Development and validation of a combined nomogram model based on deep learning contrast-enhanced ultrasound and clinical factors to predict preoperative aggressiveness in pancreatic neuroendocrine neoplasms

Diagnostic and Prognostic Ability of Contrast-Enhanced Unltrasound and Biomarkers in Hepatocellular Carcinoma Subtypes

OBJECTIVE

To investigate the diagnostic and prognostic value of contrast-enhanced ultrasound (CEUS) and clinical indicators of the vessels encapsulating tumor clusters (VETC) pattern and macrotrabecular-massive subtype in hepatocellular carcinoma (MTM-HCC).

METHODS

This retrospective study included patients who underwent preoperative CEUS and hepatectomy for HCC between August 2018 and August 2021. Multivariable logistic regression was performed to select independent correlated factors of VETC-HCC and MTM-HCC to develop nomogram models. The association between model outcomes and early postoperative HCC recurrence was assessed using Kaplan-Meier curve and Cox regression analysis.

RESULTS

The training cohort included 182 patients (54.3 ± 11.3 years, 168 males) and the validation cohort included 91 patients (54.8 ± 10.6 years, 81 males). Multivariate logistic regression analysis revealed that α-fetoprotein (AFP) levels (odds ratio [OR]: 2.26, 95% confidence interval [CI]: 1.49-3.42, p < 0.001), intratumoral nonenhancement (OR: 2.40, 95% CI: 1.02-5.64, p = 0.044), and the perfusion pattern in the CEUS arterial phase (OR: 2.27, 95% CI: 1.05-4.91, p = 0.038) were independent predictors of VETC-HCC. Besides, the former two were also independently associated with MTM-HCC (AFP level: OR: 2.36, 95% CI: 1.36-4.09, p = 0.002; intratumoral nonenhancement: OR: 3.72, 95% CI: 1.02-13.56, p = 0.046). Nomogram models were constructed based on the aforementioned indicators. Kaplan-Meier curve analysis indicated that predicted VETC-HCC or MTM-HCC exhibited higher rates of early recurrence (log-rank p < 0.001 and p = 0.002, respectively). Cox regression analysis showed that a high risk of VETC-HCC was independently correlated with early recurrence (p = 0.011).

CONCLUSION

CEUS combined with AFP levels can predict VETC-HCC/MTM-HCC and prognosis preoperatively.

Development and validation of a nomogram model based on pretreatment ultrasound and contrast-enhanced ultrasound to predict the efficacy of neoadjuvant chemotherapy in patients with borderline resectable or locally advanced pancreatic cancer

OBJECTIVES

To develop a nomogram using pretreatment ultrasound (US) and contrast-enhanced ultrasound (CEUS) to predict the clinical response of neoadjuvant chemotherapy (NAC) in patients with borderline resectable pancreatic cancer (BRPC) or locally advanced pancreatic cancer (LAPC).

METHODS

A total of 111 patients with pancreatic ductal adenocarcinoma (PDAC) treated with NAC between October 2017 and February 2022 were retrospectively enrolled. The patients were randomly divided (7:3) into training and validation cohorts. The pretreatment US and CEUS features were reviewed. Univariate and multivariate logistic regression analyses were used to determine the independent predictors of clinical response in the training cohort. Then a prediction nomogram model based on the independent predictors was constructed. The area under the curve (AUC), calibration plot, C-index and decision curve analysis (DCA) were used to assess the nomogram's performance, calibration, discrimination and clinical benefit.

RESULTS

The multivariate logistic regression analysis showed that the taller-than-wide shape in the longitudinal plane (odds ratio [OR]:0.20, p = 0.01), time from injection of contrast agent to peak enhancement (OR:3.64; p = 0.05) and Peaktumor/ Peaknormal (OR:1.51; p = 0.03) were independent predictors of clinical response to NAC. The predictive nomogram developed based on the above imaging features showed AUCs were 0.852 and 0.854 in the primary and validation cohorts, respectively. Good calibration was achieved in the training datasets, with C-index of 0.852. DCA verified the clinical usefulness of the nomogram.

CONCLUSIONS

The nomogram based on pretreatment US and CEUS can effectively predict the clinical response of NAC in patients with BRPC and LAPC; it may help guide personalized treatment.

The Value of Applying Machine Learning in Predicting the Time of Symptom Onset in Stroke Patients: Systematic Review and Meta-Analysis

BACKGROUND

Machine learning is a potentially effective method for identifying and predicting the time of the onset of stroke. However, the value of applying machine learning in this field remains controversial and debatable.

OBJECTIVE

We aimed to assess the value of applying machine learning in predicting the time of stroke onset.

METHODS

PubMed, Web of Science, Embase, and Cochrane were comprehensively searched. The C index and sensitivity with 95% CI were used as effect sizes. The risk of bias was evaluated using PROBAST (Prediction Model Risk of Bias Assessment Tool), and meta-analysis was conducted using R (version 4.2.0; R Core Team).

RESULTS

Thirteen eligible studies were included in the meta-analysis involving 55 machine learning models with 41 models in the training set and 14 in the validation set. The overall C index was 0.800 (95% CI 0.773-0.826) in the training set and 0.781 (95% CI 0.709-0.852) in the validation set. The sensitivity and specificity were 0.76 (95% CI 0.73-0.80) and 0.79 (95% CI 0.74-0.82) in the training set and 0.81 (95% CI 0.68-0.90) and 0.83 (95% CI 0.73-0.89) in the validation set, respectively. Subgroup analysis revealed that the accuracy of machine learning in predicting the time of stroke onset within 4.5 hours was optimal (training: 0.80, 95% CI 0.77-0.83; validation: 0.79, 95% CI 0.71-0.86).

CONCLUSIONS

Machine learning has ideal performance in identifying the time of stroke onset. More reasonable image segmentation and texture extraction methods in radiomics should be used to promote the value of applying machine learning in diverse ethnic backgrounds.

TRIAL REGISTRATION

PROSPERO CRD42022358898; https://www.crd.york.ac.uk/Prospero/display_record.php?RecordID=358898.

Diagnostic value of deep learning-assisted endoscopic ultrasound for pancreatic tumors: a systematic review and meta-analysis

Background and aims

Endoscopic ultrasonography (EUS) is commonly utilized in the diagnosis of pancreatic tumors, although as this modality relies primarily on the practitioner's visual judgment, it is prone to result in a missed diagnosis or misdiagnosis due to inexperience, fatigue, or distraction. Deep learning (DL) techniques, which can be used to automatically extract detailed imaging features from images, have been increasingly beneficial in the field of medical image-based assisted diagnosis. The present systematic review included a meta-analysis aimed at evaluating the accuracy of DL-assisted EUS for the diagnosis of pancreatic tumors diagnosis.

Methods

We performed a comprehensive search for all studies relevant to EUS and DL in the following four databases, from their inception through February 2023: PubMed, Embase, Web of Science, and the Cochrane Library. Target studies were strictly screened based on specific inclusion and exclusion criteria, after which we performed a meta-analysis using Stata 16.0 to assess the diagnostic ability of DL and compare it with that of EUS practitioners. Any sources of heterogeneity were explored using subgroup and meta-regression analyses.

Results

A total of 10 studies, involving 3,529 patients and 34,773 training images, were included in the present meta-analysis. The pooled sensitivity was 93% (95% confidence interval [CI], 87-96%), the pooled specificity was 95% (95% CI, 89-98%), and the area under the summary receiver operating characteristic curve (AUC) was 0.98 (95% CI, 0.96-0.99).

Conclusion

DL-assisted EUS has a high accuracy and clinical applicability for diagnosing pancreatic tumors.

Systematic review registration

https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42023391853, identifier CRD42023391853.

Pre-operative Prediction of Invasiveness in Renal Cell Carcinoma: The Role of Conventional Ultrasound and Contrast-Enhanced Ultrasound

OBJECTIVE

It is known that in patients with renal cell carcinoma (RCC), the invasiveness of the tumor is closely related to the treatment and prognosis. Currently, histologic diagnosis of RCC is typically established after surgical removal of tumors or after biopsy. The use of non-invasive imaging modalities to predict the invasiveness of RCC is of great clinical value, particularly before surgery. In this study, the differences in conventional ultrasound (US) and contrast-enhanced ultrasound (CEUS) features between invasive and non-invasive RCC were analyzed with the aim of providing more accurate and valuable information for diagnosis and treatment to clinically optimize the treatment plan in a non-invasive manner and improve the prognosis of patients.

METHODS

Conventional US and CEUS features of 163 patients (total of 164 RCCs), obtained from the Lanzhou University Second Hospital in the period ranging from March 2021 to September 2022, were retrospectively analyzed. Patients were categorized into two groups: invasive group (n = 44) and non-invasive group (n = 120), with surgical pathology as reference standard. Receiver operating characteristic curves were drawn to evaluate the feasibility of differentiation.

RESULTS

The possibility of an intrarenal lesion/kidney ratio >50% in the invasive group (13/44, 29.5%) was significantly higher than that in the non-invasive group (8/120, 6.7%) (p < 0.001). The absence of perilesional rim-like enhancement was more likely to imply invasive RCC (30/44, 68.2%) than non-invasive RCC (100/120, 83.3%) (p = 0.049) and was an independent predictor of invasive RCC. As for CEUS quantitative features, there were statistically significant differences in peak intensity (p = 0.009) or peak enhancement (p = 0.010), taking the largest range of lesion as the region of interest.

CONCLUSION

Conventional US and CEUS features may help in the differentiation of invasive RCC from non-invasive RCC and have potential application value in the pre-operative prediction of RCC invasiveness.

Predicting the Efficacy of Neoadjuvant Chemotherapy for Pancreatic Cancer Using Deep Learning of Contrast-Enhanced Ultrasound Videos

Contrast-enhanced ultrasound (CEUS) is a promising imaging modality in predicting the efficacy of neoadjuvant chemotherapy for pancreatic cancer, a tumor with high mortality. In this study, we proposed a deep-learning-based strategy for analyzing CEUS videos to predict the prognosis of pancreatic cancer neoadjuvant chemotherapy. Pre-trained convolutional neural network (CNN) models were used for binary classification of the chemotherapy as effective or ineffective, with CEUS videos collected before chemotherapy as the model input, and with the efficacy after chemotherapy as the reference standard. We proposed two deep learning models. The first CNN model used videos of ultrasound (US) and CEUS (US+CEUS), while the second CNN model only used videos of selected regions of interest (ROIs) within CEUS (CEUS-ROI). A total of 38 patients with strict restriction of clinical factors were enrolled, with 76 original CEUS videos collected. After data augmentation, 760 and 720 videos were included for the two CNN models, respectively. Seventy-six-fold and 72-fold cross-validations were performed to validate the classification performance of the two CNN models. The areas under the curve were 0.892 and 0.908 for the two models. The accuracy, recall, precision and F1 score were 0.829, 0.759, 0.786, and 0.772 for the first model. Those were 0.864, 0.930, 0.866, and 0.897 for the second model. A total of 38.2% and 40.3% of the original videos could be clearly distinguished by the deep learning models when the naked eye made an inaccurate classification. This study is the first to demonstrate the feasibility and potential of deep learning models based on pre-chemotherapy CEUS videos in predicting the efficacy of neoadjuvant chemotherapy for pancreas cancer.

Histogram array and convolutional neural network of DWI for differentiating pancreatic ductal adenocarcinomas from solid pseudopapillary neoplasms and neuroendocrine neoplasms

PURPOSE

This study aimed to investigate the diagnostic performance of the histogram array and convolutional neural network (CNN) based on diffusion-weighted imaging (DWI) with multiple b-values under magnetic resonance imaging (MRI) to distinguish pancreatic ductal adenocarcinomas (PDACs) from solid pseudopapillary neoplasms (SPNs) and pancreatic neuroendocrine neoplasms (PNENs).

METHODS

This retrospective study consisted of patients diagnosed with PDACs (n = 132), PNENs (n = 45) and SPNs (n = 54). All patients underwent 3.0-T MRI including DWI with 10 b values. The regions of interest (ROIs) of pancreatic tumor were manually drawn using ITK-SNAP software, which included entire tumor at DWI (b = 1500 s/m²). The histogram array was obtained through the ROIs from multiple b-value data. PyTorch (version 1.11) was used to construct a CNN classifier to categorize the histogram array into PDACs, PNENs or SPNs.

RESULTS

The area under the curves (AUCs) of the histogram array and the CNN model for differentiating PDACs from PNENs and SPNs were 0.896, 0.846, and 0.839 in the training, validation and testing cohorts, respectively. The accuracy, sensitivity and specificity were 90.22%, 96.23%, and 82.05% in the training cohort, 84.78%, 96.15%, and 70.0% in the validation cohort, and 81.72%, 90.57%, and 70.0% in the testing cohort. The performance of CNN with AUC of 0.865 for this differentiation was significantly higher than that of f with AUC = 0.755 (P = 0.0057) and α with AUC = 0.776 (P = 0.0278) in all patients.

CONCLUSION

The histogram array and CNN based on DWI data with multiple b-values using MRI provided an accurate diagnostic performance to differentiate PDACs from PNENs and SPNs.