Abdominal CT Images Research Articles

AATCT-IDS: A benchmark Abdominal Adipose Tissue CT Image Dataset for image denoising, semantic segmentation, and radiomics evaluation

Background and objective:The metabolic syndrome induced by obesity is closely associated with cardiovascular disease, and the prevalence is increasing globally, year by year. Obesity is a risk marker for detecting this disease. However, current research on computer-aided detection of adipose distribution is hampered by the lack of open-source large abdominal adipose datasets. Methods:In this study, a benchmark Abdominal Adipose Tissue CT Image Dataset (AATCT-IDS) containing 300 subjects is prepared and published. AATCT-IDS publics 13,732 raw CT slices, and the researchers individually annotate the subcutaneous and visceral adipose tissue regions of 3213 of those slices that have the same slice distance to validate denoising methods, train semantic segmentation models, and study radiomics. For different tasks, this paper compares and analyzes the performance of various methods on AATCT-IDS by combining the visualization results and evaluation data. Thus, verify the research potential of this data set in the above three types of tasks. Results:In the comparative study of image denoising, algorithms using a smoothing strategy suppress mixed noise at the expense of image details and obtain better evaluation data. Methods such as BM3D preserve the original image structure better, although the evaluation data are slightly lower. The results show significant differences among them. In the comparative study of semantic segmentation of abdominal adipose tissue, the segmentation results of adipose tissue by each model show different structural characteristics. Among them, BiSeNet obtains segmentation results only slightly inferior to U-Net with the shortest training time and effectively separates small and isolated adipose tissue. In addition, the radiomics study based on AATCT-IDS reveals three adipose distributions in the subject population. Conclusion:AATCT-IDS contains the ground truth of adipose tissue regions in abdominal CT slices. This open-source dataset can attract researchers to explore the multi-dimensional characteristics of abdominal adipose tissue and thus help physicians and patients in clinical practice. AATCT-IDS is freely published for non-commercial purpose at: https://figshare.com/articles/dataset/AATTCT-IDS/23807256.

The impact of skeletal muscle index at the third lumbar spine on nosocomial deterioration and short-term prognosis in acute pancreatitis: a retrospective observational study.

To investigate the impact of the third lumbar skeletal muscle index (L3-SMI) assessed by CT on the in-hospital severity and short-term prognosis of acute pancreatitis. A total of 224 patients with severe acute pancreatitis admitted to Yantaishan Hospital from January 2021 to June 2022 were selected as the subjects. Based on the in-hospital treatment outcomes, they were divided into a mortality group of 59 cases as well as a survival group of 165 cases. Upon admission, general information such as the Acute Physiology and Chronic Health Evaluation II (APACHE II) score, along with the abdominal CT images of each patient, were analyzed. The L3-SMI was calculated, and the Modified CT Severity Index (MCTSI) and Balthazar CT grade were used to assess the severity of in-hospital complications of acute pancreatitis. The evaluation value of L3-SMI for the prognosis of severe acute pancreatitis was analyzed, as well as the factors influencing the prognosis of severe acute pancreatitis. No statistically significant differences in gender, age, BMI, etiology, duration of anti-inflammatory drug use, and proportion of surgical patients between the survival and mortality groups were observed. But the mortality group showed higher proportions of patients with an elevated APACHE II score upon admission, mechanical ventilation, and renal replacement therapy, compared to the survival group, with statistically significant differences (P < 0.001). Furthermore, the mortality group had higher MCTSI scores (6.42 ± 0.69) and Balthazar CT grades (3.78 ± 0.45) than the survival group, with statistically significant differences (P < 0.001). The mortality group also had a lower L3-SMI (39.68 ± 3.25) compared to the survival group (42.71 ± 4.28), with statistically significant differences (P < 0.001). L3-SMI exhibited a negative correlation with MCTSI scores and Balthazar CT grades (r = -0.889, -0.790, P < 0.001). Logistic regression analysis, with mortality of acute pancreatitis patients as the dependent variable and MCTSI scores, Balthazar CT grades, L3-SMI, APACHE II score upon admission, mechanical ventilation, and renal replacement therapy as independent variables, revealed that MCTSI scores and L3-SMI were risk factors for mortality in acute pancreatitis patients (P < 0.001). Logistic regression analysis using the same variables confirmed that all these factors were risk factors for mortality in acute pancreatitis patients. This study confirmed that diagnosing muscle depletion using L3-SMI is a valuable radiological parameter for predicting in-hospital severity and short-term prognosis in patients with acute pancreatitis.

Diagnostic value of identifying location and amount of free gas in the abdominal cavity by multidetector computed tomography in patients with acute gastrointestinal perforation

Objective: To evaluate the relationships between the location and extent of diffusion of free intraperitoneal air by multi-slice spiral CT (MSCT) and between the location and size of acute gastrointestinal perforation. Methods: This was a descriptive case series. We examined abdominal CT images of 33 patients who were treated for intraoperatively confirmed gastrointestinal perforation (excluding appendiceal perforation) in the Department of General Surgery, Nanfang Hospital between January and September 2022. We identified five locations of intraperitoneal air: the subphrenic space, hepatic portal space, mid-abdominal wall, mesenteric space, and pelvic cavity. We allocated the 33 patients to an upper gastrointestinal perforation (n=23) and lower gastrointestinal perforation group (n=10) base on intraoperative findings and analyzed the relationships between the locations of free gas and of gastrointestinal perforation. Additionally, we established two models for analyzing the extent of diffusion of free gas in the abdominal cavity and constructed receiver operating characteristic (ROC) curves to analyze the relationships between the two models and the size of the gastrointestinal perforation. Results: In the upper gastrointestinal perforation group, free gas was located around the hepatic portal area in 91.3% (21/23) of patients: this is a significantly greater proportion than that found in the lower gastrointestinal perforation group (5/10) (P=0.016). In contrast, free gas was located in the mesenteric interspace in 8/10 patients in the lower gastrointestinal perforation group; this is a significantly greater proportion than was found in the upper gastrointestinal perforation group (8.7%, 2/23) (P<0.010). The sensitivity of diagnosis of upper gastrointestinal perforation base on the presence of hepatic portal free gas was 84.8% and the specificity 71.4%. Further, the sensitivity of diagnosis of lower gastrointestinal perforation base on the presence of mesenteric interspace free gas was 80.0% and the specificity 91.3%. The rates of presence of free gas in the subdiaphragmatic area, mid-abdominal wall, and pelvic cavity did not differ significantly between the two groups (all P>0.05). Receiver operating characteristic curves showed that when free gas was present in four or more of the studied locations in the abdominal cavity, the optimal cutoff for perforation diameter was 2 cm, the corresponding sensitivity 66.7%, and the specificity 100%, suggesting that abdominal free gas diffuses extensively when the diameter of the perforation is >2 cm. Another model revealed that when free gas is present in three or more of the studied locations, the optimal cutoff for perforation diameter is 1 cm, corresponding to a sensitivity of 91.7% and specificity of 76.2%; suggesting that free gas is relatively confined in the abdominal cavity when the diameter of the perforation is <1 cm. Conclusion: Identifying which of five locations in the abdominal cavity contains free intraperitoneal air by examining MSCT images can be used to assist in the diagnosis of the location and size of acute gastrointestinal perforations.

A case of indirect inguinal bladder hernia treated with laparoscopic transabdominal preperitoneal repair with high peritoneal incisional approach

BackgroundInguinal herniation of the urinary bladder is uncommon and those descending into the scrotum are even rarer. Although open anterior repair has been used for inguinal bladder hernia, the efficacy of laparoscopic herniorrhaphy has been reported in recent years.Case presentationA 63-year-old man presented with an irreducible right groin and scrotal bulge associated with voiding difficulty. Abdominal ultrasonography showed a dislocation of the urinary bladder descending into the right scrotum. Abdominal CT imaging revealed that a part of the bladder and small intestine was herniating into the scrotum through the internal inguinal ring and running laterally to the inferior epigastric artery. Under the diagnosis of indirect inguinal bladder hernia, the patient underwent trans-abdominal preperitoneal hernia repair (TAPP). The bladder herniated into the scrotum through the internal inguinal ring was replaced to the original position. Then the myopectineal orifice was exposed and covered with polypropylene mesh, where a horizontal peritoneal incision 4 cm above the hernia orifice, i.e., the high peritoneal incision approach (HPIA), allowed an easy peeling of the peritoneum and hernia sac. The patient’s postoperative course was uneventful and the voiding difficulty resolved. The patient continued to do well without recurrence at 20 months after surgery.ConclusionPreoperative evaluation with abdominal ultrasonography and CT scan allowed a precise diagnosis of a groin hernia with voiding difficulty. TAPP with HPIA was useful in the treatment of inguinal bladder hernia because this technique facilitated a quick confirmation of the hernia contents, secure dissection of the whole protruded bladder, and adequate replacement of the bladder to the original position without any injury.

ResNet50-Boosted UNet for Improved Liver Segmentation Accuracy

Segmentation of the liver from abdominal CT images is difficult due to changes in form, density, and the presence of malignancies. This research describes a novel strategy to improve segmentation accuracy that uses UNet as a foundation architecture and ResNet50 as a backbone architecture. This integrated design automates feature selection and spatial awareness, overcoming limitations in previous models. Experimental evaluations using the LiTS dataset show higher performance. Specifically, using the LiTS dataset, our algorithm achieves a remarkable foreground accuracy of 99.81% in liver segmentation. These results outperform existing approaches, demonstrating UNet and ResNet50's potential as valuable tools for precise liver segmentation in clinical situations. The suggested system shows promise for application in diverse medical imaging tasks other than liver segmentation, demonstrating its versatility and effectiveness in enhancing machine-assisted medical diagnostics and decision-making processes.

Dual-Attention Model Fusing CNN and Transformer for Pancreas Segmentation.

Pancreatic cancer is a highly malignant cancer of the digestive tract and is rapidly progressing and spreading clinically. Automatic and accurate pancreatic tissue segmentation in abdominal CT images is essential for the early diagnosis of pancreatic-related diseases. It is challenging that the pancreas is small in size and complex in morphology. To address this problem, we propose a dual-attention model fusing CNN and Transformer to effectively activate pancreas-related features expression. The CNN structure weights the importance of pancreas-related features at the channel level and weakens the background information. Transformer feature aggregation module constructs spatial correlations among long-distance pixels from a global perspective. This study is validated on the NIH-TCIA dataset and achieved a mean Dice Similarity Coefficient of 85.82%, which is outperforming than the state-of-the-art methods. The visualization of surface distance also demonstrates the effective segmentation of pancreas boundary details by the proposed model.

TBACkp: HER2 expression status classification network focusing on intrinsic subenvironmental characteristics of breast cancer liver metastases

The HER2 expression status in breast cancer liver metastases is a crucial indicator for the diagnosis, treatment, and prognosis assessment of patients. And typical diagnosis involves assessing the HER2 expression status through invasive procedures like biopsy. However, this method has certain drawbacks, such as being difficult in obtaining tissue samples and requiring long examination periods. To address these limitations, we propose an AI-aided diagnostic model. This model enables rapid diagnosis. It diagnoses a patient's HER2 expression status on the basis of preprocessed images, which is the region of the lesion extracted from a CT image rather than from an actual tissue sample. The algorithm of the model adopts a parallel structure, including a Branch Block and a Trunk Block. The Branch Block is responsible for extracting the gradient characteristics between the tumor sub-environments, and the Trunk Block is for fusing the characteristics extracted by the Branch Block. The Branch Block contains CNN with self-attention, which combines the advantages of CNN and self-attention to extract more meticulous and comprehensive image features. And the Trunk Block is so designed that it fuses the extracted image feature information without affecting the transmission of the original image features. The Conv-Attention is used to calculate the attention in the Trunk Block, which uses kernel dot product and is responsible for providing the weight for the self-attention in the process of using convolution induced deviation calculation. Combined with the structure of the model and the method used, we refer to this model as TBACkp. The dataset comprises the enhanced abdominal CT images of 151 patients with liver metastases from breast cancer, together with the corresponding HER2 expression levels for each patient. The experimental results are as follows: (AUC: 0.915, ACC: 0.854, specificity: 0.809, precision: 0.863, recall: 0.881, F1-score: 0.872). The results demonstrate that this method can accurately assess the HER2 expression status in patients when compared with other advanced deep learning model.

Automatic Urinary Stone Detection System for Abdominal Non-Enhanced CT Images Reduces the Burden on Radiologists.

To develop a fully automatic urinary stone detection system (kidney, ureter, and bladder) and to test it in a real clinical environment. The local institutional review board approved this retrospective single-center study that used non-enhanced abdominopelvic CT scans from patients admitted urology (uPatients) and emergency (ePatients). The uPatients were randomly divided into training and validation sets in a ratio of 3:1. We designed a cascade urinary stone map location-feature pyramid networks (USm-FPNs) and innovatively proposed a ureter distance heatmap method to estimate the ureter position on non-enhanced CT to further reduce the false positives. The performances of the system were compared using the free-response receiver operating characteristic curve and the precision-recall curve. This study included 811 uPatients and 356 ePatients. At stone level, the cascade detector USm-FPNs has the mean of false positives per scan (mFP) 1.88 with the sensitivity 0.977 in validation set, and mFP was further reduced to 1.18 with the sensitivity 0.977 after combining the ureter distance heatmap. At patient level, the sensitivity and precision were as high as 0.995 and 0.990 in validation set, respectively. In a real clinical set of ePatients (27.5% of patients contain stones), the mFP was 1.31 with as high as sensitivity 0.977, and the diagnostic time reduced by > 20% with the system help. A fully automatic detection system for entire urinary stones on non-enhanced CT scans was proposed and reduces obviously the burden on junior radiologists without compromising sensitivity in real emergency data.

Sparse deep belief network coupled with extended local fuzzy active contour model-based liver cancer segmentation from abdomen CT images.

Liver cancer from abdominal CT images must be accurately segmented for the purpose of diagnosis with treatment planning. But, the similarity in gray values between the liver and the surrounding tissues poses a challenge. To address this, a novel sparse deep belief network coupled with extended local fuzzy active contour model-based liver cancer segmentation from abdomen CT images (SDBN-ELFAC-LCS-CT) is proposed. This method incorporates dynamic adaptive pooling and residual modules in SDBN to improve the feature selection and generalization ability. Additionally, the 3D reconstruction is performed to refine segmentation results. The proposed SDBN-ELFAC-LCS-CT approach is implemented in MATLAB. The performance of the proposed SDBN-ELFAC-LCS-CT achieves dice coefficients that were up to 96.16% higher and 75.88%, 88.75%, and 71.16% lower. Volumetric overlap error compared with existing models, like basic ensembles of vanilla-style deep learning modes, increases liver segmentation from CT imageries (BEVS-LCS-CT), an incorporated 3 dimensional sparse deep belief network along enriched seagull optimization approach for liver segmentation (3DBN-ESOA-LCS-CT) and iterative convolutional encoder-decoder network and multiple scale context learning for segmenting liver (ICEDN-LCS-CT), respectively.

Segmentation of abdominal CT and cardiac MR images with multi scale visual attention

Segmentation of abdominal CT and cardiac MR images with multi scale visual attention

Quality assessment of abdominal CT images: an improved ResNet algorithm with dual-attention mechanism.

To enhance medical image classification using a Dual-attention ResNet model and investigate the impact of attention mechanisms on model performance in a clinical setting. We utilized a dataset of medical images and implemented a Dual-attention ResNet model, integrating self-attention and spatial attention mechanisms. The model was trained and evaluated using binary and five-level quality classification tasks, leveraging standard evaluation metrics. Our findings demonstrated substantial performance improvements with the Dual-attention ResNet model in both classification tasks. In the binary classification task, the model achieved an accuracy of 0.940, outperforming the conventional ResNet model. Similarly, in the five-level quality classification task, the Dual-attention ResNet model attained an accuracy of 0.757, highlighting its efficacy in capturing nuanced distinctions in image quality. The integration of attention mechanisms within the ResNet model resulted in significant performance enhancements, showcasing its potential for improving medical image classification tasks. These results underscore the promising role of attention mechanisms in facilitating more accurate and discriminative analysis of medical images, thus holding substantial promise for clinical applications in radiology and diagnostics.

Semi-supervised liver segmentation based on local regions self-supervision.

Semi-supervised learning has gained popularity in medical image segmentation due to its ability to reduce reliance on image annotation. A typical approach in semi-supervised learning is to select reliable predictions as pseudo-labels and eliminate unreliable predictions. Contrastive learning helps prevent the insufficient utilization of unreliable predictions, but neglecting the anatomical structure of medical images can lead to suboptimal optimization results. We propose a novel approach for semi-supervised liver segmentation using contrastive learning, which leverages unlabeled data and enhances the suitability of contrastive learning for liver segmentation. Contrastive learning helps prevent the inappropriate utilization of unreliable predictions, but neglecting the anatomical structure of medical images can lead to suboptimal optimization results. Therefore, we propose a semi-supervised contrastive learning method with local regions self-supervision (LRS2). On one side, we employ Shannon entropy to distinguish between reliable and unreliable predictions and reduce the dissimilarity between their representations within regional artificial units. Within each unit of the liver image, we strongly encourage unreliable predictions to acquire valuable information pertaining to the correct category by leveraging the representations of reliable predictions in their vicinity. On the other side, we introduce a dynamic reliability threshold based on the Shannon entropy of each prediction, gradually evaluating the confidence threshold of reliable predictions as predictive accuracy improves. After selecting reliable predictions, we sequentially apply erosion and dilation to refine them for better selection of qualified positive and negative samples. We evaluate our proposed method on abdominal CT images, including 131 images (train data: 77, validation data: 26, and testing data: 28) from 2017 ISBI Liver Tumors Segmentation Challenge. Our method obtains satisfactory performance in different proportion by exploiting the unreliable predictions. Compared with the result of VNet only under supervised settings (with 10, 30, 50, 70% and full labeled data), LRS2, respectively, brings an improvement of Dice coefficient by +6.11, +3.55, +4.43, and +2.25%, achieving Dice coefficients of 93.44, 93.31, 94.85, and 95.12%,respectively. In this study, we carefully select appropriate positive and negative samples from reliable regions, ensuring that anchor pixels within unreliable regions are correctly assigned to their respective categories. With a consideration of the anatomical structure present in CT images, we partition the image representations into regional units, enabling anchor pixels to capture more precise sample information. Extensive experiments confirm the effectiveness of ourmethod.

Findings of MRI, CT scan and ultrasound in Marion's disease in a young woman: a case report

BackgroundPrimary bladder neck obstruction (PBNO) is a rare condition of the lower urinary tract in young and middle-aged patients. PBNO is a urological condition affecting both sexes in which the bladder neck fails to open adequately during voiding, resulting in obstruction of urinary flow in the absence of anatomic obstruction, such as benign prostatic hypertrophy in men or genitourinary prolapse in women. PBNO may present voiding symptoms (decreased force of stream, hesitancy, intermittent stream, incomplete emptying) or irritative symptoms (frequency, urgency, urge incontinence, nocturia) or a combination of both.Case presentationA 31-year-old female without medical history presented to our department with a 4–5 months history of a sense of incomplete emptying bladder and suprapubic discomfort. In Abdominal and Pelvic CT images of her CT scan, evidence of homogenous concentric hypertrophy of the urethra with mild enhancement in the post-contrast image was seen. Also, pelvic MRI shows isointense on T1- weighted and isointense to mildly hyperintense on T2-weighted Images (to skeletal muscle signal) of the concentric hypertrophy muscular layer of the urethra. The post-contrast image demonstrates moderate enhancement in the hypertrophied urethra. A month later, while the patient had not received any treatment, a pelvic ultrasound showed concentric hypertrophy hypoechoic muscular layer of the urethra.DiscussionThe true prevalence of PBNO is unknown but Farrar et al. reported a 2% prevalence of primary bladder neck obstruction.In 1933, Marion described PBNO as caused by fibrous narrowing of the bladder neck and detrusor hyperplasia. Also, Marion described two types of disease (congenital and acquired) and treatment methods. In this study was shown isointense on T1-weighted and isointense to mildly hyperintense on T2-weighted Images of concentric hypertrophy muscular layer of urethra with moderate enhancement on the post-contrast image. A doughnut sign was also seen in our axial slices (particularly in T2-weighted images), so this sign seems to be a diagnostic finding in MRI. In our study, the diffusion sequence and ADC map showed no obvious restriction.ConclusionsAlthough PBNO has been described for decades, there is still a controversy about diagnosis and treatment. In patients suspected of PBNO, imaging methods, especially MRI, can help confirm the diagnosis or rule out other causes of bladder neck obstruction. In future studies, imaging modalities can be one of the diagnostic criteria for PBNO.

Kidney Tumor Segmentation through Semi-supervised Learning Based on Mean Teacher Using Kidney Local Guided Map in Abdominal CT Images

Kidney Tumor Segmentation through Semi-supervised Learning Based on Mean Teacher Using Kidney Local Guided Map in Abdominal CT Images

Elevating healthcare through artificial intelligence: analyzing the abdominal emergencies data set (TR_ABDOMEN_RAD_EMERGENCY) at TEKNOFEST-2022.

The artificial intelligence competition in healthcare at TEKNOFEST-2022 provided a platform to address the complex multi-class classification challenge of abdominal emergencies using computer vision techniques. This manuscript aimed to comprehensively present the methodologies for data preparation, annotation procedures, and rigorous evaluation metrics. Moreover, it was conducted to introduce a meticulously curated abdominal emergencies data set to the researchers. The data set underwent a comprehensive central screening procedure employing diverse algorithms extracted from the e-Nabız (Pulse) and National Teleradiology System of the Republic of Türkiye, Ministry of Health. Full anonymization of the data set was conducted. Subsequently, the data set was annotated by a group of ten experienced radiologists. The evaluation process was executed by calculating F1 scores, which were derived from the intersection over union values between the predicted bounding boxes and the corresponding ground truth (GT) bounding boxes. The establishment of baseline performance metrics involved computing the average of the highest five F1 scores. Observations indicated a progressive decline in F1 scores as the threshold value increased. Furthermore, it could be deduced that class 6 (abdominal aortic aneurysm/dissection) was relatively straightforward to detect compared to other classes, with class 5 (acute diverticulitis) presenting the most formidable challenge. It is noteworthy, however, that if all achieved outcomes for all classes were considered with a threshold of 0.5, the data set's complexity and associated challenges became pronounced. This data set's significance lies in its pioneering provision of labels and GT-boxes for six classes, fostering opportunities for researchers. The prompt identification and timely intervention in cases of emergent medical conditions hold paramount significance. The handling of patients' care can be augmented, while the potential for errors is minimized, particularly amidst high caseload scenarios, through the application of AI. • The data set used in artificial intelligence competition in healthcare (TEKNOFEST-2022) provides a 6-class data set of abdominal CT images consisting of a great variety of abdominal emergencies. • This data set is compiled from the National Teleradiology System data repository of emergency radiology departments of 459 hospitals. • Radiological data on abdominal emergencies is scarce in literature and this annotated competition data set can be a valuable resource for further studies and new AI models.

Abstract 284: Renal Contrast Excretion After Myelography: A Valuable Tool in The Diagnosis of Cerebrospinal‐Venous Fistulae

Introduction Cerebrospinal‐venous fistulae (CSF‐VFs) represent a pathologic, low‐flow direct connection between the intrathecal space, usually at the level of a diverticular dilation of a nerve root sleeve, and the thoracolumbar venous system. They can cause CSF hypovolemia resulting in an often‐debilitating intracranial hypotension syndrome. Lateral decubitus CT or DSA myelography (CTM, DSM) are the gold‐standard modalities for diagnosing these entities. Their high spatial and temporal resolution allows visualization of the abnormal connection between the two spaces. Yet, the small size and low flow make CSF‐VFs radiographically subtle, and may not be consistently visualized even after a correctly‐performed CTM/DSM, requiring repeated invasive procedures to definitively establish or exclude their presence. Opacification of the renal excretory system 30‐60 minutes after lateral decubitus myelograms is thought to be a highly sensitive and specific indirect sign of the presence of a CSF‐VF 1,2, as this amount of time is considered insufficient to allow arachnoid reabsorption of contrast. Renal pelvises hyperdensity or its absence after a negative decubitus CTM/DSM can prompt further myelographic investigation or definitively rule out the presence of a CSF‐VF on the side of the examination, respectively. Methods We present three patients with severely symptomatic CSF‐VFs treated at our hospital between 2021 and 2023. All patients were females (range: 50‐60 years old). All three patients underwent bilateral diagnostic lateral decubitus DSM/CTM. Exams were performed at least 72 hours apart in order to allow complete clearance of intrathecal contrast. After each myelographic acquisition, patients remained in the lateral decubitus position until abdominal CT imaging was performed at 30 and 45 minutes to detect the presence of contrast in the renal collecting system. Results Patients 1 and 2, had a CSF‐VF between nerve root dural diverticula and the azygos system that was directly visualized on lateral decubitus CTM (patient 1: right T10‐T11, patient 2: left T8‐T9). Contrast was detectable in the renal pelvises after myelography was performed on the affected side, and not detectable after the contralateral investigation. One fistula was treated via CT‐guided targeted blood patch and fibrin glue injection and the other via endovascular Onyx embolization of the tributary Intercostal veins. Patient 3 had an extremely slow flow CSF‐VF at the right T11‐12 level that was identified only after 5 right‐sided decubitus myelograms with both modalities. The repeated exams were based on the presence of contrast in the renal pelvises after each right‐sided procedure and on the absence of renal opacification after the left‐sided myelograms. Eventually, a CSF‐VF was identified. Successful Onyx embolization of the foraminal veins of the R T11‐12 levels was performed. All three patients had complete resolution of symptoms. Conclusion Contrast presence in the renal excretory system after a lateral decubitus CTM/DSM is a potentially highly sensitive and specific indirect diagnostic sign that can establish or exclude the presence of a CSF‐VF on the examined side. It could be particularly helpful when the fistula is not initially detected, as it can prompt aggressive continuation of the myelographic work‐up. Similarly, it can exclude its presence on one side, sparing the patient invasive and low yield diagnostic procedures.

FRI205 Splenic Infarct: A Rare Presentation Of Cushing Syndrome

Abstract Disclosure: G. Virella: None. Splenic Infarct: A Rare Presentation of Cushing Syndrome Geover, Virella M.D. 1,2, Carlos Santiago Botero Suarez, M.D.1,2, Alfredo Palomino Abella, M.D. 1,2, Mustafa Kinaan, M.D. 1,2 1) University of Central Florida HCA Healthcare GME, Greater Orlando, Florida2) Department of Internal Medicine, University of Central Florida College of Medicine, Orlando, Florida Disclaimer: This research was supported (in whole or in part) by HCA Healthcare and/or an HCA Healthcare affiliated entity. The views expressed in this publication represent those of the author(s) and do not necessarily represent the official views of HCA Healthcare or any of its affiliated entities.Introduction Cushing’s syndrome (CS) is associated with a hypercoagulable state leading to increased incidence of thromboembolic events. The mechanism behind this phenomenon is due to an imbalance between activity of pro- and anticoagulant pathways. We report a case of newly diagnosed CS in a patient presenting with acute splenic and inferior vena cava (IVC) infarcts. Case Description A 59-year-old woman with no significant medical history presents with acute left upper quadrant pain. She denied other symptoms. Physical examination was remarkable for tachycardia, left upper quadrant tenderness, and centripetal adiposity with BMI of 19. Labs showed marked leukocytosis (WBC 26 x 109/L) and thrombocytopenia (platelet 96,000 cells/uL). She was also found to be diabetic (HgA1C 12.5%). Abdominal CT imaging showed a wedge-shaped 4.3 x 3.1 x 2.3 cm splenic infarction, peri-splenic abscess, and thrombosis of the left renal and gonadal veins extending into the IVC. Additionally, there was a left adrenal incidentaloma measuring 4.5 x 3.8 x 2.9 cm. MRI of the left adrenal gland was consistent with an adenoma. The right adrenal gland was unremarkable. Screening for pheochromocytoma and hyperaldosteronism was normal. A morning serum cortisol was 33.4 mcg/dL, and ACTH level of &amp;lt; 1.5 pg/mL. An overnight 1-mg dexamethasone suppression test revealed post-dexamethasone cortisol of 4.58 mcg/dL. These findings were consistent with Cushing syndrome secondary to adrenal cause. The patient was initially treated with heparin infusion and underwent a successful thrombectomy of the IVC and gonadal vein. She later underwent a successful laparoscopic left adrenalectomy and splenectomy. Pathology reports confirmed an adrenocortical adenoma. A random post-surgical cortisol was 22.8. The patient had an uneventful postoperative course and was discharged home with outpatient endocrine follow-up planned to evaluate for hypercortisolism resolution and management of diabetes. Conclusion Cushing Syndrome patients have been shown to have an almost 18-fold higher rate of venous thromboembolism (VTE). This is an often-overlooked complication of hypercortisolism. The pathogenesis is not fully understood but is believed to be related to shortening of activated partial thromboplastin time (aPTT), increased clot lysis time, and glucocorticoid-induced increase in procoagulant factors such as fibrinogen, von Willebrand factor, and factor VIII. Our case will closely demonstrate the link between hypercortisolism and hypercoagulability. Presentation: Friday, June 16, 2023

Automatic liver tumor segmentation from CT images using random forest algorithm

Automatic liver segmentation is challenging, and the tumor segmenting process adds more complexity. Based on the grey levels and shape, separating the liver and tumor from abdominal CT images is critical. In our paper suggests a more effective approach by using Gabor features (GF) to segment liver tumors from CT images and three alternative neural network algorithms to address these problems: RF, CNN and ANN. This thesis uses the same collection of classifiers and GF to first segment a variety of Gabor liver images. The organ (liver) is then extracted from an abdominal CT image using liver segmentation, which is done by three classifiers: ANN, CNN, RF trained on Gabor filter and the tumor segmentation is done by the human visual system (HVS). For pixel-wise segmentation, reliable and accurate ML techniques were used. For the liver segmentation, the classification accuracy was 99.55, 97.88 and 98.13% for RF, CNN and ANN, respectively. From the extracted image of liver, the classification accuracy for tumor was 99.52, 98.07 and 98.45% for RF, CNN and ANN, respectively.

Comparison of Nonneurological Structures at Risk During Anterior-to-Psoas Versus Transpsoas Surgical Approaches Using Abdominal CT Imaging From L1 to S1.

The kidneys, ribs, liver, spleen, and iliac crests can pose access-related issues to the disc space during both anterior-to-psoas (ATP) and transpsoas (TP) surgical approaches. The aim of this study was to identify and compare the presence and degree of obstruction caused by these structures for the ATP and TP approaches bilaterally from L1 to S1 using abdominal computed tomography. Presence of obstruction by a given structure was recorded if the structure was within ATP or TP borders. Degree of obstruction was calculated as the quotient of the structure measurement within the ATP or TP approach divided by the entire corridor length at the point of obstruction. The percentage of time the left kidney was present during the ATP vs TP approaches at L1 to L2 was 44% vs 89% (P < 0.001), at L2 to L3 was 26% vs 75% (P < 0.001), and at L3 to L4 was 5% vs 19% (P < 0.001). For the right kidney, these values were 37% vs 78% (P < 0.001), 43% vs 71% (P < 0.001), and 11% vs 18% (P < 0.001). The percentage of time the left rib was present during ATP vs TP approaches was 41% vs 81% (P < 0.001) at L1 to L2 and 11% vs 26% (P = 0.413) at L2 to L3. With respect to the liver, the ATP approach was obstructed 56%, 30%, and 9% of the time at the levels of L1 to L2, L2 to L3, and L3 to L4; the liver was not present in L1 to L4 TP approach. This study is the first to both characterize and compare nonneurological structures at risk during ATP and TP fusion approaches bilaterally from L1 to S1 using abdominal computed tomography. Findings suggest the ATP approach poses less structures at risk relative to the TP approach with respect to the kidneys, ribs, and iliac crests bilaterally. The TP approach offers advantages compared with ATP approach with respect to the liver and spleen. Findings from this study are clinically relevant for ATP and TP surgical approach planning.

Hybrid CNN model employing patch-based exemplar for accessory spleen detection in abdominal CT images

Hybrid CNN model employing patch-based exemplar for accessory spleen detection in abdominal CT images