monitor systemic diseases and health span.In recent years, artificial Intelligence may provide insight into exploring the potential covert association behind and reveal some early ocular architecture changes in individuals with systemic diseases. The investigators conducted a survey to explore the association between the eye and systemic diseases via deep learning, to develop and evaluate different deep learning models to predict the systemic diseases such as hepatobiliary disease by using ocular images.