Novel Strategy for the Management of Cervical Multicystic Diseases.

PURPOSE: To investigate the clinical practices of diagnosing multicystic cervical lesions as a means to develop a more appropriate diagnostic algorithm for gastric-type adenocarcinoma (GAS) and its precursors. METHODS: Clinical information for 159 surgically treated patients for multicystic disease of the uterine cervix was collected from 15 hospitals. We performed a central review of the MRI and pathological findings. The MRI findings were categorized into four types including two newly proposed imaging features based on the morphology and distribution of cysts, and the diagnosis accuracy was assessed. Among the four MRI types, types 1 and 2 were categorized as benign lesions that included LEGH; type 3 were precancerous lesions (with an assumption of atypical LEGH); and type 4 were malignant lesions. RESULTS: The central pathological review identified 56 cases of LEGH, seven with GAS, four with another form of carcinoma, and 92 with benign disease. In clinical practice, over-diagnosis of malignancy (suspicion of MDA) occurred for 12/19 cases (63.2%) and under-diagnosis of malignancy occurred for 4/11 (36%). Among the 118 patients who had a preoperative MRI and underwent a hysterectomy, type 3 or 4 MRI findings in conjunction with abnormal cytology were positively indicative of premalignancy or malignancy, with a sensitivity and specificity of 61.1% and 96.7%, respectively. CONCLUSIONS: Although the correct preoperative diagnosis of cervical cancer with a multicystic lesion is challenging, the combination of cytology and MRI findings creates a more appropriate diagnostic algorithm that significantly improves the diagnostic accuracy for differentiating benign disease from premalignancy and malignancy.

Image Quality and Lesion Detectability of Pancreatic Phase Thin-Slice Computed Tomography Images With a Deep Learning-Based Reconstruction Algorithm.

OBJECTIVE: To evaluate the image quality and lesion detectability of pancreatic phase thin-slice computed tomography (CT) images reconstructed with a deep learning-based reconstruction (DLR) algorithm compared with filtered-back projection (FBP) and hybrid iterative reconstruction (IR) algorithms. METHODS: Fifty-three patients who underwent dynamic contrast-enhanced CT including pancreatic phase were enrolled in this retrospective study. Pancreatic phase thin-slice (0.625 mm) images were reconstructed with each FBP, hybrid IR, and DLR. Objective image quality and signal-to-noise ratio of the pancreatic parenchyma, and contrast-to-noise ratio of pancreatic lesions were compared between the 3 reconstruction algorithms. Two radiologists independently assessed the image quality of all images. The diagnostic performance for the detection of pancreatic lesions was compared among the reconstruction algorithms using jackknife alternative free-response receiver operating characteristic analysis. RESULTS: Deep learning-based reconstruction resulted in significantly lower image noise and higher signal-to-noise ratio and contrast-to-noise ratio than hybrid IR and FBP ( P < 0.001). Deep learning-based reconstruction also yielded significantly higher visual scores than hybrid IR and FBP ( P < 0.01). The diagnostic performance of DLR for detecting pancreatic lesions was highest for both readers, although a significant difference was found only between DLR and FBP in one reader ( P = 0.02). CONCLUSIONS: Deep learning-based reconstruction showed improved objective and subjective image quality of pancreatic phase thin-slice CT relative to other reconstruction algorithms and has potential for improving lesion detectability.

Diagnostic accuracy of MRI for evaluating myometrial invasion in endometrial cancer: a comparison of MUSE-DWI, rFOV-DWI, and DCE-MRI.

OBJECTIVES: To compare the image quality of high-resolution diffusion-weighted imaging (DWI) using multiplexed sensitivity encoding (MUSE) versus reduced field-of-view (rFOV) techniques in endometrial cancer (EC) and to compare the diagnostic performance of these techniques with that of dynamic contrast-enhanced (DCE) MRI for assessing myometrial invasion of EC. METHODS: MUSE-DWI and rFOV-DWI were obtained preoperatively in 58 women with EC. Three radiologists assessed the image quality of MUSE-DWI and rFOV-DWI. For 55 women who underwent DCE-MRI, the same radiologists assessed the superficial and deep myometrial invasion using MUSE-DWI, rFOV-DWI, and DCE-MRI. Qualitative scores were compared using the Wilcoxon signed-rank test. Receiver operating characteristic analysis was performed to compare the diagnostic performance. RESULTS: Artifacts, sharpness, lesion conspicuity, and overall quality were significantly better with MUSE-DWI than with rFOV-DWI (p < 0.05). The area under the curve (AUC) of MUSE-DWI, rFOV-DWI, and DCE-MRI for the assessment of myometrial invasion were not significantly different except for significantly higher AUC of MUSE-DWI than that of DCE-MRI for superficial myometrial invasion (0.76 for MUSE-DWI and 0.64 for DCE-MRI, p = 0.049) and for deep myometrial invasion (0.92 for MUSE-DWI and 0.80 for DCE-MRI, p = 0.022) in one observer, and that of rFOV-DWI for deep myometrial invasion in another observer (0.96 for MUSE-DWI and 0.89 for rFOV-MRI, p = 0.048). CONCLUSION: MUSE-DWI exhibits better image quality than rFOV-DWI. MUSE-DWI and rFOV-DWI shows almost equivalent diagnostic performance compared to DCE-MRI for assessing superficial and deep myometrial invasion in EC although MUSE-DWI may be helpful for some radiologists.

New trend in artificial intelligence-based assistive technology for thoracic imaging.

Although there is no solid agreement for artificial intelligence (AI), it refers to a computer system with intelligence similar to that of humans. Deep learning appeared in 2006, and more than 10 years have passed since the third AI boom was triggered by improvements in computing power, algorithm development, and the use of big data. In recent years, the application and development of AI technology in the medical field have intensified internationally. There is no doubt that AI will be used in clinical practice to assist in diagnostic imaging in the future. In qualitative diagnosis, it is desirable to develop an explainable AI that at least represents the basis of the diagnostic process. However, it must be kept in mind that AI is a physician-assistant system, and the final decision should be made by the physician while understanding the limitations of AI. The aim of this article is to review the application of AI technology in diagnostic imaging from PubMed database while particularly focusing on diagnostic imaging in thorax such as lesion detection and qualitative diagnosis in order to help radiologists and clinicians to become more familiar with AI in thorax.

Clinical applications of artificial intelligence in liver imaging.

This review outlines the current status and challenges of the clinical applications of artificial intelligence in liver imaging using computed tomography or magnetic resonance imaging based on a topic analysis of PubMed search results using latent Dirichlet allocation. LDA revealed that "segmentation," "hepatocellular carcinoma and radiomics," "metastasis," "fibrosis," and "reconstruction" were current main topic keywords. Automatic liver segmentation technology using deep learning is beginning to assume new clinical significance as part of whole-body composition analysis. It has also been applied to the screening of large populations and the acquisition of training data for machine learning models and has resulted in the development of imaging biomarkers that have a significant impact on important clinical issues, such as the estimation of liver fibrosis, recurrence, and prognosis of malignant tumors. Deep learning reconstruction is expanding as a new technological clinical application of artificial intelligence and has shown results in reducing contrast and radiation doses. However, there is much missing evidence, such as external validation of machine learning models and the evaluation of the diagnostic performance of specific diseases using deep learning reconstruction, suggesting that the clinical application of these technologies is still in development.

New response evaluation criteria using early morphological change in imatinib treatment for patients with gastrointestinal stromal tumor.

BACKGROUND: The introduction of molecularly targeted drugs, including imatinib, has greatly improved the prognosis of gastrointestinal stromal tumor (GIST), and based on the different response image, the methods of response evaluation have been established for GISTs. Furthrmore, the best response evaluation using them has been reported to be associated with progression-free survival (PFS) in imatinib treatment. However, since it is more important to predict the clinical outcomes of imatinib treatment in "early treatment phase", new predicting factor in earlier stage is desired to work out the whole strategy of each patient. Early morphological change (EMC) was previously reported as a predictive marker for molecularly targeted drugs in metastatic colorectal cancer. The purpose of the present study was to verify the efficacy of EMC in predicting the outcome in patients with GIST receiving imatinib at early evaluation. METHODS: We retrospectively reviewed 66 patients. EMC in computed tomography (CT) image was evaluated, and the patients were categorized into two groups: active MR (morphological response) (+) group and active MR (-) group. We investigated the association between the presence of active MR and clinical outcomes. RESULTS: Forty-five patients had active MR ( +). The median progression-free survival (PFS) in patients with/without active MR was 49/23 months (P = 0.0039). CONCLUSION: The evaluation criteria based on EMC could be a sensitive method to predict the clinical outcome of imatinib treatment for patients with unresectable GIST.

MRI of Borderline Epithelial Ovarian Tumors: Pathologic Correlation and Diagnostic Challenges.

Borderline epithelial ovarian tumors are a distinct pathologic entity characterized by increased epithelial proliferation and nuclear atypia, but without frank stromal invasion. Borderline tumor (BT) is now considered to represent an intermediate phase in the stepwise progression from benign to malignant ovarian epithelial tumor. Since BTs commonly manifest at early stages in women of reproductive age and are associated with a good prognosis, making the correct diagnosis is important in determining whether a patient is a candidate for fertility-sparing surgery. There are six histologic BT subtypes (serous, mucinous, seromucinous, endometrioid, clear cell, and Brenner), and each has different MRI features, reflecting their unique histologic architectures. Radiologists should be aware of the MRI features that can suggest BTs. These features include a hyperintense papillary architecture with hypointense internal branching, which can be observed with serous and seromucinous BTs on T2-weighted images; aggregates of microcysts that have hypointensity on T2-weighted images and reticular enhancement on contrast-enhanced T2-weighted images, which can be seen with mucinous BTs; and moderately high signal intensity on diffusion-weighted images along with relatively high apparent diffusion coefficient values, which can be observed regardless of the histologic subtype. Nevertheless, because the imaging features of BTs overlap with those of many benign lesions (eg, cystadenoma and cystadenofibroma, decidualized endometriosis, and polypoid endometriosis) and malignant tumors (ovarian cancers and metastases), histologic confirmation is required for the final diagnosis. Special emphasis is placed on the MRI features of BTs, pathologic correlation, and the challenges related to diagnosis. ©RSNA, 2022.

The prediction of postoperative anorectal dysfunction after low anterior resection for lower rectal cancer by measuring the volume of defecation-related muscles.

PURPOSE: Approximately 90% of patients are thought to develop bowel dysfunction after low anterior resection (LAR). Although some prognostic factors have been reported, the volumes of defecation-related muscles have not been examined. This retrospective study investigated the association between the preoperative volume of defecation-related muscles and major LARS. METHODS: Forty-six patients who underwent LAR for rectal cancer between 2013 and 2020 in our institution were analyzed. They had no local residual tumor or local recurrence at the time of the study and no problems with their defecation function pre-operatively. Defecation-related muscle volume measurements were made before surgery, and the patients answered a low anterior resection syndrome (LARS) score questionnaire after surgery. The muscle volume was adjusted by the patient's height squared. RESULTS: Twenty-seven patients (58.7%) developed major LARS. In the univariate analysis, sex, lateral lymph node dissection, and diverting ileostomy as well as muscle volume of the external anal sphincter, pubococcygeal + iliococcygeus muscle, and puborectal muscle were associated with major LARS. In a multivariate analysis, pubococcygeal + iliococcygeus muscle (< 5.96 ml/m2) was the only factor (p = 0.02). CONCLUSIONS: Measuring the volume of the defecation-related muscles may aid in predicting major LARS.

Efficacy of positron emission tomography in diagnosis of lateral lymph node metastases in patients with rectal Cancer: a retrospective study.

BACKGROUND: The presence of lateral pelvic lymph node (LLN) metastasis is an essential prognostic factor in rectal cancer patients. Thus, preoperative diagnosis of LLN metastasis is clinically important to determine the therapeutic strategy. The aim of this study was to evaluate the efficacy of preoperative positron emission tomography/computed tomography (PET/CT) in the diagnosis of LLN metastasis. METHODS: Eighty-four patients with rectal cancer who underwent LLN dissection at Osaka University were included in this study. The maximum standardized uptake value (SUVmax) of the primary tumor and LLN were preoperatively calculated using PET/CT. Simultaneously, the short axis of the lymph node was measured using multi-detector row computed tomography (MDCT). The presence of metastases was evaluated by postoperative pathological examination. RESULTS: Of the 84 patients, LLN metastases developed in the left, right, and both LLN regions in 6, 7, and 2 patients, respectively. The diagnosis of the metastases was predicted with a sensitivity of 82%, specificity of 93%, positive predictive value of 58%, negative predictive value of 98%, false positive value of 7%, and false negative value of 18% when the cutoff value of the LLN SUVmax was set at 1.5. The cutoff value of the short axis set at 7 mm on MDCT was most useful in diagnosing LLN metastases, but SUVmax was even more useful in terms of specificity. CONCLUSIONS: The cutoff value of 1.5 for lymph node SUVmax in PET is a reasonable measure to predict the risk of preoperative LLN metastases in rectal cancer patients.

Comparison of HASTE with multiple signal averaging versus conventional turbo spin echo sequence: a new option for T2-weighted MRI of the female pelvis.

OBJECTIVES: This study was conducted in order to evaluate the feasibility of a newly developed half-Fourier acquisition single-shot turbo spin echo (HASTE) sequence for T2-weighted MRI of the female pelvis, in which scanning and post-processing filtering techniques were optimized to minimize blurring and the resultant reduction in signal-to-noise ratio (SNR) was compensated by signal averaging of multiple excitations (multi-NEX), and to compare its image quality with that of a conventional turbo spin echo (TSE) sequence. METHODS: Fifty females with gynecological disease were prospectively enrolled. They underwent pelvic MRI including sagittal T2-weighted imaging obtained with multi-NEX HASTE (mHASTE) and TSE with acquisition times of 1 m 9 s and 3 m 26 s, respectively. At the prospective reading, mHASTE was reviewed first, and any changes in the image interpretation detected during the subsequent TSE reading were recorded. The SNR and tissue contrast were quantitatively measured. Two radiologists independently assessed tissue conspicuity and overall image quality considering noise, artifacts, and blurring, and visual grading characteristics (VGC) analysis was performed to compare the two sequences. RESULTS: Adding TSE to mHASTE did not change image interpretations in any case. mHASTE produced an equivalent SNR and significantly higher contrast between the myometrium and junctional zone (p < 0.05) compared with TSE. The area under the VGC curve (AUCVGC) showed that the conspicuity of the uterus, ovaries, and lesions and the overall image quality were significantly higher on mHASTE than on TSE for both readers (AUCVGC, 0.740-0.880; p < 0.01). CONCLUSION: mHASTE is a feasible option for T2-weighted MRI of the female pelvis, as it allows faster and higher-quality image acquisition than conventional TSE. KEY POINTS: • High-quality T2-weighted imaging is essential for magnetic resonance imaging of the female pelvis. • A new HASTE sequence was developed, in which a parallel imaging technique, a wide bandwidth, and a post-processing filter were used to minimize blurring, and the resultant reduction in the signal-to-noise ratio was compensated using signal averaging of multiple excitations (multi-NEX). • Multi-NEX HASTE allows fast and high-quality T2-weighted image acquisition.

Assessment of Myometrial Invasion in Premenopausal Grade 1 Endometrial Carcinoma: Is Magnetic Resonance Imaging a Reliable Tool in Selecting Patients for Fertility-Preserving Therapy?

OBJECTIVES: The aim of this study was to evaluate the diagnostic ability of magnetic resonance imaging (MRI) in premenopausal women with G1 endometrial carcinoma. METHODS: Twenty-six patients underwent T2W, diffusion weighted, and dynamic contrast-enhanced 3-T MRI. The degree of myometrial invasion was pathologically classified into no invasion, shallow (3 mm or less), and more. Two radiologists assessed myometrial invasion on MRI. Diagnostic accuracy, sensitivity, specificity, positive and negative predictive values, AUC, and interobserver agreement were analyzed. RESULTS: For assessing myometrial invasion, mean accuracy, sensitivity, specificity, positive predictive values, negative predictive values, and AUC, respectively, were as follows: 63%, 42%, 85%, 79%, 47%, and 0.75. Mean interobserver agreement was fair (k = 0.36). Shallow invasions were underestimated as no invasion on MRI in all 6 cases. CONCLUSIONS: Magnetic resonance imaging produced false-negative result on half of patients. The misjudgments tended to happen in patients with shallow invasion.

Experimental studies on artifacts and tumor enhancement on gadoxetic acid-enhanced arterial phase liver MRI in a rabbit VX2 tumor model.

Background Rapid injection of gadoxetic acid is reported to produce more frequent artifacts and lower vascular enhancement on arterial phase liver magnetic resonance imaging (MRI). However, its effect on tumor enhancement and the mechanism of the artifacts remain unclear. Purpose To evaluate the effect of rapid injection of gadoxetic acid on artifacts and tumor enhancement during arterial phase liver MRI, and on arterial blood gases (ABGs) which may explain the cause of the artifacts. Material and Methods ABG analysis was performed in 13 free-breathing rabbits after rapid injection (1 mL/s; injection time = 0.6-0.8 s) of gadoxetic acid (0.025 mmol/kg). Dynamic liver MRI was performed in six anesthetized rabbits with VX2 tumors under a ventilation stoppage after rapid and slow injection (0.25 mL/s; injection time = 2.4-3.2 s) of gadoxetic acid. Artifacts and signal enhancement on arterial phase imaging were compared with those obtained after rapid injection of gadopentetic acid (Gd-DTPA, 0.1 mmol/kg) using a Friedman test or Kruskal-Wallis test. Results ABG analysis did not find any significant changes. Artifacts were not related to injection protocols ( P = 0.95). Aortic enhancement with slow injection of gadoxetic acid was significantly higher than that with rapid injection ( P < 0.05), and was comparable to that with Gd-DTPA injection. Tumor enhancement obtained with gadoxetic acid was not significantly different between rapid and slow injection, and was significantly lower than that with Gd-DTPA injection ( P < 0.05). Conclusion Rapid injection of gadoxetic acid did not affect ABGs and may not be the cause of the artifacts. It lowered vascular enhancement but not arterial tumor enhancement.

Preoperative staging of endometrial cancer using reduced field-of-view diffusion-weighted imaging: a preliminary study.

OBJECTIVES: To compare the image quality and diagnostic performance of reduced field-of-view (rFOV) versus conventional full field-of-view (fFOV) diffusion-weighted (DW) imaging of endometrial cancer. METHODS: Fifty women with endometrial cancer underwent preoperative rFOV and fFOV DW imaging. Two radiologists compared the image qualities of both techniques, and five radiologists assessed superficial and deep myometrial invasion using both techniques. The statistical analysis included the Wilcoxon signed-rank test and paired t-test for comparisons of image quality and mean diagnostic values. RESULTS: Distortion, tumour delineation, and overall image quality were significantly better with rFOV DW imaging, compared to fFOV DW imaging (P < 0.05); however, the former was inferior in noise (P < 0.05). Regarding superficial invasion, the mean accuracies of the techniques did not differ statistically (rFOV, 58.0% versus fFOV, 56.0%; P = 0.30). Regarding deep myometrial invasion, rFOV DW imaging yielded significantly better mean accuracy, specificity, and positive predictive values (88.4%, 97.8%, and 91.7%, respectively), compared with fFOV DW imaging (84.8%, 94.1%, and 77.4%, respectively; P = 0.009, 0.005, and 0.011, respectively). CONCLUSIONS: Compared with fFOV DW imaging, rFOV DW imaging yielded less distortion, improved image quality and, consequently, better diagnostic performance for deep myometrial invasion of endometrial cancer. KEY POINTS: • rFOV DWI yields better assessment of deep myometrial invasion in endometrial cancer. • rFOV DWI could not sufficiently evaluate superficial invasion in endometrial cancer. • Distortion, tumour delineation, and overall image quality were improved with rFOV DWI.

MRI Findings of Inverted Urothelial Papilloma of the Bladder.

OBJECTIVE: The objective of this study was to characterize MRI findings of inverted urothelial papilloma of the bladder. MATERIALS AND METHODS: Data pertaining to 16 patients with 18 pathologically proven inverted urothelial papillomas of the bladder who had undergone MRI were retrospectively collected from seven institutions. The shape and surface characteristics of the tumors were evaluated using T2-weighted MR images. In addition, the signal intensity of inverted urothelial papillomas was visually assessed on T1-weighted, T2-weighted, and DW images and on early and delayed phase contrast-enhanced images. RESULTS: The shape of the 18 inverted urothelial papillomas of the bladder was classified as polypoid with a stalk for 16 tumors (89%) and polypoid without a stalk for two tumors (11%). All stalks were surrounded by urine in the bladder. A total of 15 of the tumor surfaces (83%) were nonpapillary and three (17%) were papillary. All 12 of the inverted urothelial papillomas for which evaluable T1-weighted images were available were isointense with the bladder wall. The lesions had a slightly higher signal intensity than the bladder wall in 15 of the patients (83%) and showed isointensity with the bladder wall in three patients (17%). A total of three patients (17%) had tiny hyperintense foci noted on T2-weighted images. All 16 of the inverted urothelial papillomas examined by DWI had very high signal intensity. All 13 of the lesions for which early phase images were obtained using dynamic contrast-enhanced MRI showed strong enhancement. When compared with early phase images, delayed phase images of the same 13 lesions showed that enhancement was stronger in two lesions (15%), similar in six lesions (46%), and weaker in five lesions (38%). CONCLUSION: On MRI, the typical appearance of inverted urothelial papillomas of the bladder is a polypoid shape with a nonpapillary surface and a thin short stalk surrounded by urine. Cystic foci are also occasionally seen within the tumor.

Magnetic resonance imaging findings in atypical polypoid adenomyoma.

OBJECTIVES: This article describes magnetic resonance imaging (MRI) findings in 3 cases of atypical polypoid adenomyoma (APAM). METHODS: Clinical and MRI manifestations of 3 patients with APAM were evaluated. High b value diffusion-weighted and dynamic contrast-enhanced images were performed. The size, shape, site of origin, and signal intensity (SI) of MRI findings were evaluated. RESULTS: All patients (age range, 37-47 years; mean age, 40 years) had a chief complaint of atypical genital bleeding with no history of pregnancy. In 2 cases, cytology of the endometrium was positive, and pathological analysis of curettage specimens indicated endometrioid adenocarcinoma. The MRI revealed an endometrial polypoid mass arising from the upper corpus (50%) or lower uterine segment (50%). Except for 1 tumor that seemed to invade the myometrium of the uterine wall due to its irregular margin, all tumors were well circumscribed. On T2-weighted MRI scans, the SI of the tumor was homogenous and marginally hyperintense, and contained markedly hyperintense cystic foci. On T1-weighted images, all the tumors were isointense relative to the myometrium, and in 1 case, a number of cystic foci showed high SIs. In a dynamic contrast-enhanced study, solid portions other than the cystic foci showed contrast enhancement in the arterial phase and demonstrated a washout pattern or plateau pattern in the late phase. All tumors demonstrated high SIs on diffusion-weighted images. CONCLUSIONS: Differentiating APAM from other malignant polypoid uterine endometrial tumors may still pose difficulties, especially when the tumor seems to invade the myometrium on MRI. Comprehensive clinical information about the patient including age, pathological assessment of the endometrium, and MRI findings should be essential to provide indication for a diagnosis of APAM.

Low-Dose Pelvic Computed Tomography Using Adaptive Iterative Dose Reduction 3-Dimensional Algorithm: A Phantom Study.

OBJECTIVE: To evaluate the image quality and radiation dose reduction in pelvic computed tomography (CT) achieved with an adaptive iterative dose reduction 3-dimensional (AIDR 3D) algorithm using a phantom model. METHODS: Two phantoms were scanned using a 320-detector row CT scanner with 8 tube current levels, and the images were reconstructed with a standard filtered back projection (FBP) algorithm and with an AIDR 3D algorithm. RESULTS: Compared with FBP, AIDR 3D reduced image noise and improved contrast-to-noise ratios. The diagnostic performance for detection of low-contrast targets of AIDR 3D images obtained with 100 mA at 120 kVp was almost as good as that of the FBP images obtained with 200 mA. CONCLUSIONS: The AIDR 3D algorithm substantially reduced image noise and improved the image quality of pelvic CT images compared with those obtained with the FBP algorithm and can thus be considered a promising technique for low-dose pelvic CT examinations.

A case of gastric-type adenocarcinoma of the uterine cervix associated with lobular endocervical glandular hyperplasia: radiologic-pathologic correlation.

Gastric-type adenocarcinoma (GAS) of the uterine cervix is a recently defined subtype of mucinous adenocarcinoma. GAS is proposed to include minimal deviation adenocarcinoma (MDA) as a very well-differentiated form and has been suggested to arise from lobular endocervical glandular hyperplasia (LEGH). We report the magnetic resonance imaging (MRI) findings of a GAS associated with LEGH. On MRI, the LEGH component was detected as multiple cystic lesions arranged in a "cosmos pattern", while the GAS was depicted as a predominantly solid lesion containing obvious adenocarcinoma and MDA components, which appeared as mass-like and infiltrative components, respectively. The GAS exhibited tiny cysts on three-dimensional T2-weighted images, high intensity on diffusion-weighted images mostly due to T2 shine-through effect according to apparent diffusion coefficient (ADC) map, and reticular enhancement on dynamic contrast-enhanced MRI, which reflected numerous dilated glandular structures of the tumor. Low ADC was only observed at the deepest invasion front of the obvious adenocarcinoma component. Our case suggests that the MRI features of GAS vary depending on the tumor's histological components, and it is important to be aware of these imaging features when evaluating LEGH on MRI.

Hepatic and pancreatic extracellular volume fraction analysis using contrast-enhanced CT in patients with diabetes mellitus and pre-diabetes.

PURPOSE: Liver and pancreatic fibrosis is associated with diabetes mellitus (DM), and liver fibrosis is associated with pancreatic fibrosis. This study aimed to investigate the relationship between the hepatic and pancreatic extracellular volume fractions (fECVs), which correlate with tissue fibrosis, and their relationships with DM and pre-DM (pDM). MATERIAL AND METHODS: We included 100 consecutive patients with known or suspected liver and/or pancreatic diseases who underwent contrast-enhanced CT. Patients were classified as nondiabetes, pDM, and DM with hemoglobin A1c (HbA1c) levels of < 5.7%, 5.7%-6.5%, and ≥ 6.5% or fasting plasma glucose (FPG) levels of < 100, 100-125 mg/dL, and ≥ 126 mg/dL, respectively. Subtraction images between unenhanced and equilibrium-phase images were prepared. The liver and the pancreas were automatically extracted using a high-speed, three-dimensional image analysis system, and their respective mean CT values were calculated. The enhancement degree of the aorta (Δaorta) was measured. fECV was calculated using the following equation: fECV = (100 - hematocrit) * Δliver or pancreas/Δaorta. Differences were investigated in hepatic and pancreatic fECVs among the three groups, and the correlation between each two in hepatic fECV, pancreatic fECV, and HbA1c was determined. RESULTS: The pancreatic fECV, which was positively correlated with the hepatic fECV and HbA1c (r = 0.51, P < 0.001, and r = 0.51, P < 0.001, respectively), significantly differed among the three groups (P < 0.001) and was significantly greater in DM than in pDM or nondiabetes and in pDM with nondiabetes (P < 0.001). Hepatic fECV was significantly greater in DM than in nondiabetes (P < 0.05). CONCLUSION: The pancreatic fECV and pDM/DM are closely related.

Fairness of artificial intelligence in healthcare: review and recommendations.

In this review, we address the issue of fairness in the clinical integration of artificial intelligence (AI) in the medical field. As the clinical adoption of deep learning algorithms, a subfield of AI, progresses, concerns have arisen regarding the impact of AI biases and discrimination on patient health. This review aims to provide a comprehensive overview of concerns associated with AI fairness; discuss strategies to mitigate AI biases; and emphasize the need for cooperation among physicians, AI researchers, AI developers, policymakers, and patients to ensure equitable AI integration. First, we define and introduce the concept of fairness in AI applications in healthcare and radiology, emphasizing the benefits and challenges of incorporating AI into clinical practice. Next, we delve into concerns regarding fairness in healthcare, addressing the various causes of biases in AI and potential concerns such as misdiagnosis, unequal access to treatment, and ethical considerations. We then outline strategies for addressing fairness, such as the importance of diverse and representative data and algorithm audits. Additionally, we discuss ethical and legal considerations such as data privacy, responsibility, accountability, transparency, and explainability in AI. Finally, we present the Fairness of Artificial Intelligence Recommendations in healthcare (FAIR) statement to offer best practices. Through these efforts, we aim to provide a foundation for discussing the responsible and equitable implementation and deployment of AI in healthcare.