Background: The accuracy of estimating microvascular invasion (MVI) preoperatively in hepatocellular carcinoma (HCC) by clinical observers is low. Most recent studies constructed MVI predictive models utilizing radiological and/or radiomics features extracted from computed tomography (CT) images. These methods, however, rely heavily on human experiences and require manual tumor contouring. We developed a deep learning-based framework for preoperative MVI prediction by using CT images of arterial phase (AP) with simple tumor labeling and without the need of manual feature extraction. The model was further validated on CT images that were originally scanned at multiple different hospitals. Methods: CT images of AP were acquired for 309 patients from China Medical University Hospital (CMUH). Images of 164 patients, who took their CT scanning at 54 different hospitals but were referred to CMUH, were also collected. Deep learning (ResNet-18) and machine learning (support vector machine) models were constructed with AP images and/or patients’ clinical factors (CFs), and their performance was compared systematically. All models were independently evaluated on two patient cohorts: validation set (within CMUH) and external set (other hospitals). Subsequently, explainability of the best model was visualized using gradient-weighted class activation map (Grad-CAM). Results: The ResNet-18 model built with AP images and patients’ clinical factors was superior than other models achieving a highest AUC of 0.845. When evaluating on the external set, the model produced an AUC of 0.777, approaching its performance on the validation set. Model interpretation with Grad-CAM revealed that MVI relevant imaging features on CT images were captured and learned by the ResNet-18 model. Conclusions: This framework provide evidence showing the generalizability and robustness of ResNet-18 in predicting MVI using CT images of AP scanned at multiple different hospitals. Attention heatmaps obtained from model explainability further confirmed that ResNet-18 focused on imaging features on CT overlapping with the conditions used by radiologists to estimate MVI clinically.

The ingestion of foreign bodies such as coins, fish bones, plastic toy parts, batteries, and needles is common in children. Although the majority of ingested foreign bodies pass through the gastrointestinal tract unaided, some children require either nonsurgical or surgical intervention. The medical records of children who presented to the pediatric emergency department of a single tertiary referral center between December 2001 and May 2006 were reviewed. A total of 87 patients underwent an endoscopic procedure because of suspected foreign body ingestion and foreign bodies were identified by endoscopy in 74 patients (85.1%). The mean age of these 74 patients was 3.4 years (range, 6 months to 13 years). The most common site of foreign body lodgment was the esophagus (n=38,51.4%); other sites included the stomach (n=33,44.6%) and duodenum (n=3, 4.0%). The types of foreign bodies included coins (n=42, 56.8%), button batteries (n = 16, 21.6%), sharp objects (n = 9, 12.2%), chicken bones (n = 2, 2.7%) and others (n=5, 6.7%). Only two foreign bodies (button batteries) in the duodenum could not be removed successfully by endoscopy. Instead, they were moved into the intestine and then eliminated spontaneously the following day. There were no major complications caused by foreign body ingestion or endoscopic procedures. The outcome of all patients was uneventful without morbidity or mortality. In our experience, endoscopic removal of foreign bodies under general anesthesia is an effective and safe method in children; the method also prevents erosion and perforation of the gastrointestinal tract.

Right paraduodenal hernia is an unusual congenital malformation that causes intestinal obstruction. It is seldom diagnosed preoperatively and a small bowel follow-through series is the best diagnostic method. However, multidetector computed tomography (MDCT) can be an alternative diagnostic method. We report the first case of right paraduodenal hernia that was diagnosed preoperatively by MDCT. A 15-year-old boy presented with right lower abdominal pain, nausea and vomiting. Abdominal MDCT with coronal reformation images clearly revealed encapsulated small bowel loops in the right side of the abdomen and displacement of the ascending colon to the left side. He received surgical repair and recovered uneventfully.

Right lower quadrant (RLQ) pain is a common complaint in children presenting at emergency departments. This study analyzed the etiologies of RLQ pain, and compared the clinical presentations, laboratory test results and imaging findings in patients with appendicitis with those in other groups of patients. We also investigated if active observation resulted in delayed diagnosis, to the detriment of patients. Medical records for the period January 2006 to July 2006 were reviewed for children (age < 18 years) who presented to the emergency department of one medical center, complaining of RLQ pain. Out of a total of 100 patients (age range: 2–17 years; mean: 11 years), 46 patients presented with only one symptom of RLQ pain, while 32 patients had ≥ 2 associated signs or symptoms, including fever, nausea/vomiting, diarrhea, or rebound tenderness. Imaging studies, including abdominal sonography and/or computed tomography, were performed in 73 patients; 44 underwent surgery for presumed appendicitis and one received surgery for a right paraduodenal hernia. Eleven patients underwent surgery because of peritoneal signs, and eight because of persistent or aggravated RLQ pain. Postoperative pathologic examinations revealed 53 cases of appendicitis, six normal appendices, and other morbidities (1 perforated peptic ulcer, 1 pelvic inflammatory disease, 1 ovarian cyst, 1 diverticulitis, and 1 right paraduodenal hernia). Thirty‐three patients were discharged after several hours of observation (range: 0.5–18 hours; mean: 4 hours), and three patients were admitted for further observations. All were discharged without operation. There were significant differences in the incidences of fever (p = 0.004) and rebound tenderness (p = 0.019), and in white cell counts (p < 0.001), neutrophil percentages (p < 0.001), and C‐reactive protein levels (p < 0.001) between patients with appendicitis and patients with other causes of RLQ pain. Clinical signs and symptoms, laboratory tests, and imaging studies can be used to differentiate between the causes of RLQ pain. Patients without the classical features of appendicitis or peritonitis can be safely managed by active observation and repeated physical examinations.

Although multiparametric magnetic resonance imaging (mpMRI) is widely used to assess the volume of prostate cancer, it often underestimates the histological tumor boundary. The aim of this study was to evaluate the feasibility of combining prostate health index (PHI) and mpMRI to estimate the histological tumor diameter and determine the safety margin during treatment of prostate cancer. We retrospectively enrolled 72 prostate cancer patients who underwent radical prostatectomy and had received PHI tests and mpMRI before surgery. We compared the discrepancy between histological and radiological tumor diameter stratified by Prostate Imaging-Reporting and Data System (PI-RADS) score, and then assessed the influence of PHI on the discrepancy between low PI-RADS (2 or 3) and high PI-RADS (4 or 5) groups. The mean radiological and histological tumor diameters were 1.60 cm and 2.13 cm, respectively. The median discrepancy between radiological and histological tumor diameter of PI-RADS 4 or 5 lesions was significantly greater than that of PI-RADS 2 or 3 lesions (0.50 cm, IQR (0.00–0.90) vs. 0.00 cm, IQR (−0.10–0.20), p = 0.02). In the low PI-RADS group, the upper limit of the discrepancy was 0.2 cm; so the safety margin could be set at 0.1 cm. In the high PI-RADS group, the upper limits of the discrepancy were 1.2, 1.6, and 2.2 cm in men with PHI < 30, 30–60, and > 60; so the safety margin could be set at 0.6, 0.8, and 1.1 cm, respectively. Radiological tumor diameter on mpMRI often underestimated the histological tumor diameter, especially for PI-RADS 4 or 5 lesions. Combining mpMRI and PHI may help to better estimate the histological tumor diameter.