Gastric cancer remains a significant global health concern, coercing the need for advancements in imaging techniques for ensuring accurate diagnosis and effective treatment planning. Artificial intelligence (AI) has emerged as a potent tool for gastric-cancer imaging, particularly for diagnostic imaging and body morphometry. This review article offers a comprehensive overview of the recent developments and applications of AI in gastric cancer imaging. We investigated the role of AI imaging in gastric cancer diagnosis and staging, showcasing its potential to enhance the accuracy and efficiency of these crucial aspects of patient management. Additionally, we explored the application of AI body morphometry specifically for assessing the clinical impact of gastrectomy. This aspect of AI utilization holds significant promise for understanding postoperative changes and optimizing patient outcomes. Furthermore, we examine the current state of AI techniques for the prognosis of patients with gastric cancer. These prognostic models leverage AI algorithms to predict long-term survival outcomes and assist clinicians in making informed treatment decisions.However, the implementation of AI techniques for gastric cancer imaging has several limitations. As AI continues to evolve, we hope to witness the translation of cutting-edge technologies into routine clinical practice, ultimately improving patient care and outcomes in the fight against gastric cancer.

Background/Aims: Posthepatectomy liver failure (PHLF) is a major complication that increases mortality in patients with hepatocellular carcinoma after surgical resection. The aim of this retrospective study was to evaluate the utility of magnetic resonance elastography-assessed liver stiffness (MRE-LS) for the prediction of PHLF and to develop an MRE-LS-based risk prediction model. Methods: A total of 160 hepatocellular carcinoma patients who underwent surgical resection with available preoperative MRE-LS data were enrolled. Clinical and laboratory parameters were collected from medical records. Logistic regression analyses were conducted to identify the risk factors for PHLF and develop a risk prediction model. Results: PHLF was present in 24 patients (15%). In the multivariate logistic analysis, high MRE-LS (kPa; odds ratio [OR] 1.49, 95% confidence interval [CI] 1.12 to 1.98, p=0.006), low serum albumin (≤3.8 g/dL; OR 15.89, 95% CI 2.41 to 104.82, p=0.004), major hepatic resection (OR 4.16, 95% CI 1.40 to 12.38, p=0.014), higher albumin-bilirubin score (>–0.55; OR 3.72, 95% CI 1.15 to 12.04, p=0.028), and higher serum α-fetoprotein (>100 ng/mL; OR 3.53, 95% CI 1.20 to 10.40, p=0.022) were identified as independent risk factors for PHLF. A risk prediction model for PHLF was established using the multivariate logistic regression equation. The area under the receiver operating characteristic curve (AUC) of the risk prediction model was 0.877 for predicting PHLF and 0.923 for predicting grade B and C PHLF. In leave-one-out cross-validation, the risk model showed good performance, with AUCs of 0.807 for all-grade PHLF and 0. 871 for grade B and C PHLF. Conclusions Our novel MRE-LS-based risk model had excellent performance in predicting PHLF, especially grade B and C PHLF.

This study evaluated the test-retest repeatability and measurement variability of ultrasonographic shear wave elastography (SWE) for liver stiffness in a rat liver fibrosis model. Methods: In 31 Sprague-Dawley rats divided into three groups (high-dose, low-dose, and control), liver fibrosis was induced by intraperitoneal administration of thioacetamide for 8 weeks. A dedicated radiographer performed SWE to measure liver stiffness in kilopascals in two sessions at a 3-day interval. We calculated correlations between liver stiffness and histopathologic results, measurement variability in each session using coefficients of variation (CoVs) and interquartile/median (IQR/M), and test-retest repeatability between both sessions using the repeatability coefficient. Results: Different levels of liver fibrosis in each group were successfully induced in the animal model. The mean liver stiffness values were 8.88±1.48 kPa in the control group, 11.62±1.70 kPa in the low-dose group, and 11.91±1.73 kPa in the high-dose group. The correlation between collagen areas and liver stiffness values was moderate (r=0.6). In all groups, the second session yielded lower CoVs (i.e., more reliable results) for liver stiffness than the first session, suggesting a training effect for the operator. The mean IQR/M values were also lower in the second session than in the first session, which had four outliers (0.21 vs. 0.12, P<0.001). The test-retest repeatability coefficient was 3.75 kPa and decreased to 2.82 kPa after removing the four outliers. Conclusion: The use of ultrasonographic SWE was confirmed to be feasible and repeatable for evaluating liver fibrosis in preclinical trials. Operator training might reduce variability in liver stiffness measurements.

OBJECTIVE: To test the feasibility of non-contrast magnetic resonance imaging (MRI) in a sequential imaging study for characterizing computed tomography (CT)-detected arterial-enhancing nodules that do not washout in patients at risk of hepatocellular carcinoma (HCC).MATERIALS AND METHODS: In this retrospective study, 134 patients (mean age ± standard deviation, 56.8 ± 10.0 years) with 151 arterial enhancing-only nodules measuring up to 2 cm during multiphasic CT that were subsequently evaluated using gadoxetic acid-enhanced MRI in treatment-naïve at-risk patients from three tertiary referral centers were included. Tentative diagnostic criteria for HCC and hepatic malignancy were defined as the presence of one of eight MRI features favoring HCC in combinations of the following sequences: T2-weighted imaging (T2WI), diffusion-weighted imaging (DWI), T1-weighted dual gradient-echo in-phase and out-of-phase imaging (Dual-GRE), and hepatobiliary phase imaging (HBP). Typical hemangiomas and arterioportal shunts were excluded from the analysis. Diagnostic performance for HCC and hepatic malignancy was calculated and compared between the abbreviated MRI and full-sequence gadoxetic acid-enhanced MRI.RESULTS: Of 151 nodules (mean size, 1.2 cm) 68 HCCs and 83 non-HCC benignities and malignancies were included. The combination of T2WI, DWI, and Dual-GRE showed per-lesion sensitivity, specificity, and accuracy of 88.2%, 90.4%, and 89.4%, respectively, comparable to those of full-sequence MRI. Applying the same sequence combination to diagnose hepatic malignancy had per-lesion sensitivity, specificity, and accuracy of 86.8%, 97.3%, and 92.1%. In nodules < 1 cm, adding HBP increased sensitivity by up to 13% without compromising the specificity or accuracy.CONCLUSION: The non-contrast MRI protocol comprising T2WI, DWI, and Dual-GRE showed reasonable and comparable performance to full-sequence MRI for discriminating HCC and primary liver malignancies in CT-detected indeterminate arterial enhancing-only nodules in at-risk patients, and can be potentially used for sequential imaging in place of a full-sequence MRI. In nodules < 1 cm, HBP may still be needed to preserve sensitivity.
Carcinoma, Hepatocellular ; Hemangioma ; Humans ; Liver ; Magnetic Resonance Imaging ; Retrospective Studies ; Sensitivity and Specificity ; Tertiary Care Centers

Abdominal Muscles ; Adipose Tissue ; Artificial Intelligence ; Dataset ; Intra-Abdominal Fat ; Learning ; Muscle, Skeletal ; Muscles ; Sarcopenia ; Spine ; Subcutaneous Fat ; Tomography, X-Ray Computed

PURPOSE: The purpose of this study was to evaluate the prognostic value of skeletal muscle depletion measured on computed tomography (CT) in patients with non-metastatic invasive breast cancer. METHODS: This retrospective study included 577 consecutive women (mean age ± standard deviation: 48.9 ± 10.2 years with breast cancer who underwent a preoperative positron-emission tomography (PET)/CT scan and curative surgery between January 2012 and August 2014. The total abdominal muscle area (TAMA), subcutaneous fat area (SFA), and visceral fat area (VFA) were measured on CT images at the L3 vertebral level. Univariate and multivariate Cox proportional-hazard regression analyses were performed to evaluate whether there was an association between sarcopenia and overall survival (OS) outcome. RESULTS: Of the 577 women, 49 (8.5%) died after a mean of 46 months. The best TAMA threshold for predicting OS was 83.7 cm². The multivariate Cox proportional-hazard analysis revealed that sarcopenia (TAMA ≤ 83.70 cm²) was a strong prognostic biomarker (hazard ratio [HR], 1.951; 95% confidence interval [CI], 1.061–3.586), along with large tumor size, axillary lymph node metastasis, high nuclear grade, estrogen receptor status, and adjuvant radiation therapy. In the subgroup analysis of patients aged ≥ 50 years, TAMA (≤ 77.14 cm²) was a significant independent factor (HR, 2.856; 95% CI, 1.218–6.695). CONCLUSION Skeletal muscle depletion measured on CT was associated with worse OS outcome in patients with non-metastatic breast cancer.

Recently, sarcopenia has garnered renewed interest. Sarcopenia is a disease characterized by decreased skeletal muscle mass and strength/function, which can impair the quality of life and increase physical disability, adverse metabolic effects, and mortality. Imaging tools for evaluating and diagnosing sarcopenia have developed rapidly. Radiologists should be aware of sarcopenia and its clinical implications. We review current knowledge about sarcopenia, its pathophysiological impact, and advantages and disadvantages of methods for evaluation of sarcopenia focusing on body composition imaging modalities such as whole-body dual-energy X-ray absorptiometry, CT, and MRI. Controversial issues are discussed, including the lack of consensus and standardization of the disease definition, imaging modality, measurement methods, and diagnostic cutoff points.
Absorptiometry, Photon ; Body Composition ; Consensus ; Magnetic Resonance Imaging ; Mortality ; Muscle, Skeletal ; Quality of Life ; Sarcopenia

OBJECTIVE: The reliability of size measurements of liver metastases from neuroendocrine tumors (NETs) on contrast-enhanced computed tomography (CT) phases made by different readers may be hampered due to transient, variable rim enhancement in arterial phase (AP) or portal venous phase (PVP) images. We aimed to assess the reliability of tumor size measurements in pre- and post-contrast scans. MATERIALS AND METHODS: The study coordinator selected target lesions according to Response Evaluation Criteria in Solid Tumors 1.1 guidelines in 44 consecutive patients with pathologically confirmed NET liver metastases. Two blinded readers measured the longest diameters of target lesions on pre-contrast, AP, and PVP images twice with a 4-week interval. Inter- and intra-observer agreements were evaluated using Bland-Altman plots and 95% limit of agreement (LOA) calculations. RESULTS: Of the 79 target lesions (approximate mean size of 3 cm), 45 showed rim enhancement. Inter-observer agreement assessed based on LOA was highest in pre-contrast CT images (−6.1–5.7 mm), followed by PVP (−7.9–7.1 mm) and AP (−8.5–7.4 mm) images. Intra-observer agreement showed the same trend: −2.8–2.9 mm and −2.9–2.9 mm for readers 1 and 2, respectively, on pre-contrast CT, −2.8–2.9 mm and −3.0–3.2 mm, respectively, on PVP, and −3.2–4.2 mm and −3.4–3.2 mm, respectively, on AP images. Mean tumor diameters differed significantly among the phases in the following increasing order: pre-contrast CT, PVP, and AP images. CONCLUSION: There was better inter- and intra-observer agreement in size measurements of NET liver metastases on precontrast scans than on AP and PVP scans. Pre-contrast CT may be the optimal for measuring NET liver metastases if its accuracy is proven.
Humans ; Liver* ; Loa ; Neoplasm Metastasis ; Neuroendocrine Tumors* ; Response Evaluation Criteria in Solid Tumors

OBJECTIVE: To compare the breathing effects on dynamic contrast-enhanced (DCE)-MRI between controlled aliasing in parallel imaging results in higher acceleration (CAIPIRINHA)-volumetric interpolated breath-hold examination (VIBE), radial VIBE with k-space-weighted image contrast view-sharing (radial-VIBE), and conventional VIBE (c-VIBE) sequences using a dedicated phantom experiment. MATERIALS AND METHODS: We developed a moving platform to simulate breathing motion. We conducted dynamic scanning on a 3T machine (MAGNETOM Skyra, Siemens Healthcare) using CAIPIRINHA-VIBE, radial-VIBE, and c-VIBE for six minutes per sequence. We acquired MRI images of the phantom in both static and moving modes, and we also obtained motion-corrected images for the motion mode. We compared the signal stability and signal-to-noise ratio (SNR) of each sequence according to motion state and used the coefficients of variation (CoV) to determine the degree of signal stability. RESULTS: With motion, CAIPIRINHA-VIBE showed the best image quality, and the motion correction aligned the images very well. The CoV (%) of CAIPIRINHA-VIBE in the moving mode (18.65) decreased significantly after the motion correction (2.56) (p < 0.001). In contrast, c-VIBE showed severe breathing motion artifacts that did not improve after motion correction. For radial-VIBE, the position of the phantom in the images did not change during motion, but streak artifacts significantly degraded image quality, also after motion correction. In addition, SNR increased in both CAIPIRINHA-VIBE (from 3.37 to 9.41, p < 0.001) and radial-VIBE (from 4.3 to 4.96, p < 0.001) after motion correction. CONCLUSION: CAIPIRINHA-VIBE performed best for free-breathing DCE-MRI after motion correction, with excellent image quality.
Acceleration ; Artifacts ; Magnetic Resonance Imaging ; Respiration* ; Signal-To-Noise Ratio

OBJECTIVE: To prospectively evaluate the performance of diffusion-weighted imaging (DWI) to monitor bowel inflammation after medical therapy for Crohn's disease (CD). MATERIALS AND METHODS: Before and following 1–2 years of medical therapy, between October 2012 and May 2015, 18 randomly selected adult CD patients (male:female, 13:5; mean age ± SD, 25.8 ± 7.9 years at the time of enrollment) prospectively underwent MR enterography (MRE) including DWI (b = 900 s/mm²) and ileocolonoscopy. Thirty-seven prospectively defined index lesions (one contiguous endoscopy-confirmed inflamed area chosen from each inflamed anatomical bowel segment; 1–4 index lesions per patient; median, 2 lesions) were assessed on pre- and post-treatment MRE and endoscopy. Visual assessment of treatment responses on DWI in 4 categories including complete remission and reduced, unchanged or increased inflammation, and measurements of changes in apparent diffusion coefficient (ΔADC), i.e., pre-treatment–post-treatment, were performed by 2 independent readers. Endoscopic findings and CD MRI activity index (CDMI) obtained using conventional MRE served as reference standards. RESULTS: ΔADC significantly differed between improved (i.e., complete remission and reduced inflammation) and unimproved (i.e., unchanged or increased inflammation) lesions: mean ± SD (× 10⁻³ mm²/s) of -0.65 ± 0.58 vs. 0.06 ± 0.15 for reader 1 (p = 0.022) and -0.68 ± 0.56 vs. 0.10 ± 0.26 for reader 2 (p = 0.025). DWI accuracy for diagnosing complete remission or improved inflammation ranged from 76% (28/37) to 84% (31/37). A significant negative correlation was noted between ΔADC and ΔCDMI for both readers with correlation coefficients of -0.438 and -0.461, respectively (p < 0.05). CONCLUSION: DWI is potentially a feasible tool to monitor quantitatively and qualitatively bowel inflammation of CD after medical treatment.
Adult ; Crohn Disease* ; Diffusion ; Endoscopy ; Follow-Up Studies ; Humans ; Inflammation* ; Longitudinal Studies* ; Magnetic Resonance Imaging ; Prospective Studies*