The Applications of Artificial Intelligence in Digestive System Neoplasms: A Review

Deep Learning for Automatic Gross Tumor Volumes Contouring in Esophageal Cancer Based on Contrast-Enhanced Computed Tomography Images: A Multi-Institutional Study

PURPOSE

To develop and externally validate an automatic artificial intelligence (AI) tool for delineating gross tumor volume (GTV) in patients with esophageal squamous cell carcinoma (ESCC), which can assist in neo-adjuvant or radical radiation therapy treatment planning.

METHODS AND MATERIALS

In this multi-institutional study, contrast-enhanced CT images from 580 eligible ESCC patients were retrospectively collected. The GTV contours delineated by 2 experts via consensus were used as ground truth. A 3-dimensional deep learning model was developed for GTV contouring in the training cohort and internally and externally validated in 3 validation cohorts. The AI tool was compared against 12 board-certified experts in 25 patients randomly selected from the external validation cohort to evaluate its assistance in improving contouring performance and reducing variation. Contouring performance was measured using dice similarity coefficient (DSC) and average surface distance. Additionally, our previously established radiomics model for predicting pathologic complete response was used to compare AI-generated and ground truth contours, to assess the potential of the AI contouring tool in radiomics analysis.

RESULTS

The AI tool demonstrated good GTV contouring performance in multicenter validation cohorts, with median DSC values of 0.865, 0.876, and 0.866 and median average surface distance values of 0.939, 0.789, and 0.875 mm, respectively. Furthermore, the AI tool significantly improved contouring performance for half of 12 board-certified experts (DSC values, 0.794-0.835 vs 0.856-0.881, P = .003-0.048), reduced the intra- and interobserver variations by 37.4% and 55.2%, respectively, and saved contouring time by 77.6%. In the radiomics analysis, 88.7% of radiomic features from ground truth and AI-generated contours demonstrated stable reproducibility, and similar pathologic complete response prediction performance for these contours (P = .430) was observed.

CONCLUSIONS

Our AI contouring tool can improve GTV contouring performance and facilitate radiomics analysis in ESCC patients, which indicates its potential for GTV contouring during radiation therapy treatment planning and radiomics studies.

Radiomics Nomogram with Added Nodal Features Improves Treatment Response Prediction in Locally Advanced Esophageal Squamous Cell Carcinoma: A Multicenter Study

OBJECTIVE

We aimed to develop and validate a radiomics nomogram and determine the value of radiomic features from lymph nodes (LNs) for predicting pathological complete response (pCR) to neoadjuvant chemoradiotherapy (NCRT) in patients with locally advanced esophageal squamous cell carcinoma (ESCC).

METHODS

In this multicenter retrospective study, eligible participants who had undergone NCRT followed by radical esophagectomy were consecutively recruited. Three radiomics models (modelT, modelLN, and modelTLN) based on tumor and LN features, alone and combined, were developed in the training cohort. The radiomics nomogram was developed by incorporating the prediction value of the radiomics model and clinicoradiological risk factors using multivariate logistic regression, and was evaluated using the receiver operating characteristic curve, validated in two external validation cohorts.

RESULTS

Between October 2011 and December 2018, 116 patients were included in the training cohort. Between June 2015 and October 2020, 51 and 27 patients from two independent hospitals were included in validation cohorts 1 and 2, respectively. The radiomics modelTLN performed better than the radiomics modelT for predicting pCR. The radiomics nomogram incorporating the predictive value of the radiomics modelTLN and heterogeneous after NCRT outperformed the clinicoradiological model, with an area under the curve (95% confidence interval) of 0.833 (0.765-0.894) versus 0.764 (0.686-0.833) [p = 0.088, DeLong test], 0.824 (0.718-0.909) versus 0.692 (0.554-0.809) [p = 0.012], and 0.902 (0.794-0.984) versus 0.696 (0.526-0.857) [p = 0.024] in all three cohorts.

CONCLUSIONS

Radiomic features from LNs could provide additional value for predicting pCR in ESCC patients, and the radiomics nomogram provided an accurate prediction of pCR, which might aid treatment decision.