Application of artificial intelligence in gastroenterology: Potential role in clinical practice

Development and validation of a spontaneous preterm birth risk prediction algorithm based on maternal bioinformatics: A single-center retrospective study

BACKGROUND

Spontaneous preterm birth (sPTB) is a primary cause of adverse neonatal outcomes. The objective of this study is to analyze the factors influencing the occurrence of sPTB in pregnant women and to construct and validate a predictive model for sPTB risk based on big data from clinical and laboratory assessments during pregnancy.

METHODS

A retrospective analysis was conducted on the clinical data of 3,082 pregnant women, categorizing those who delivered before 37 weeks of gestation as the sPTB group and those who delivered at or after 37 weeks as the full-term group. The performance of five machine learning models was compared using metrics such as the AUC, accuracy, sensitivity, specificity, and precision to identify the optimal predictive model. The top 10 predictive variables were selected based on their significance in disease prediction. The data were then divided into a training set (70%) and a validation set (30%) for validation. External data were also utilized to validate the model's predictive performance.

RESULTS

A total of 24 indicators with significant differences were identified. In terms of predicting the risk of preterm birth, the XGBoost algorithm demonstrated the most outstanding performance, with an AUCROC of 0.89 (95% CI: 0.88-0.90). The top 10 critical indicators included ALP, AFP, ALB, HCT, TC, DBP, ALT, PLT, height, and SBP, which are essential for constructing an accurate predictive model. The model exhibited stable performance on both the training and validation sets, with AUC values of 0.93 and 0.87, respectively. Furthermore, the external testing set also showed superior performance, with an AUC of 0.79.

CONCLUSIONS

At the time of delivery, ALP, AFP, ALB, HCT, TC, DBP, ALT, PLT, height, and SBP are influential factors for sPTB in pregnant women. The XGBoost algorithm, constructed based on these factors, demonstrated the most outstanding performance.

Noninvasive Computed Tomography-Based Deep Learning Model Predicts In Vitro Chemosensitivity Assay Results in Pancreatic Cancer

OBJECTIVES

We aimed to predict in vitro chemosensitivity assay results from computed tomography (CT) images by applying deep learning (DL) to optimize chemotherapy for pancreatic ductal adenocarcinoma (PDAC).

MATERIALS AND METHODS

Preoperative enhanced abdominal CT images and the histoculture drug response assay (HDRA) results were collected from 33 PDAC patients undergoing surgery. Deep learning was performed using CT images of both the HDRA-positive and HDRA-negative groups. We trimmed small patches from the entire tumor area. We established various prediction labels for HDRA results with 5-fluorouracil (FU), gemcitabine (GEM), and paclitaxel (PTX). We built a predictive model using a residual convolutional neural network and used 3-fold cross-validation.

RESULTS

Of the 33 patients, effective response to FU, GEM, and PTX by HDRA was observed in 19 (57.6%), 11 (33.3%), and 23 (88.5%) patients, respectively. The average accuracy and the area under the receiver operating characteristic curve (AUC) of the model for predicting the effective response to FU were 93.4% and 0.979, respectively. In the prediction of GEM, the models demonstrated high accuracy (92.8%) and AUC (0.969). Likewise, the model for predicting response to PTX had a high performance (accuracy, 95.9%; AUC, 0.979).

CONCLUSIONS

Our CT patch-based DL model exhibited high predictive performance in projecting HDRA results. Our study suggests that the DL approach could possibly provide a noninvasive means for the optimization of chemotherapy.

Development and validation of a scoring system to predict esophagogastroduodenoscopy necessity

OBJECTIVE

This study aimed to develop and validate a scoring system for predicting the need for esophagogastroduodenoscopy (EGD) in clinical practice to enhance accuracy and reduce misapplications.

METHODS

From February 2021 to April 2022, outpatients scheduled for EGD at the Department of Gastroenterology in our hospital were recruited. Patients completed the system evaluation by providing clinical symptoms, relevant medical history, and endoscopic findings. Patients were randomly divided into the training and validation cohorts (at 2:1 ratio). The optimal algorithm was selected from five alternatives including a parallel test. Six physicians participated in a human-computer comparative validation. Sensitivity and negative likelihood ratio (-LR) were used as the primary indicators.

RESULTS

Altogether 865 patients were enrolled, with 578 in the training cohort and 287 in the validation cohort. The scoring system comprised 21 variables, including age, 13 typical clinical symptoms, and seven medical history variables. The parallel test was selected as the final algorithm. Positive EGD findings were reported in 54.5% of the training cohort and 62.7% of the validation cohort. The scoring system demonstrated a sensitivity of 79.0% in the training cohort and 83.9% in the validation cohort, with -LR being 0.627 and 0.615, respectively. Compared to physicians, the scoring system exhibited higher sensitivity (84.0% vs 68.7%, P = 0.02) and a lower -LR (1.11 vs 2.41, P = 0.439).

CONCLUSIONS

We developed a scoring system to predict the necessity of EGD using a parallel test algorithm, which was user-friendly and effective, as evidenced by single-center validation.

Detection of colorectal adenomas using artificial intelligence models in patients with chronic hepatitis C

BACKGROUND

Hepatitis C virus is known for its oncogenic potential, especially in hepatocellular carcinoma and non-Hodgkin lymphoma. Several studies have shown that chronic hepatitis C (CHC) has an increased risk of the development of colorectal cancer (CRC).

AIM

To analyze this positive relationship and develop an artificial intelligence (AI)-based tool using machine learning (ML) algorithms to stratify these patient populations into risk groups for CRC/adenoma detection.

METHODS

To develop the AI automated calculator, we applied ML to train models to predict the probability and the number of adenomas detected on colonoscopy. Data sets were split into 70:30 ratios for training and internal validation. The Scikit-learn standard scaler was used to scale values of continuous variables. Colonoscopy findings were used as the gold standard and deep learning architecture was used to train six ML models for prediction. A Flask (customizable Python framework) application programming interface (API) was used to deploy the trained ML model with the highest accuracy as a web application. Finally, Heroku was used for the deployment of the web-based API to https://adenomadetection.herokuapp.com.

RESULTS

Of 415 patients, 206 had colonoscopy results. On internal validation, the Bernoulli naive Bayes model predicted the probability of adenoma detection with the highest accuracy of 56%, precision of 55%, recall of 55%, and F1 measure of 54%. Support vector regressor predicted the number of adenomas with the least mean absolute error of 0.905.

CONCLUSION

Our AI-based tool can help providers stratify patients with CHC for early referral for screening colonoscopy. Along with providing a numerical percentage, the calculator can also comment on the number of adenomatous polyps a gastroenterologist can expect, prompting a higher adenoma detection rate.

Quality indicators in esophagogastroduodenoscopy

Esophagogastroduodenoscopy (EGD) has been used to diagnose a wide variety of upper gastrointestinal diseases. In particular, EGD is used to screen high-risk subjects of gastric cancer. Quality control of EGD is important because the diagnostic rate is examiner-dependent. However, there is still no representative quality indicator that can be uniformly applied in EGD. There has been growing awareness of the importance of quality control in improving EGD performance. Therefore, we aimed to review the available and emerging quality indicators for diagnostic EGD.