Beyond regulatory compliance: evaluating radiology artificial intelligence applications in deployment.

The implementation of artificial intelligence (AI) applications in routine practice, following regulatory approval, is currently limited by practical concerns around reliability, accountability, trust, safety, and governance, in addition to factors such as cost-effectiveness and institutional information technology support. When a technology is new and relatively untested in a field, professional confidence is lacking and there is a sense of the need to go above the baseline level of validation and compliance. In this article, we propose an approach that goes beyond standard regulatory compliance for AI apps that are approved for marketing, including independent benchmarking in the lab as well as clinical audit in practice, with the aims of increasing trust and preventing harm.

Fully automated deep-learning section-based muscle segmentation from CT images for sarcopenia assessment.

AIM: To develop a fully automated deep-learning-based approach to measure muscle area for assessing sarcopenia on standard-of-care computed tomography (CT) of the abdomen without any case exclusion criteria, for opportunistic screening for frailty. MATERIALS AND METHODS: This ethically approved retrospective study used publicly available and institutional unselected abdominal CT images (n=1,070 training, n=31 testing). The method consisted of two sequential steps: section detection from CT volume followed by muscle segmentation on single-section. Both stages used fully convolutional neural networks (FCNN), based on a UNet-like architecture. Input data consisted of CT volumes with a variety of fields of view, section thicknesses, occlusions, artefacts, and anatomical variations. Output consisted of segmented muscle area on a CT section at the L3 vertebral level. The muscle was segmented into erector spinae, psoas, and rectus abdominus muscle groups. Output was tested against expert manual segmentation. RESULTS: Threefold cross-validation was used to evaluate the model. Section detection cross-validation error was 1.41 ± 5.02 (in sections). Segmentation cross-validation Dice overlaps were 0.97 ± 0.02, 0.95 ± 0.04, and 0.94 ± 0.04 for erector spinae, psoas, and rectus abdominus, respectively, and 0.96 ± 0.02 for the combined muscle area, with R2 = 0.95/0.98 for muscle attenuation/area in 28/31 hold-out test cases. No statistical difference was found between the automated output and a second annotator. Fully automated processing took <1 second per CT examination. CONCLUSIONS: A FCNN pipeline accurately and efficiently automates muscle segmentation at the L3 vertebral level from unselected abdominal CT volumes, with no manual processing step. This approach is promising as a generalisable tool for opportunistic screening for frailty on standard-of-care CT.

O-RADS MRI score: analysis of misclassified cases in a prospective multicentric European cohort.

OBJECTIVE: To retrospectively review the causes of categorization errors using O-RADS-MRI score and to determine the presumptive causes of these misclassifications. METHODS: EURAD database was retrospectively queried to identify misclassified lesions. In this cohort, 1194 evaluable patients with 1502 pelvic masses (277 malignant / 1225 benign lesions) underwent standardized MRI to characterize adnexal masses with histology or 2 years' follow-up as a reference standard. An expert radiologist reviewed cases with two junior radiologists and lesions termed misclassified if malignant lesion was scored ≤ 3, a benign lesion was scored ≥ 4, the site of origin was incorrect, or a non-adnexal mass was incorrectly categorized as benign or malignant. RESULTS: There were 139 / 1502 (9.2%) misclassified masses in 116 women including 109 adnexal and 30 non-adnexal masses. False-negative cases corresponded to 16 borderline or invasive malignant adnexal masses rated score ≤ 3 (16 / 139, 11.5%). False-positive cases corresponded to 88 benign masses were rated score 4 (67 / 139, 48.2%) or 5 (18 / 139,12.9%) or considered suspicious non-adnexal lesions (3 / 139, 2.2%). Misclassifications were only due to origin error in 12 adnexal masses (8 benign, 4 malignant) (8.6%, 12 / 139) and 23 non-adnexal masses (18 benign, 5 malignant,16.5%, 23 / 139) perceived respectively as non-adnexal and adnexal masses. Interpretive error (n = 104), failure to recognize technical insufficient exams (n = 9), and perceptual errors (n = 4) were found. Most interpretive was due to misinterpretation of solid tissue or incorrect assignment of mass origin. Eighty-four out of 139 cases were correctly reclassified by the readers with strict adherence to the score rules. CONCLUSION: Most errors were due to misinterpretation of solid tissue or incorrect assignment of mass origin. KEY POINTS: • Prospective assignment of O-RADS-MRI score resulted in misclassification of 9.25% of sonographically indeterminate pelvic masses. • Most errors were interpretive (74.8%) due to misinterpretation of solid tissue as defined by the lexicon or incorrect assignment of mass origin. • Pelvic inflammatory disease is a common source of misclassification (8.9%) (12 / 139).

Machine learning in whole-body MRI: experiences and challenges from an applied study using multicentre data.

Machine learning is now being increasingly employed in radiology to assist with tasks such as automatic lesion detection, segmentation, and characterisation. We are currently involved in an National Institute of Health Research (NIHR)-funded project, which aims to develop machine learning methods to improve the diagnostic performance and reduce the radiology reading time of whole-body magnetic resonance imaging (MRI) scans, in patients being staged for cancer (MALIBO study). We describe here the main challenges we have encountered during the course of this project. Data quality and uniformity are the two most important data traits to be considered in clinical trials incorporating machine learning. Robust data pre-processing and machine learning pipelines have been employed in MALIBO, a task facilitated by the now freely available machine learning libraries and toolboxes. Another important consideration for achieving the desired clinical outcome in MALIBO, was to effectively host the resulting machine learning output, along with the clinical images, for reading in a clinical environment. Finally, a range of legal, ethical, and clinical acceptance issues should be considered when attempting to incorporate computer-assisting tools into clinical practice. The road from translating computational methods into potentially useful clinical tools involves an analytical, stepwise adaptation approach, as well as engagement of a multidisciplinary team.

Adnexal lesions: Imaging strategies for ultrasound and MR imaging.

Adnexal lesions are routinely encountered in general practice. Ultrasound is the first line of investigation in determining the benign or malignant potential of an adnexal lesion. In the cases of classic simple cysts, hemorrhagic cysts, endometriomas, dermoids and obviously malignant lesions, ultrasound may be sufficient for management recommendations. In cases where there is an isolated adnexal lesion, without peritoneal disease or serum CA-125 elevation, and in lesions considered indeterminate on ultrasound, MR imaging with incorporation of the ADNEx MR score can increase the specificity for the diagnosis of benignity or malignancy. This article will review the imaging evaluation of adnexal lesions and how to incorporate the ADNEx MR score to help guide clinical management.

Histogram analysis of apparent diffusion coefficient from whole-body diffusion-weighted MRI to predict early response to chemotherapy in patients with metastatic colorectal cancer: preliminary results.

AIM: To evaluate apparent diffusion coefficient (ADC) histogram analysis parameters, acquired from whole-body diffusion-weighted magnetic resonance imaging (DW-MRI), as very early predictors of response to chemotherapy in patients with metastatic colorectal cancer (mCRC). MATERIALS AND METHODS: This was a single-institution prospective study, approved by the West Midlands-South Birmingham research ethics committee. All patients gave fully informed consent prior to imaging. Sixteen patients with histologically confirmed mCRC were enrolled to the study and 11 were successfully scanned with whole-body DW-MRI before (baseline) and 10.8±2.7 days after commencing chemotherapy (follow-up). Therapy response was assessed by RECIST 1.1. Mean ADC and histogram parameters (skewness, kurtosis, 25th, 50th, and 75th percentiles) were compared between progressors and non-progressors at baseline and follow-up. Receiver operating characteristics (ROC) analysis was performed for the statistically significant parameters. Data from metastases were also compared to normative tissue data acquired from healthy volunteers. RESULTS: Three patients had progressive disease (progressors) and eight had partial response/stable disease (non-progressors). Mean, 25th, 50th, and 75th percentiles were significantly lower for progressors at baseline (p=0.012, 0.012, 0.012 and 0.025 respectively) with areas under the ROC curves (AUC)=0.58, 0.50, 0.58 and 0.63, respectively. Skewness and kurtosis were significantly lower for non-progressors at follow-up (p=0.001 and 0.003 respectively) with AUC=0.67 and 0.79 respectively. CONCLUSION: ADC histogram analysis shows potential in discriminating progressive from non-progressive disease in patients with mCRC, who underwent whole-body DW-MRI. The technique can potentially be tested as a response assessment methodology in larger trials.

European society of urogenital radiology (ESUR) guidelines: MR imaging of pelvic endometriosis.

Endometriosis is a common gynaecological condition of unknown aetiology that primarily affects women of reproductive age. The accepted first-line imaging modality is pelvic ultrasound. However, magnetic resonance imaging (MRI) is increasingly performed as an additional investigation in complex cases and for surgical planning. There is currently no international consensus regarding patient preparation, MRI protocols or reporting criteria. Our aim was to develop clinical guidelines for MRI evaluation of pelvic endometriosis based on literature evidence and consensus expert opinion. This work was performed by a group of radiologists from the European Society of Urogenital Radiology (ESUR), experts in gynaecological imaging and a gynaecologist expert in methodology. The group discussed indications for MRI, technical requirements, patient preparation, MRI protocols and criteria for the diagnosis of pelvic endometriosis on MRI. The expert panel proposed a final recommendation for each criterion using Oxford Centre for Evidence Based Medicine (OCEBM) 2011 levels of evidence. KEY POINTS: • This report provides guidelines for MRI in endometriosis. • Minimal and optimal MRI acquisition protocols are provided. • Recommendations are proposed for patient preparation, best MRI sequences and reporting criteria.

Use of an internal reference in semi-quantitative dynamic contrast-enhanced MRI (DCE MRI) of indeterminate adnexal masses.

OBJECTIVE: Semi-quantitative dynamic contrast-enhanced MRI (DCE MRI) has proven useful in discriminating benign from borderline/malignant adnexal lesions. Our aim was to assess if the use of a lesion-to-internal-reference ratio improved the performance in characterizing adnexal masses and which internal reference was suitable. METHODS: Semi-quantitative DCE MRI images of 71 indeterminate adnexal lesions were retrospectively reviewed. A region of interest was manually drawn onto the enhancing solid component, psoas muscle and normal outer myometrium. The DCE parameters were evaluated, and the lesion-to-internal-reference ratios were calculated. RESULTS: When the wash in rate of the lesion was higher than that of the myometrium, 97% specificity and 12% sensitivity for borderline/malignancy was reached. When the maximum relative enhancement and maximum absolute enhancement (SImax) of the lesion was less than those of the psoas, 100% specificity for benignity was achieved. The highest area under the curve (AUC) (0.807) was achieved using a SImax lesion-myometrium ratio. A slightly lower AUC (0.799) was achieved using a SImax lesion-psoas ratio, but the psoas muscle was more frequently measurable in the same slice as the lesion ROI. Although the AUC was higher, when using ratios instead of individual DCE values, this was not significantly different. CONCLUSION: DCE MRI has added diagnostic value in the assessment of adnexal lesions, and the use of internal references enables high specificity for malignancy and benignity. Lesion-internal-reference ratios have no added diagnostic value over DCE values alone. ADVANCES IN KNOWLEDGE: Both psoas muscle and myometrium are suitable internal references in the DCE assessment of adnexal lesions enabling high specificity for malignancy and benignity.

The prevalence of incidentally detected adrenal enlargement on CT.

AIM: To assess the prevalence and the department's detection rate of adrenocortical hyperplasia. MATERIALS AND METHODS: All computed tomography (CT) examinations of the adrenal glands between February and April 2011 were reviewed. The study excluded patients with known underlying cancer, abdominal trauma, or endocrine disease. The adrenal gland was deemed enlarged if its body was greater than 10 mm diameter, or a limb greater than 5 mm. RESULTS: There were a total of 564 eligible CT studies during this period. A total of 64 cases of incidental adrenal enlargement were found giving a prevalence of 11.3%. Only nine cases were reported in the contemporaneous CT report. CONCLUSION: The results of the present study show that incidental adrenal enlargement has a significant prevalence. It is often dismissed during reporting, and awareness needs to be raised in the radiological community. Equally, the clinical and biochemical significance needs to be assessed with endocrine correlation.

Diffusion-weighted and multiphase contrast-enhanced MRI as surrogate markers of response to neoadjuvant sunitinib in metastatic renal cell carcinoma.

BACKGROUND: Current imaging criteria for categorising disease response in metastatic renal cell carcinoma (mRCC) correlate poorly with overall survival (OS) in patients on anti-angiogenic therapies. We prospectively assess diffusion-weighted and multiphase contrast-enhanced (MCE) MR imaging (MRI) as markers of outcome. METHODS: Treatment-naive mRCC patients on a phase II trial using sunitinib completed an MRI substudy. Whole-tumour apparent diffusion coefficient (ADC) maps and histograms were generated, and mean ADC and AUC(low) (proportion of the tumour with ADC values lying below the 25th percentile of the ADC histogram) recorded. On MCE-MRI, regions of interest were drawn around the most avidly enhancing components to analyse enhancement parameters. Baseline (n=26) and treatment-related changes in surviving patients (n=20) were correlated with OS. Imaged metastases were also analysed. RESULTS: Forty-seven per cent of the patients showed significant changes in whole-tumour mean ADC following therapy, but there was no correlation with outcome. Patients with a high baseline AUC(low) and greater-than-median AUC(low) increase had reduced OS (HR=3.67 (95% confidence interval (CI)=1.23-10.9), P=0.012 and HR=3.72 (95% CI=0.98-14.21), P=0.038, respectively). There was no correlation between MCE-MRI parameters and OS. Twenty-eight metastases were analysed and showed positive correlation with primary tumour mean ADC for individual patients (r=0.607; P<0.001). CONCLUSION: Primary RCC ADC histogram analysis shows dynamic changes with sunitinib. Patients in whom the tumour ADC histogram demonstrated high baseline AUC(low) or a greater-than-median increase in AUC(low) with treatment had reduced OS.

In vitro and in vivo repeatability of abdominal diffusion-weighted MRI.

OBJECTIVE: To study the in vitro and in vivo (abdomen) variability of apparent diffusion coefficient (ADC) measurements at 1.5 T using a free-breathing multislice diffusion-weighted (DW) MRI sequence. METHODS: DW MRI images were obtained using a multislice spin-echo echo-planar imaging sequence with b-values=0, 100, 200, 500, 750 and 1000 s mm(-2). A flood-field phantom was imaged at regular intervals over 100 days, and 10 times on the same day on 2 occasions. 10 healthy volunteers were imaged on two separate occasions. Mono-exponential ADC maps were fitted excluding b=0. Paired analysis was carried out on the liver, spleen, kidney and gallbladder using multiple regions of interest (ROIs) and volumes of interest (VOIs). RESULTS: The in vitro coefficient of variation was 1.3% over 100 days, and 0.5% and 1.0% for both the daily experiments. In vivo, there was no statistical difference in the group mean ADC value between visits for any organ. Using ROIs, the coefficient of reproducibility was 20.0% for the kidney, 21.0% for the gallbladder, 24.7% for the liver and 28.0% for the spleen. For VOIs, values fall to 7.7%, 6.4%, 8.6% and 9.6%, respectively. CONCLUSION: Good in vitro repeatability of ADC measurements provided a sound basis for in vivo measurement. In vivo variability is higher and when considering single measurements in the abdomen as a whole, only changes in ADC value greater than 23.1% would be statistically significant using a two-dimensional ROI. This value is substantially lower (7.9%) if large three-dimensional VOIs are considered.

Diffusion-weighted imaging in the assessment of tumour grade in endometrial cancer.

OBJECTIVE: Endometrial cancer is the most common gynaecological malignancy in developed countries. Histological grade and subtype are important prognostic factors obtained by pipelle biopsy. However, pipelle biopsy "samples" tissue and a high-grade component that requires more aggressive treatment may be missed. The purpose of the study was to assess the use of diffusion-weighted MRI (DW-MRI) in the assessment of tumour grade in endometrial lesions. METHOD: 42 endometrial lesions including 23 endometrial cancers and 19 benign lesions were evaluated with DW-MRI (1.5T with multiple b-values between 0 and 750 s mm(-2)). Visual evaluation and the calculation of mean and minimum apparent diffusion coefficient (ADC) value were performed and correlated with histology. RESULTS: The mean and minimum ADC values for each histological grade were 1.02 ± 0.29×10(-3) mm(2) s(-1) and 0.74 ± 0.24×10(-3) mm(2) s(-1) (grade 1), 0.88 ± 0.39×10(-3) mm(2) s(-1) and 0.64 ± 0.36×10(-3) mm(2) s(-1) (grade 2), and 0.94 ± 0.32×10(-3) mm(2) s(-1) and 0.72 ± 0.36×10(-3) mm(2) s(-1) (grade 3), respectively. There was no statistically significant difference between tumour grades. However, the mean ADC value for endometrial carcinoma was 0.97 ± 0.31, which was significantly lower (p<0.0001) than that of benign endometrial pathology (1.50 ± 0.14). Applying a cut-off mean ADC value of less than 1.28 × 10(-3) mm(2) s(-1)we obtained a sensitivity, specificity, positive predictive value and negative predictive value for malignancy of 87%, 100%, 100% and 85.7%, respectively. CONCLUSION: Tumour mean and minimum ADC values are not useful in differentiating histological tumour grade in endometrial carcinoma. However, mean ADC measurement can provide useful information in differentiating benign from malignant endometrial lesions. This information could be clinically relevant in those patients where pre-operative endometrial sampling is not possible.

Magnetic resonance lymphography: a novel technique for lymph node assessment in gynecologic malignancies.

Lymph node status determines both prognosis and treatment choice in gynecological malignancies. Both CT and MRI are standard techniques used to detect lymph node involvement but these techniques have low sensitivity and specificity. Magnetic resonance lymphography (MRL) using iron oxide nanoparticles is a new technique for lymph node assessment which has shown much promise. This article discusses the applications of MRL in the setting of gynecological malignancy.

MRI appearances of borderline ovarian tumours.

This review was performed to describe the range of magnetic resonance imaging (MRI) appearances of borderline ovarian tumours. The MRI findings in 26 patients with 31 borderline ovarian tumours (mean age: 40.1 years, range: 14-85 years) were retrospectively reviewed. For each tumour, site, size, MRI characteristics, and enhancement following gadolinium administration were recorded. There were 20 serous and 11 mucinous borderline ovarian subtypes. Nine of 26 patients demonstrated bilateral disease on MRI; synchronous contralateral ovarian disease included three benign, five serous borderline, and one serous invasive tumour. A history of a metachronous mucinous borderline tumour was identified in one patient. MRI appearances were classified into four morphological categories: group 1 (6/31, 19%), unilocular cysts; group 2 (6/31, 19%), minimally septate cysts with papillary projections; group 3 (14/31, 45%), markedly septate lesions with plaque-like excrescences; and group 4 (5/31, 16%), predominantly solid with exophytic papillary projections, all of serous subtype. There was a significant difference in mean volume between serous (841.5 cm(3)) and mucinous (6358.2 cm(3)) subtypes (p=0.009). All tumours demonstrated at least one MRI feature suggestive of malignancy. The present review demonstrates the variable MRI appearances of borderline ovarian tumours along with imaging features suggestive of tumour subtype. In patients in whom the clinical features are suggestive of a borderline ovarian tumour (young age and normal or minimally elevated CA125), the ability to predict a borderline disease using morphological features observed on MRI would be extremely helpful in surgical planning, with the potential to offer fertility or ovary-preserving surgery. Future studies are required to further this aim.

Detection of neuroendocrine liver metastases with MnDPDP-enhanced MRI.

PURPOSE: The sensitivity of computerised tomography (CT) in detecting neuroendocrine liver metastases is variable and three-phase imaging is advocated. However, patients are often young and may require prolonged follow-up, thus a technique that avoids radiation exposure would be desirable. Our purpose was to assess the diagnostic performance of MRI, before and after administration of mangafodipir trisodium (MnDPDP), in the detection of neuroendocrine liver metastases. METHODS: Patients who had undergone single-phase or multi-phase contrast-enhanced MD-CT for neuroendocrine liver metastases were invited to have MRI. Two independent observers made quantitative measurements (number and size of lesions). All measurements were made on each available CT phase and all MRI sequences independently, and repeated after an interval to assess reproducibility. The final number of lesions was agreed on by consensus of three observers. A qualitative assessment (contrast and spatial resolution) and preferred modality were agreed on by consensus. RESULTS: 265 lesions were detected by consensus in 11 patients. Non-contrast CT was available in 4/11, arterial phase in 6/11 and portal phase in 10/11 patients. When compared with the consensus number of lesions, MD-CT identified 17% on non-contrast, 44% on arterial and 43% on portal venous imaging. Lesion detection on MRI was 48% on T(1)W, 52% on T(2)W and 92% on MnDPDP-MRI. The number of lesions detected on MnDPDP-MRI was closest to the final consensus reading (variance = 0.994, p = 0.0027). The reproducibility of lesion size measurements was best on MnDPDP-MRI (variance = 0.033, p = 0.0021). The preferred modality subjectively was MnDPDP-MRI in 9/11 cases and T(2)W MRI in 2/11. CONCLUSION: MRI is a robust technique in the demonstration of neuroendocrine liver metastases. It is highly reproducible in both detecting the number and measuring the size of lesions. We recommend T(2)W MRI and MnDPDP-MRI in detection and follow-up of neuroendocrine liver metastases.

The appearance of the adrenal glands on computed tomography in multiple endocrine neoplasia type 1.

AIMS: To review the morphology of the adrenal glands in multiple endocrine neoplasia type 1 (MEN1) on computed tomography (CT) to compare the results with established normal values for adrenal size and nodularity and to correlate adrenal size with serum cortisol secretory dynamics. MATERIALS AND METHODS: Two observers independently reviewed the adrenal CT in 28 patients with MEN1, measuring the maximum width of the body of the gland and the medial and lateral limbs. Incidence and location of nodules >5 mm within the gland were recorded. Following exclusion of known cases of Cushing's syndrome, adrenal gland size was compared with previously documented normative data. Adrenal gland size was compared between patients with normal and abnormal cortisol dynamics. RESULTS: Comparison of mean adrenal size in MEN1 patients with normative data showed that the adrenal limbs were significantly larger in MEN1 than normal (P<0.0001 in all four limbs). Adrenal body was also significantly larger (P<0.05). Nodules were demonstrated in 17 (60%) of patients (versus 0.4-2% in the normal population). No statistically significant correlation was demonstrated between adrenal limb hyperplasia and abnormal cortisol dynamics. CONCLUSIONS: In patients with MEN1, adrenal limb hyperplasia and adrenal nodules are significantly more common than in the normal population, a phenomenon not previously documented in a quantitative manner. There was no significant correlation between adrenal limb hyperplasia and abnormal cortisol dynamics.

Computed tomography features of recurrent ovarian carcinoma according to time to relapse.

BACKGROUND: Although the sites of recurrent ovarian cancer are individually described in the literature, patterns of recurrent disease are poorly understood. PURPOSE: To describe CT patterns of disease in recurrent ovarian cancer. To emphasize common patterns, recognise subtle and unusual sites of recurrent disease. MATERIALS AND METHODS: We identified patients between 1981-2004 presenting with clinical recurrence or elevated CA 125 after complete primary clinical and radiological response. CT imaging at primary diagnosis, during and after treatment and at recurrence was retrospectively reviewed. Site, distribution, stage of disease and time to relapse was recorded. RESULTS: 400 patients were treated for ovarian cancer. 214(54%) achieved complete primary response. 161(75%) relapsed with complete imaging available in 67 patients. Of the 67 patients, 14 (21%) recurred within 1 year, 44 (66%) relapsed between 1-5 years. Therefore 87% of relapses occurred within 5 years following primary treatment. Five (8%) relapsed between 5-10 years and 4 (6%) relapsed after 10 years. Commonest pattern of relapse was pelvic mass in 35 (48%) patients, solitary in 15 (22%). 27 (45%) relapsed with peritoneal thickening, 27 (45%) had small or large bowel serosal disease, 22 (33%) had enlarged lymphadenopathy, 6 as sole manifestation of recurrence, 20 (30%) presented with unusual sites of recurrence: 6 splenic, 10 hepatic, 2 biliary, 3 brain and 2 muscle. CONCLUSION: Our study is the first to describe common patterns of recurrence in ovarian cancer. Most frequent site is pelvis, followed by peritoneum, serosal surfaces and nodal disease. 30% presented with disease at 'unusual' sites.

An atlas of the pelvic lymph node regions to aid radiotherapy target volume definition.

AIMS: The implementation of advanced three-dimensional radiotherapy planning techniques requires accurate target volume localisation. We have previously developed guidelines to aid definition of the pelvic lymph node regions, and the aim of this study was to produce a CT atlas. MATERIALS AND METHODS: The guidelines were applied to a CT scan of a patient to receive adjuvant radiotherapy. RESULTS: Reference CT images of the pelvis were generated, illustrating the nodal regions and a typical target volume for adjuvant pelvic radiotherapy for gynaecological cancer. CONCLUSION: These images can be used as an aid for target volume definition of the pelvic nodal regions.

Evaluation of endometrial carcinoma on magnetic resonance imaging.

Our aims were to assess diagnostic performance of T2-weighted (T2W) and dynamic gadolinium-enhanced T1-weighted (T1W) magnetic resonance imaging (MRI) in the preoperative assessment of myometrial and cervical invasion by endometrial carcinoma and to identify imaging features that predict nodal metastases. Two radiologists retrospectively reviewed MR images of 96 patients with endometrial carcinoma. Tumor size, depth of myometrial and cervical invasion, and nodal enlargement were recorded and then correlated with histology. The sensitivity, specificity, positive and negative predictive values (PPV and NPV) for the identification of any myometrial invasion (superficial or deep) were 0.94, 0.50, 0.93, 0.55 on T2W and 0.92, 0.50, 0.92, 0.50 on dynamic T1W, and for deep myometrial invasion were 0.84, 0.78, 0.65, 0.91 on T2W and 0.72, 0.88, 0.72, 0.88 on dynamic T1W. The sensitivity, specificity, PPV and NPV for any cervical invasion (endocervical or stromal) were 0.65, 0.87, 0.57, 0.90 on T2W and 0.50, 0.90, 0.46, 0.92 on dynamic T1W, and for cervical stromal involvement were 0.69, 0.95, 0.69, 0.95 on T2W and 0.50, 0.96, 0.57, 0.95 on dynamic T1W. Leiomyoma or adenomyosis were seen in 73% of misdiagnosed cases. Sensitivity and specificity for the detection of nodal metastases was 66% and 73%, respectively. Fifty percent of patients with cervical invasion on MRI had nodal metastases. In conclusion, MRI has a high sensitivity for detecting myometrial invasion and a high NPV for deep invasion. MRI has a high specificity and NPV for detecting cervical invasion. Dynamic enhancement did not improve diagnostic performance. MRI may allow accurate categorization of cases into low- or high-risk groups ensuring suitable extent of surgery and adjuvant therapy.