Gallbladder Polyp Classification in Ultrasound Images Using an Ensemble Convolutional Neural Network Model

Advances and current research status of early diagnosis for gallbladder cancer

Gallbladder cancer (GBC) is the most common malignant tumor in the biliary system, characterized by high malignancy, aggressiveness, and poor prognosis. Early diagnosis holds paramount importance in ameliorating therapeutic outcomes. Presently, the clinical diagnosis of GBC primarily relies on clinical-radiological-pathological approach. However, there remains a potential for missed diagnosis and misdiagnose in the realm of clinical practice. We firstly analyzed the blood-based biomarkers, such as carcinoembryonic antigen and carbohydrate antigen 19-9. Subsequently, we evaluated the diagnostic performance of various imaging modalities, including ultrasound (US), endoscopic ultrasound (EUS), computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography/computed tomography (PET/CT) and pathological examination, emphasizing their strengths and limitations in detecting early-stage GBC. Furthermore, we explored the potential of emerging technologies, particularly artificial intelligence (AI) and liquid biopsy, to revolutionize GBC diagnosis. AI algorithms have demonstrated improved image analysis capabilities, while liquid biopsy offers the promise of non-invasive and real-time monitoring. However, the translation of these advancements into clinical practice necessitates further validation and standardization. The review highlighted the advantages and limitations of current diagnostic approaches and underscored the need for innovative strategies to enhance diagnostic accuracy of GBC. In addition, we emphasized the importance of multidisciplinary collaboration to improve early diagnosis of GBC and ultimately patient outcomes. This review endeavoured to impart fresh perspectives and insights into the early diagnosis of GBC.

Artificial intelligence in abdominal and pelvic ultrasound imaging: current applications

BACKGROUND

In recent years, the integration of artificial intelligence (AI) techniques into medical imaging has shown great potential to transform the diagnostic process. This review aims to provide a comprehensive overview of current state-of-the-art applications for AI in abdominal and pelvic ultrasound imaging.

METHODS

We searched the PubMed, FDA, and ClinicalTrials.gov databases for applications of AI in abdominal and pelvic ultrasound imaging.

RESULTS

A total of 128 titles were identified from the database search and were eligible for screening. After screening, 57 manuscripts were included in the final review. The main anatomical applications included multi-organ detection (n = 16, 28%), gynecology (n = 15, 26%), hepatobiliary system (n = 13, 23%), and musculoskeletal (n = 8, 14%). The main methodological applications included deep learning (n = 37, 65%), machine learning (n = 13, 23%), natural language processing (n = 5, 9%), and robots (n = 2, 4%). The majority of the studies were single-center (n = 43, 75%) and retrospective (n = 56, 98%). We identified 17 FDA approved AI ultrasound devices, with only a few being specifically used for abdominal/pelvic imaging (infertility monitoring and follicle development).

CONCLUSION

The application of AI in abdominal/pelvic ultrasound shows promising early results for disease diagnosis, monitoring, and report refinement. However, the risk of bias remains high because very few of these applications have been prospectively validated (in multi-center studies) or have received FDA clearance.

WFUMB Commentary Paper on Artificial intelligence in Medical Ultrasound Imaging

Artificial intelligence (AI) is defined as the theory and development of computer systems able to perform tasks normally associated with human intelligence. At present, AI has been widely used in a variety of ultrasound tasks, including in point-of-care ultrasound, echocardiography, and various diseases of different organs. However, the characteristics of ultrasound, compared to other imaging modalities, such as computed tomography (CT) and magnetic resonance imaging (MRI), poses significant additional challenges to AI. Application of AI can not only reduce variability during ultrasound image acquisition, but can standardize these interpretations and identify patterns that escape the human eye and brain. These advances have enabled greater innovations in ultrasound AI applications that can be applied to a variety of clinical settings and disease states. Therefore, The World Federation of Ultrasound in Medicine and Biology (WFUMB) is addressing the topic with a brief and practical overview of current and potential future AI applications in medical ultrasound, as well as discuss some current limitations and future challenges to AI implementation.

Establishing a radiomics model using contrast-enhanced ultrasound for preoperative prediction of neoplastic gallbladder polyps exceeding 10 mm

BACKGROUND

A key challenge in the medical field is managing gallbladder polyps (GBP) > 10 mm, especially when their nature is uncertain. GBP with a diameter exceeding 10 mm are associated with an increased risk of gallbladder cancer, making the key to their management the differentiation between benign and malignant types. The current practice, due to the inability to predict accurately, leads to excessive surgeries and ineffective follow-ups, increasing patient risks and medical burdens.

PURPOSE

This study aims to establish an imaging radiomics model using clinical data and contrast-enhanced ultrasound (CEUS) to predict neoplastic GBP exceeding 10 mm in diameter preoperatively.

MATERIALS AND METHODS

Data from 119 patients with GBP > 10 mm of unknown origin were analyzed. A total of 1197 features were extracted from the GBP area using conventional ultrasound (US) and CEUS. Significant features were identified using the Mann-Whitney U test and further refined with a least absolute shrinkage and selection operator (LASSO) regression model to construct radiomic features. By integrating clinical characteristics, a radiomics nomogram was developed. The diagnostic efficacy of the preoperative logistic regression (LR) model was validated using receiver operating characteristic (ROC) curves, calibration plots, and the Hosmer-Lemeshow test. CEUS is an examination based on conventional ultrasound, and conventional two-dimensional ultrasound still poses significant challenges in differential diagnosis. CEUS has a high accuracy rate in diagnosing the benign or malignant nature of gallbladder space-occupying lesions, which can significantly reduce the preoperative waiting time for related examinations and provide more reliable diagnostic information for clinical practice.

RESULTS

Feature selection via Lasso led to a final LR model incorporating high-density lipoprotein, smoking status, basal width, and Rad_Signature. This model, derived from machine learning frameworks including Support Vector Machine (SVM), Logistic Regression (LR), Multilayer Perceptron (MLP), k-Nearest Neighbors (KNN), and eXtreme Gradient Boosting (XGBoost) with fivefold cross-validation, showed AUCs of 0.95 (95% CI: 0.90-0.99) and 0.87 (95% CI: 0.72-1.0) in internal validation. The model exhibited excellent calibration, confirmed by calibration graphs and the Hosmer-Lemeshow test (P = 0.551 and 0.544).

CONCLUSION

The LR model accurately predicts neoplastic GBP > 10 mm preoperatively. Radiomics with CEUS is a powerful tool for analysis of GBP > 10 mm. The model not only improves diagnostic accuracy and reduces healthcare costs but also optimizes patient management through personalized treatment plans, enhancing clinical outcomes and ensuring resources are more precisely allocated to patients who need surgery.

Predicting Neoplastic Gallbladder Polyps: The Role of Current Surgical Indications and Preoperative Images

Background/Aims

Cholecystectomy for gallbladder (GB) polyps is performed primarily based on preoperative images. This study examined the accuracy of surgical indications commonly used in clinical practice for detecting neoplastic polyps and investigated further clues for predicting neoplastic polyps.

Methods

This retrospective study included 385 patients who underwent a cholecystectomy for GB polyps. The predictive performances of seven surgical indications were compared by fitting the receiver operating characteristic curves. Logistic regression analysis was used to identify the candidate variables associated with predicting neoplastic polyps.

Results

Neoplastic polyps were identified in 18.9% (n=62) of the 385 patients assessed. The neoplastic group contained more females than males, larger polyps, more frequent solitary lesions, and lower platelet counts than the non-neoplastic group. Current surgical indications revealed an unsatisfactory prediction for neoplastic polyps. The optimal cutoff polyp size for neoplastic polyps by ultrasound (US) was larger than by computed tomography (CT) (12 mm vs. 10 mm). The proportion of pathologic neoplastic polyps was higher when both US and CT images were used than that predicted using a single test. Logistic regression analysis revealed larger polyps, increasing age, female sex, and lower platelet count to be associated with neoplastic polyps.

Conclusions

The current indications for cholecystectomy in GB polyps have a low predictive value for neoplastic lesions that can lead to overtreatment. Combining the polyp size from US and CT images may reduce unnecessary surgery. In addition, knowledge of the patient's age, sex, and platelet count could help make more selective surgical decisions for neoplastic polyps.

Diagnosis of Gallbladder Disease Using Artificial Intelligence: A Comparative Study

Gallbladder (GB) disease is a common pathology that needs correct and early diagnosis for the optimum medical treatment. Early diagnosis is crucial as any delay or misdiagnosis can worsen the patient situation. Incorrect diagnosis could also lead to an escalation in patient symptoms and poorer clinical outcomes. The use of Artificial Intelligence (AI) techniques, ranging from Machine Learning (ML) to Deep Learning (DL) to predict disease progression, identify abnormalities, and estimate mortality rates associated with GB disorders has increased over the past decade. To this end, this paper provides a comprehensive overview of the AI approaches used in the diagnosis of GB illnesses. This review compiles and compares relevant papers from the last decade to show how AI might enhance diagnostic precision, speed, and efficiency. Therefore, this survey gives researchers the opportunity to find out both the diagnosis of GB diseases and AI techniques in one place. The maximum accuracy rate by ML was when using SVM with 96.67%, whilst the maximum accuracy rate by DL was by utilising a unique structure of VGG, GoogleNet, ResNet, AlexNet and Inception with 98.77%. This could provide a clear path for further investigations and algorithm’s development to boost diagnostic results to improve the patient’s condition and choose the appropriate treatment.

Advances in the management of gallbladder polyps: establishment of predictive models and the rise of gallbladder-preserving polypectomy procedures

Gallbladder polyps are a common biliary tract disease whose treatment options have yet to be fully established. The indication of "polyps ≥ 10 mm in diameter" for cholecystectomy increases the possibility of gallbladder excision due to benign polyps. Compared to enumeration of risk factors in clinical guidelines, predictive models based on statistical methods and artificial intelligence provide a more intuitive representation of the malignancy degree of gallbladder polyps. Minimally invasive gallbladder-preserving polypectomy procedures, as a combination of checking and therapeutic approaches that allow for eradication of lesions and preservation of a functional gallbladder at the same time, have been shown to maximize the benefits to patients with benign polyps. Despite the reported good outcomes of predictive models and gallbladder-preserving polypectomy procedures, the studies were associated with various limitations, including small sample sizes, insufficient data types, and unknown long-term efficacy, thereby enhancing the need for multicenter and large-scale clinical studies. In conclusion, the emergence of predictive models and minimally invasive gallbladder-preserving polypectomy procedures has signaled an ever increasing attention to the role of the gallbladder and clinical management of gallbladder polyps.

Risk stratification model for incidentally detected gallbladder polyps: A multicentre study

PURPOSE

We aimed to develop a 4-level risk stratification model using a scoring system based on conventional ultrasound to improve the diagnosis of gallbladder polyp.

METHOD

Patients with histopathologically confirmed gallbladder polyps were consecutively recruited from three medical centres. Conventional ultrasound findings and clinical characteristics were acquired prior to cholecystectomy. Risk factors for neoplastic and malignant polyps were used to build a risk stratification system via interobserver agreement and multivariate logistic regression analysis. The model was retrospectively trained using 264 pre-surgical samples and prospectively validated using 106 pre-surgical samples. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC) and malignant polyp rate.

RESULTS

In total, 370 patients (mean age, 51.68 ± 14.41 years, 156 men) were enrolled in this study. Size (≥12 mm), shape (oblate or round), single, vascularity, gallbladder stone or sludge were considered risk factors for neoplastic polyps. Size (≥14 mm), shape (oblate), single, disrupted gallbladder wall, and gallbladder stone or sludge were risk factors for malignant polyps (all p < 0.05). In the scoring system, the sensitivity, specificity, and AUC of score ≥ 9 in diagnosing neoplastic polyps were 0.766, 0.788, and 0.876 respectively; and the sensitivity, specificity, and AUC of score ≥ 15 in diagnosing malignant polyps were 0.844, 0.926, and 0.949 respectively. In our model, the malignancy rates at the four levels were 0 % (0/24), 1.28 % (2/156), 9.26 % (5/54), and 70.37 % (38/54), respectively.

CONCLUSIONS

The 4-level risk stratification model based on conventional ultrasound imaging showed excellent performance in classifying gallbladder polyps.

Value of Micro Flow Imaging in the Prediction of Adenomatous Polyps

OBJECTIVE

The aim of this study was to assess the value of micro flow imaging (MFI) in distinguishing adenomatous polyps from cholesterol polyps.

METHODS

A total of 143 patients who underwent cholecystectomy for gallbladder polyps were retrospectively analyzed. B-mode ultrasound (BUS), color Doppler flow imaging (CDFI), MFI and contrast-enhanced ultrasound (CEUS) were performed before cholecystectomy. The weighted kappa consistency test was used to evaluate the agreement of vascular morphology among CDFI, MFI and CEUS. Ultrasound image characteristics, including BUS, CDFI and MFI images, were compared between adenomatous polyps and cholesterol polyps. The independent risk factors for adenomatous polyps were selected. The diagnostic performance of MFI combined with BUS in determining adenomatous polyps was compared with CDFI combined with BUS.

RESULTS

Of the 143 patients, 113 cases were cholesterol polyps, and 30 cases were adenomatous polyps. The vascular morphology of gallbladder polyps was more clearly depicted by MFI than CDFI, and it had better agreement with CEUS. Differences in maximum size, height/width ratio, hyperechoic spot and vascular intensity on CDFI and MFI images were significant between adenomatous polyps and cholesterol polyps (p < 0.05). The maximum size, height/width ratio, and vascular intensity on MFI images were independent risk factors for adenomatous polyps. For MFI combined with BUS, sensitivity, specificity and accuracy were 90.00%, 94.69% and 93.70%, respectively. Area under the receiver operating characteristic curve (AUC) of MFI combined with BUS was significantly higher than that of CDFI combined with BUS (AUC = 0.923 vs. 0.784).

CONCLUSION

Compared with CDFI combined with BUS, MFI combined with BUS had higher diagnostic performance in determining adenomatous polyps.

Incidental Findings of Gallbladder and Bile Ducts-Management Strategies: General Aspects, Gallbladder Polyps and Gallbladder Wall Thickening-A World Federation of Ultrasound in Medicine and Biology (WFUMB) Position Paper

The World Federation of Ultrasound in Medicine and Biology (WFUMB) is addressing the issue of incidental findings with a series of position papers to give advice on characterization and management. The biliary system (gallbladder and biliary tree) is the third most frequent site for incidental findings. This first part of the position paper on incidental findings of the biliary system is related to general aspects, gallbladder polyps and other incidental findings of the gallbladder wall. Available evidence on prevalence, diagnostic work-up, malignancy risk, follow-up and treatment is summarized with a special focus on ultrasound techniques. Multiparametric ultrasound features of gallbladder polyps and other incidentally detected gallbladder wall pathologies are described, and their inclusion in assessment of malignancy risk and decision- making on further management is suggested.

Diagnosing gangrenous cholecystitis on computed tomography using deep learning: A preliminary study

Aim

To compare deep learning and experienced physicians in diagnosing gangrenous cholecystitis using computed tomography images and explore the feasibility of diagnostic assistance for acute cholecystitis requiring emergency surgery.

Methods

This retrospective study included 25 patients with pathologically confirmed gangrenous cholecystitis and 129 patients with noncomplicated acute cholecystitis who underwent computed tomography between 2016 and 2021 at two institutions. All available computed tomography images at the time of the initial diagnosis were used for the analysis. A deep learning model based on a convolutional neural network was trained using 1,517 images of 112 patients (18 patients with gangrenous cholecystitis and 94 patients with acute cholecystitis) and tested with 68 images of 42 patients (seven patients with gangrenous cholecystitis and 35 patients with acute cholecystitis). Three blinded, experienced physicians independently interpreted the test images. The sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve were compared between the convolutional neural network and the reviewers.

Results

The convolutional neural network (sensitivity, 0.70; 95% confidence interval [CI], 0.44–0.87, specificity, 0.93; 95% CI, 0.88–0.96, accuracy, 0.89; 95% CI, 0.81–0.95, area under the receiver operating characteristic curve, 0.84; 95% CI, 0.68–1.00) had achieved a better diagnostic performance than the reviewers (ex. sensitivity, 0.55; 95% CI, 0.30–0.77, specificity, 0.67; 95% CI, 0.62–0.71, accuracy, 0.65; 95% CI, 0.57–0.72, area under the receiver operating characteristic curve, 0.63; 95% CI, 0.44–0.82; P = 0.048 for area under the receiver operating characteristic curve versus convolutional neural network).

Conclusions

Deep learning had a better diagnostic performance than experienced reviewers in diagnosing gangrenous cholecystitis and has potential applicability for assisting in identifying indications for emergency surgery in the future.