Current Medical Imaging - Volume 20, Issue 1, 2024

Recent advances in deep learning have aided the well-being business in Medical Imaging of numerous disorders like brain tumours, a serious malignancy caused by unregulated and aberrant cell portioning. The most frequent and widely used machine learning algorithm for visual learning and image identification is CNN.

In this article, the convolutional neural network (CNN) technique is used. Augmentation of data and processing of images is used to classify scan imagery of brain MRI as malignant or benign. The performance of the proposed CNN model is compared with pre-trained models: VGG-16, ResNet-50, and Inceptionv3 using the technique which is transfer learning.

Even though the experiment was conducted on a relatively limited dataset, the experimental results reveal that the suggested scratched CNN model accuracy achieved is 94%, VGG-16 was extremely effective and had a very low complexity rate with an accuracy of 90%, whereas ResNet- 50 reached 86% and Inception v3 obtained 64% accuracy.

When compared to previous pre-trained models, the suggested model consumes significantly less processing resources and achieves significantly higher accuracy outcomes and a reduction in losses.

We aimed to evaluate the effectiveness of high-dose and low-dose use of radioactive material in intraoperative gamma probe application methods in patients diagnosed with primary hyperparathyroidism and planned for surgery.

47 patients with primary hyperparathyroidism underwent minimally-invasive parathyroid surgery (MIS) after preoperative imaging studies consisting of ultrasonography (USG) and sestamibi parathyroid scintigraphy (SPS) showed a possible primary parathyroid lesion (PPL). All patients received Tc-99 sestamibi on day-of-surgery imaging (DOSI) to help with the localization of a primary parathyroid lesion (PPL) via both DOSI and intraoperative gamma probe (IGP). Patients in Group I were administered 20-25 mCi Tc-99m sestamibi (methoxy isobutyl isonitrile) and images were obtained at the 20th and 120th minutes. Patients in Group II were administered 8-10 mCi doses and images were obtained at the 20th and 40th minutes. Two nuclear medicine specialists independently evaluated the images. Lesions in the localizations determined by DOSI and IGP were compared with the histopathological results of these lesions.

47 patients, 35 females, and 12 males were included in the study. The mean age of 28 patients in the first group given the high dose was 54 (41-60), and the mean age was 48 (42-57) in the second group given the low dose (p=0.011). In the group given low-dose radioactive material during intraoperative gamma probe application, the observer's sensitivity, specificity, positive, and negative predictive values for finding pathology were 61.1, 100, 100, and 12.5, respectively. In the group given high-dose radioactive material, the same values were 90.9, 33.3, 50, and 83.3, respectively. While the success of MIS increases with the use of DOSI and IGP in large lesions, the success decreases with the prolongation of the accumulation time of the given dose.

In the intraoperative gamma probe technique used in primary hyperparathyroidism patients, the method used with low-dose radioactive material has lower sensitivity but higher specificity in estimating the post-operative pathology compared to the high-dose technique.

The etiology of rotator cuff tears (RCTs) have been investigated for years and many underlying causes have been identified. Shoulder joint morphology is one of the extrinsic causes of RCTs.

Morphometric measurements on MRI sections determined which parameters are an important indicator of RCT in patients with shoulder pain. The aim of this study was to determine the risk factors in the etiology of RCTs by evaluating the shoulder joint morphology with the help of previously defined radiological parameters.

Between January 2019-December 2020, 408 patients (40-70 years old) who underwent shoulder MRI and met the criteria were included in the study. There were 202 patients in the RCT group and 206 patients in the control group. Acromion type, acromial index (AI), critical shoulder angle (CSA), acromiohumeral distance (AHD), lateral acromial angle (LAA), acromial angulation (AA), acromion-greater tuberosity impingement index (ATI), and glenoid version angle (GVA) were measured from the MRI images of the patients.

AI (0.64 vs. 0.60, p = 0.003) CSA (35.3° vs. 32.4°, p = 0.004), ATI (0.91 vs. 0.83, P < 0.001), and AA (13.6° vs. 11.9°, p = 0.011) values were higher in the RCT group than in the control group and the difference was significant. AHD (8.1 mm vs. 9.9 mm, P < 0.001), LAA (77.2° vs. 80.9°, p = 0.004) and GVA (-3.9° vs. -2.5°, P < 0.001) values were lower in the RCT group than in the control group, and again the difference was significant. According to the receiver operating characteristic curve analysis, the cutoff values were 0.623 for AI and 0.860 for ATI.

Acromion type, AI, CSA, AHD, LAA, AA, ATI, and GVA are suitable radiological parameters to evaluate shoulder joint morphology. High AI, CSA, AA, ATI, GVA and low AHD and LAA are risk factors for RCT.

The objective of this study was to describe the CT and MRI features of sclerosing angiomatoid nodular transformation (SANT) of the spleen with pathologic correlation.

Ten patients with surgically resected and pathologically confirmed SANTs were included. Clinical history was reviewed, and gross pathologic, histologic, and immunohistochemical findings were recorded. CT and MRI examinations were evaluated by two radiologists.

Patients included seven men and three women, with a mean age of 42.9±16.7 years. Pathologic features of SANTs involved multiple angiomatous nodules in a radiating pattern with a central stellate fibrous scar and evidence of hemosiderin deposition. 9 cases showed a lobulated demarcated margin, 8 cases a slight hypoattenuating, 1 isoattenuating, and 1 case with two lesions demonstrated a slight hyperattenuating margin, respectively. Multiple scattered punctate calcifications were involved in 2 cases. 5 cases manifested hypointensity on in-phase imaging, 1 iso-intensity, and 4 iso-hypointensity on out-of-phase imaging. Progressive and centripetal enhancement were exhibited in 10 cases, spoke-wheel pattern in 3 cases, and nodular enhancement in 4 cases, respectively. The central fibrous scar was identified in 8 cases during delayed enhancement.

Characteristics of SANTs on CT/MRI reflected the underlying pathology. Hypointensity on DWI and T2WI, and change of signal on T1 chemical-shift imaging were found to be due to hemosiderin deposition and fibrous tissue. Typical feature was a solitary, round, lobulated mass with a fibrous scar. Progressive and centripetal enhancement, spoke-wheel pattern, nodular enhancement, and delayed enhancement of central fibrous scar were observed.

The majority of the existing diagnostic reference levels (DRLs) that have been established for computed tomography (CT) are based on various anatomical locations, such as the head, chest, abdomen, etc. However, DRLs are initiated to improve radiation protection by conducting a comparison of similar examinations with similar objectives. The aim of this study was to explore the feasibility of establishing dose baselines based on common CT protocols for patients who underwent enhanced CT abdomen and pelvis exams.

Dose length product total (tDLPs), volumetric CT dose index (CTDIvol), size-specific dose estimate (SSDE), effective dose (E), and scan acquisition parameters for a total of 216 adult patients, who underwent an enhanced CT abdomen and pelvis exams over a one-year period, were obtained and retrospectively analyzed. Spearman coefficient and one-way ANOVA tests were used to check significant differences between dose metrics and the different CT protocols.

The data exhibited 9 different CT protocols to acquire an enhanced CT abdomen and pelvis exam at our institute. Out of these, 4 were found more common, i.e., CT protocols were acquired for a minimum of 10 cases. Triphasic liver demonstrated the highest mean and median tDLPs across all 4 CT protocols. Triphasic liver protocol registered the highest E followed by gastric sleeve protocol with a mean of 28.7 and 24.7 mSv, respectively. Significant differences (p < 0.0001) were found between the tDLPs of anatomical location and the CT protocol.

Evidently, wide variability exists across CT dose indices and patient dose metrics relying on anatomical-based dose baseline, i.e., DRLs. Patient dose optimizations require establishing dose baselines based on CT protocols rather than the anatomical location.

Pancreatic cancer is one of the most serious problems that has taken many lives worldwide. The diagnostic procedure using the traditional approaches was manual by visually analyzing the large volumes of the dataset, making it time-consuming and prone to subjective errors. Hence the need for the computer-aided diagnosis system (CADs) emerged that comprises the machine and deep learning approaches for denoising, segmentation and classification of pancreatic cancer.

There are different modalities used for the diagnosis of pancreatic cancer, such as Positron Emission Tomography/Computed Tomography (PET/CT), Magnetic Resonance Imaging (MRI), Multiparametric-MRI (Mp-MRI), Radiomics and Radio-genomics. Although these modalities gave remarkable results in diagnosis on the basis of different criteria. CT is the most commonly used modality that produces detailed and fine contrast images of internal organs of the body. However, it may also contain a certain amount of gaussian and rician noise that is necessary to be preprocessed before segmentation of the required region of interest (ROI) from the images and classification of cancer.

This paper analyzes different methodologies used for the complete diagnosis of pancreatic cancer, including the denoising, segmentation and classification, along with the challenges and future scope for the diagnosis of pancreatic cancer.

Various filters are used for denoising and image smoothening and filters as gaussian scale mixture process, non-local means, median filter, adaptive filter and average filter have been used more for better results.

In terms of segmentation, atlas based region-growing method proved to give better results as compared to the state of the art whereas, for the classification, deep learning approaches outperformed other methodologies to classify the images as cancerous and non- cancerous. These methodologies have proved that CAD systems have become a better solution to the ongoing research proposals for the detection of pancreatic cancer worldwide.

The aim of this study was to assess the value of 3.0T magnetic resonance (MR) Diffusion tensor imaging (DTI) in the diagnosis of lumbosacral nerve root compression.

The radiology reports, and clinical records of 34 patients with nerve root compression caused by lumbar disc herniation or bulging and 21 healthy volunteers who had undergone magnetic resonance imaging (MRI) and DTI scan were retrospectively reviewed. The differences in fractional anisotropy (FA) and apparent diffusion coefficient (ADC) between compressed and non-compressed nerve roots from patients and the normal nerve roots from healthy volunteers were compared. Meanwhile, the nerve root fiber bundles were observed and analyzed.

The average FA and ADC values of the compressed nerve roots were 0.254 ± 0.307 and 1.892 ± 0.346 10^-3mm²/s, respectively. The average FA and ADC values of the non-compressed nerve roots were 0.377 ± 0.659 and 1.353 ± 0.344 10^-3mm²/s, respectively. The FA value of compressed nerve roots was significantly lower than that of non-compressed nerve roots (P < 0.01). The ADC value of compressed nerve roots was significantly higher than that of non-compressed nerve roots. There were no significant differences between the left and right nerve roots of normal volunteers in FA and ADC values (P > 0.05). The nerve roots at different levels of L3-S1 had significantly different FA and ADC values (P < 0.01). Incomplete fiber bundles with extrusion deformation, displacement or partial defect were observed in the compressed nerve root fiber bundles. The real diagnosis of the clinical situation of the nerve can provide neuroscientists with an important computer tool to help them infer and understand the possible working mechanism from the experimental data of behavior and electrophysiology.

The compressed lumbosacral nerve roots can be accurately localized through 3.0T magnetic resonance DTI, which is instructive for accurate clinical diagnosis and preoperative localization.

To investigate the performance of a novel radiological-metabolic scoring (RM-S) system to predict mortality and intensive care unit (ICU) requirements among COVID-19 patients and to compare performance with the chest computed-tomography severity-scoring (C-CT-SS). The RM-S was created from scoring systems such as visual coronary-artery-calcification scoring (V-CAC-S), hepatic-steatosis scoring (HS-S) and pancreatic-steatosis scoring (PS-S).

Between May 2021 and January 2022, 397 patients with COVID-19 were included in this retrospective cohort study. All demographic, clinical and laboratory data and chest CT images of patients were retrospectively reviewed. RM-S, V-CAC-S, HS-S, PS-S and C-CT-SS scores were calculated, and their performance in predicting mortality and ICU requirement were evaluated by univariate and multivariable analyses.

A total of 32 (8.1%) patients died, and 77 (19.4%) patients required ICU admission. Mortality and ICU admission were both associated with older age (p < 0.001). Sex distribution was similar in the deceased vs. survivor and ICU vs. non-ICU comparisons (p = 0.974 and p = 0.626, respectively). Multiple logistic regression revealed that mortality was independently associated with having a C-CT-SS score of ≥ 14 (p < 0.001) and severe RM-S category (p = 0.010), while ICU requirement was independently associated with having a C-CT-SS score of ≥ 14 (p < 0.001) and severe V-CAC-S category (p = 0.010).

RM-S, C-CT-SS, and V-CAC-S are useful tools that can be used to predict patients with poor prognoses for COVID-19. Long-term prospective follow-up of patients with high RM-S scores can be useful for predicting long COVID.

To compare main lacrimal gland values through shear-wave elastography (SWE) in patients with low Schirmer value and unspecified Sjögren's syndrome (SS) with healthy controls.

Admitted to the ophthalmology department with Schirmer value <10 mm, randomly selected 46 eyes of 46 patients evaluated for Sjögren's syndrome (SS) in the rheumatology department between December 2022 and April 2023 were classified as low Schirmer group (LSG). Randomly selected 48 eyes of 48 patients at a similar age with Schirmer value >10 mm were included as controls. Main lacrimal gland SWE measurements in LSG and control groups were recorded and compared as meter/second (m/sec).

Mean SWE values of the main lacrimal gland were measured as 2.78 ± 0.66 m/sec and 2.26 ± 0.29 m/sec in LSG and controls. SWE measurements were significantly higher in LSG patients than in controls (p<0.001). No significant correlation was found in the analysis between the Schirmer and the main lacrimal gland SWE values in LSG patients (p=0.702, r=0.058). No significant correlation was also detected between the Schirmer and main lacrimal gland SWE values in controls (p=0.097, r=0.242). No significant relationship was also found between age, gender, body mass index (BMI), and SWE values (p=0.351, p=0.493, p=0.328, respectively).

Mean SWE value of the main lacrimal gland was determined as significantly higher in patients with aqueous lacrimal insufficiency without SS than in controls. We consider that SWE measurements may be an imaging method to support the diagnosis of aqueous lacrimal insufficiency and used in follow-ups of those with dry eye syndrome (DES) in the future.

Distinguishing between IHCC and HCC is important because of their differences in treatment and prognosis. The hybrid Positron Emission Tomography/magnetic Resonance Imaging (PET/MRI) system has become more widely accessible, with oncological imaging becoming one of its most promising applications.

The objective of this study was to see how well 18F-fluorodeoxyglucose (18F-FDG) PET/MRI could be used for differential diagnosis and histologic grading of primary hepatic malignancies.

We retrospectively evaluated 64 patients (53 patients with HCC, 11 patients with IHCC) with histologically proven primary hepatic malignancies using 18F-FDG/MRI. The Apparent Diffusion Coefficient (ADC), Coefficient of Variance (CV) of the ADC, and standardized uptake value (SUV) were calculated.

The mean SUVmax value was higher for IHCC (7.7 ± 3.4) than for HCC (5.2 ± 3.1) (p = 0.019). The area under the curve (AUC) was 0.737, an optimal 6.98 cut-off value providing 72% sensitivity and 79% specificity. The ADCcv value in IHCC was statistically significantly higher than in HCC (p=0.014). ADC mean values in HCCs were significantly higher in low-grade tumors than in high-grade tumors. The AUC value was 0.73, and the optimal cut-off point was 1.20x10-6 mm²/s, giving 62% sensitivity and 72% specificity. The SUVmax value was also found to be statistically significantly higher in the high-grade group. The ADCcv value in the HCC low-grade group was found to be lower than in the high-grade group (p=0.036).

18F FDG PET/MRI is a novel imaging technique that can aid in the differentiation of primary hepatic neoplasms as well as tumor-grade estimation.

Acute pancreatitis (AP) is one of the most common digestive emergencies, and vascular complication is one of the primary reasons for death, with splanchnic venous thrombosis being the most common. Although extra-splanchnic venous thrombosis is rare, it carries the risk of life-threatening secondary pulmonary embolism.

We have, herein, reported a case of AP complicated by rare brachiocephalic vein thrombosis and superior vena cava thrombosis. A 40 years old woman was diagnosed with severe AP for abdominal pain 21 days ago. The patient received symptomatic treatment, including acid suppression, enzyme suppression, lipid-lowering, fluid infusion, anti-infection, and continuous renal replacement therapy. The patient was discharged after symptomatic relief. Recently, the patient was admitted again for middle-upper abdominal pain and discomfort. On admission, her blood platelet, D-Dimer, fibrin degradation products (FDP), and triglyceride levels have been found to be increased; abdominal enhanced CT showed pancreatic necrosis and an accumulation of peripancreatic necrosis and fluid, while chest enhanced CT suggested thrombosis in the right brachiocephalic vein and superior vena cava. The patient, however, improved and was discharged after anticoagulation combined with insulin and trypsin inhibitors.

In diagnosing and treating AP, dynamic monitoring of D-dimer levels is necessary for the timely detection of the development of thrombotic complications.

This study aimed to explore the imaging characteristics of patients with pulmonary tuberculosis complicated with pulmonary embolism and analyze the prognosis of the condition, thereby reducing the mortality and misdiagnosis rate of complications in this type of pulmonary tuberculosis.

In this retrospective study, a total of 70 patients diagnosed with pulmonary embolism by computed tomography pulmonary angiography (CTPA) from January 2016 to May 2021 in Anhui Chest Hospital were included. Among them, 35 patients with pulmonary embolism combined with pulmonary tuberculosis were set as the study group, and the other 35 patients with pulmonary embolism only were set as the control group. The imaging findings of chest CT examination, the incidence of pulmonary hypertension, the level of N-terminal proto-B-type brain natriuretic peptide (NT-proBNP), and the prognosis of patients were compared between the two groups. The incidence of deep venous embolism was evaluated by ultrasonography of the lower extremity.

In the study group, the median age of patients was 71 years, and the ratio of males to females was 2.5 to 1. In the control group, the median age was 66 years old, and the male-to-female ratio was 2.2 to 1. There were 16 cases (16/35, 45.71%) in the study group and 10 cases (10/35, 28.57%) in the control group with an increased level of NT-proBNP. Pulmonary hypertension occurred in 10 patients (10/35, 28.57%) in the study group and 7 patients (7/35, 20.00%) in the control group. Patients who lost follow-up included 5 in the study group (5/35, 14.29%) and 3 in the control group (3/35, 8.57%). There were 17 cases (17/35, 48.57%) in the study group and 3 cases (3/35, 8.57%) in the control group with pulmonary artery widening, and the difference was significant (P < 0.001). There were 13 deaths in the study group (13/35, 37.14%) and 1 death in the control group (1/35, 2.86%), and the difference was significant (P <0.001).

Special signs of pulmonary artery widening, pulmonary hypertension of varying degrees, and increased levels of NT-proBNP of varying degrees can be found in patients with pulmonary tuberculosis complicated with pulmonary embolism, and the three signs are positively correlated. The mortality of patients with pulmonary tuberculosis complicated with pulmonary embolism is significantly higher than that of patients with pulmonary embolism alone. Pulmonary tuberculosis and pulmonary embolism both occur in the ipsilateral lung, causing clinical symptoms to cover each other, thereby making diagnosis difficult.

Adenofibroma is a rare benign Müllerian mixed tumor composed of epithelial and mesenchymal cells. This tumor may occasionally be associated with toremifene therapy which is used as an adjuvant drug for breast cancer.

We describe a case of a 55-year-old woman with adenofibroma of the endometrium. This patient was receiving toremifene after surgery and neoadjuvant chemotherapy for breast cancer. She underwent a total abdominal hysterectomy and bilateral salpingectomy. There was no evidence of tumor residual or recurrence at 32 months of MRI follow-up.

In conclusion, we report a rare case of endometrial adenofibroma in a patient receiving toremifene. It must be borne in mind that long-term toremifene therapy may increase the frequency of endometrial neoplasms.

One of the most common issues following pedicle screw fixation is pedicle screw loosening. There are, however, few trustworthy methods for predicting screw loosening. The goal of the current study was to identify an efficient technique for using preoperative CT scanning to predict screw loosening in older patients and to offer recommendations for preoperative surgical planning.

The current analysis included retrospectively all patients who underwent lumbar pedicle screw fixation for degenerative lumbar diseases in our department between January 2015 and January 2022. Hounsfield units were used to assess each vertebra's attenuation in a CT scan (HU). Postoperative X-ray testing identified screw loosening. Using IBMSPSS 24.00 software, one-way analysis of variance (ANOVA) and receiver operating characteristic (ROC) curve analysis were performed.

Over a mean follow-up period of 28.4±11.5 (range 12-44 months) months, screw loosening was noted in 53 of 242 patients (136 male and 106 female, average age 58.7±7.3 years). Gender, BMI, smoking habits, and whether or not a patient had diabetes or spondylolisthesis were not shown to be significantly different among the patients (P>0.05). The difference between the average lumbar vertebral HU values in the screw-loosening group and the control group was significant (P<0.01) at 120.3±31.5HU and 138.6±37.6HU, respectively. The average HU value of L1-L4 exhibited an area under the curve (AUC) of 0.691 (95% CI: 0.614-0.784), according to ROC curve analysis. A HU cut-off value of 122 HU is a likely cut-off point to predict screw loosening with a sensitivity of 70% and a specificity of 58%.

The use of screw augmentation techniques can be decided using a prospective CT scan HU value-based prediction. An independent risk factor for screw loosening in an instrumented lumbar vertebra is a cutoff L1-L5 average HU value of 122 HU.

Although iron deposition has been identified as a significant migraine trigger, the key structures in episodic migraine (EM) remain unknown.

The aim of this study is to investigate cerebral iron deposition in EM using an advanced voxel-based quantitative susceptibility mapping (QSM).

A multi-echo gradient echo sequence MR was performed in 15 episodic migraine patients (EMs) and 27 normal control subjects (NCs). The reconstructed quantitative susceptibility mapping images and voxel-based analysis were performed over the entire brain. The susceptibility value of all brain regions with altered iron deposition was extracted, and the correlations between susceptibility value and clinical variables (including HAMA, HAMD, MoCA, VAS, MIDAS score, diseased duration, and headache frequency) were calculated.

EM patients presented increased susceptibility value in the left putamen and bilateral substantia nigra (SN) compared with NC. There was no correlation between susceptibility value and the clinical variables.

Increased brain iron deposition in the extrapyramidal system may be a biomarker for migraine, and abnormal iron metabolism may be involved in the extrapyramidal mechanism. The QSM technique would be an optimal and simple tool for clinical practice and research in iron measurement.

Primary thymic mucinous adenocarcinoma is an extremely rare and aggressive tumor with poor prognosis. The tumor may present as a heterogeneous solid or cystic mass accompanied by calcifications. However, clinical and radiologic features of the tumor are not well known due to the rarity of the disease, which makes accurate diagnosis difficult.

Here we present a rare case of primary thymic mucinous adenocarcinoma in the anterior mediastinum, including computed tomography (CT) and magnetic resonance imaging (MRI) findings. Chest computed tomography revealed a large anterior mediastinal mass with extensive calcifications with poor enhancement. MRI showed that anterior mediastinal mass showed intermediate signal intensity on T1-weighted images (T1WI), high SI on T2-weighted images (T2WI), and heterogeneous enhancement. Biopsy was performed and the anterior mediastinal tumor was diagnosed as thymic mucinous adenocarcinoma by histopathologic examination and immunohistochemical staining.

Thymic mucinous adenocarcinomas could be included in differential diagnoses of anterior mediastinal tumors showing extensive calcification, and common imaging findings of mucinous adenocarcinoma such as T2 high signal intensity and heterogeneous enhancement on MRI may be helpful to diagnose thymic mucinous adenocarcinoma.

The distinction between benign and borderline epithelial ovarian tumors is important because treatment and follow-up strategies differ.

We aimed to evaluate benign, borderline, and malignant epithelial ovarian tumors using MRI features and contributed to the preoperative evaluation.

MRIs of 81 patients (20 bilateral), including 31 benign, 27 borderline, and 23 malignant, who had pelvic imaging between 2013-2020, were evaluated retrospectively. The evaluation was made blindly to the pathology result by two radiologists with MRI scoring and features that we determined. MRI evaluation was performed with T1 TSE, T2 TSE, fat-suppressed T2 TSE, and before and after contrast T1 fat-suppressed and non-fat-suppressed TSE images. The numbers and findings obtained in scoring were evaluated by Chi-Square, ordinal logistic regression, and 2 and 3 category ROC analysis.

The total score varied between 7 and 24. Among the three groups, a significant difference was found in terms of T1, T2 signal intensity (p <0.01), size (p = 0.055), solid area (p <0.001), septa number (p <0.05), ovarian parenchyma (p = 0.001), ascites (p <0.001), peritoneal involvement (p <0.001), laterality (p <0.001), contrast enhancement pattern (p <0.001). On the other hand, no significant difference was found in terms of wall thickness, lymph node involvement and endometrial thickness (p> 0.05). Cut-off values were found as 11.5 and 18.5 in the 3-category ROC analysis performed for the score (VUS: 0.8109). Patients with a score below 11.5 were classified as benign, those between 11.5-18.5 as borderline, and those over 18.5 as malignant.

The differentiation of borderline tumors from benign and malignant tumors by MRI scoring will contribute to the preoperative diagnosis.

Transvaginal and transabdominal ultrasonography has become a widely used investigative method in the diagnostic workup of endometriosis, as well as for the postoperative follow-up. The variety of lesions, the distorted anatomy caused by adhesions and the fibrosis process represent the main challenges of the ultrasound evaluation. Regarding the recurrence of endometriosis, the diagnostic criteria are being imprecise, being adjusted according to the development of ultrasound techniques.

To this study, extensive research has been performed interrogating PubMed, Embase and Web of Science databases to identify published research including patients with endometriosis who underwent surgery. Included patients had postoperative ultrasound investigations to detect evidence of endometriosis recurrence. The selected timeframe was 5 years. We conducted a literature review on ultrasound markers of endometriosis recurrence.

In this analysis, 2023 patients from 9 studies were included. The recurrence rate detected was 17.26%. The postoperative treatment was different in the selected studies and they include the administration of progesterone, oral combined contraceptives, GnRh antagonists, aromatase inhibitors and intrauterine devices with prolonged release of progesterone.

The recurrence rate is different in the selected studies as a result of the recurrence definition used by each author and the minimum dimension of the lesions, in case of the cystic pattern. Innovative techniques of differential diagnosis by ultrasound are proposed, one of them being the textural analysis performed by computer-aided diagnosis – CAD. In order to standardize the recommendations regarding imaging techniques, we propose an algorithm for following up with patients in the postoperative period.

The transabdominal or endovaginal ultrasound performed regularly represents a key factor to determine the recurrence of endometriosis in the postoperative period and the imaging reassessment is recommended to be performed at a 6-month interval.