A Healthcare Physician Can Be Trained to Perform Intestinal Ultrasound in Children With Inflammatory Bowel Disease.

OBJECTIVES: Training healthcare physicians to perform intestinal ultrasound (IUS) during outpatient visits with equal accuracy as radiologists could improve clinical management of IBD patients. We aimed to assess whether a healthcare-physician can be trained to perform IUS, with equal accuracy compared with experienced radiologists in children with iBD, and to assess inter-observer agreement. METHODS: Consecutive children, 6 to 18 years with IBD or suspicion of IBD, who underwent ileo-colonoscopy were enrolled. iUS was performed independently by a trained healthcare-physician and a radiologist in 1 visit. Training existed of an international training curriculum for IUS. Operators were blinded for each other's IUS, and for the ileocolonoscopy. Difference in accuracy of IUS by the healthcare-physician and radiologist was assessed using areas under the ROC curve (AUROC). Inter-observer variability was assessed in terminal ileum (TI), transverse colon (TC) and descending-colon (DC), for disease activity (ie, bowel wall thickness [BWT] >2 mm with hyperaemia or fat-proliferation, or BWT >3 mm). RESULTS: We included 73 patients (median age 15, interquartile range [IQR]:13-17, 37 [51%] female, 43 [58%] with Crohn disease). AUROC ranged between 0.71 and 0.81 for the healthcare-physician and between 0.67 and 0.79 for radiologist (P  > 0.05). Inter-observer agreement for disease activity per segment was moderate (K: 0.58 [SE: 0.09], 0.49 [SE: 0.12], 0.52 [SE: 0.11] respectively for TI, TC, and DC). CONCLUSIONS: A healthcare- physician can be trained to perform IUS in children with IBD with comparable diagnostic accuracy as experienced radiologists. The interobserver agreement is moderate. Our findings support the usage of IUS in clinical management of children with IBD.

Accuracy of controlled attenuation parameter compared with ultrasound for detecting hepatic steatosis in children with severe obesity.

OBJECTIVES: To determine the diagnostic accuracy of controlled attenuation parameter (CAP) on FibroScan® in detecting and grading steatosis in a screening setting and perform a head-to-head comparison with conventional B-mode ultrasound. METHODS: Sixty children with severe obesity (median BMI z-score 3.37; median age 13.7 years) were evaluated. All underwent CAP and US using a standardized scoring system. Magnetic resonance spectroscopy proton density fat fraction (MRS-PDFF) was used as a reference standard. RESULTS: Steatosis was present in 36/60 (60%) children. The areas under the ROC (AUROC) of CAP for the detection of grade ≥ S1, ≥ S2, and ≥ S3 steatosis were 0.80 (95% CI: 0.67-0.89), 0.77 (95% CI: 0.65-0.87), and 0.79 (95% CI: 0.66-0.88), respectively. The AUROC of US for the detection of grade ≥ S1 steatosis was 0.68 (95% CI: 0.55-0.80) and not significantly different from that of CAP (p = 0.09). For detecting ≥ S1 steatosis, using the optimal cutoffs, CAP (277 dB/m) and US (US steatosis score ≥ 2) had a sensitivity of 75% and 61% and a specificity of 75% and 71%, respectively. When using echogenicity of liver parenchyma as only the scoring item, US had a sensitivity of 70% and specificity of 46% to detect ≥ S1 steatosis. The difference in specificity of CAP and US when using only echogenicity of liver parenchyma of 29% was significant (p = 0.04). CONCLUSION: The overall performance of CAP is not significantly better than that of US in detecting steatosis in children with obesity, provided that the standardized scoring of US features is applied. When US is based on liver echogenicity only, CAP outperforms US in screening for any steatosis (≥ S1). KEY POINTS: • The areas under the ROC curves of CAP and ultrasound (US) for detecting grade ≥ S1 steatosis were 0.80 and 0.68, respectively, and were not significantly different (p = 0.09). • For detecting grade ≥ S1 steatosis in severely obese children, CAP had a sensitivity of 75% and a specificity of 75% at its optimal cutoff value of 277 dB/m. • For detecting grade ≥ S1 steatosis in clinical practice, both CAP and US can be used, provided that the standardized scoring of US images is used.

Individualized lung recruitment during high-frequency ventilation in preterm infants is not associated with lung hyperinflation and air leaks.

UNLABELLED: Lung recruitment during high-frequency ventilation (HFV) in preterm infants with respiratory distress syndrome (RDS) has been associated with an increased risk of lung hyperinflation and air leaks. Individualizing the lung recruitment procedure to the severity of lung disease of each patient might reduce these risks. In this prospective cohort study, we evaluated chest X-ray (CXR) characteristics during individualized oxygenation-guided lung recruitment with HFV in preterm infants with RDS, before and after surfactant therapy. Two pediatric radiologists scored radiolucency, the presence of lung hyperinflation, and/or air leaks following lung recruitment during HFV in 69 infants before and 39 infants after surfactant treatment. Following lung recruitment, the median radiolucency score was 2, with 44 (64 %) infants having a score ≤2. Only mild to moderate hyperinflation was seen in 13 (19 %) infants, with no air leaks. After the surfactant, the radiolucency score improved in 62 % of 39 paired CXRs (p < 0.001). Mild to moderate hyperinflation was seen in nine (24 %) patients. During the entire admission, only four (6 %) of the patients developed air leaks. CONCLUSION: The risk of significant hyperinflation and air leaks is low when using an individualized oxygenation-guided recruitment procedure during HFV in preterm infants with RDS. WHAT IS KNOWN: • Lung recruitment during high-frequency ventilation in preterm infants with respiratory distress syndrome is associated with an increased risk of lung hyperinflation and air leaks. What is New: • The risk of lung hyperinflation and air leaks is low when using an individualized oxygenation-guided lung recruitment procedure during high-frequency ventilation in preterm infants with respiratory distress syndrome.

Reliability and responsiveness of the Juvenile Arthritis MRI Scoring (JAMRIS) system for the knee.

OBJECTIVES: To assess the reliability and responsiveness of a new Juvenile Arthritis MRI Scoring (JAMRIS) system for evaluating disease activity of the knee. METHODS: Twenty-five juvenile idiopathic arthritis (JIA) patients with clinical knee involvement were studied using open-bore 1-T MRI. MRI features of synovial hypertrophy, bone marrow changes, cartilage lesions and bone erosions were independently scored by five readers using the JAMRIS system. In addition, the JAMRIS system was determined to be a follow-up parameter by two readers to evaluate the response to therapy in 15 consecutive JIA patients. RESULTS: Inter-reader (ICCs 0.86-0.95) and intra-reader reliability (ICCs 0.92-1.00) for the scoring of JAMRIS features was good. Reliability of the actual scores and changes in scores over time was good for all items: ICCs 0.89-1.00, 0.87-1.00, respectively. Concerning therapy response, the mean synovial hypertrophy scores decreased significantly (mean 1.1 point; P < 0.001, SRM = -0.65). No change was observed with respect to bone marrow change, cartilage lesion and bone erosion scores. CONCLUSIONS: The JAMRIS proved to be a simple and highly reliable assessment score in the evaluation of JIA disease activity of the knee. The JAMRIS system may serve as an objective and accurate outcome measure in future research and clinical trials.

MRI of the breast in patients with DCIS to exclude the presence of invasive disease.

OBJECTIVES: Core biopsy underestimates invasion in more than 20% of patients with preoperatively diagnosed ductal carcinoma in situ (DCIS) without evidence of invasion (pure DCIS). The aim of the current study was to evaluate the efficacy of preoperative magnetic resonance imaging (MRI) to discriminate between patients with DCIS who are at high risk of invasive breast cancer and patients at low risk. METHODS: One hundred and twenty-five patients, preoperatively diagnosed with pure DCIS (128 lesions; 3 bilateral) by core-needle biopsy, were prospectively included. Clinical, mammographic, histological (core biopsy) and MRI features were assessed. All patients underwent breast surgery. Analyses were performed to identify features associated with presence of invasion. RESULTS: Eighteen lesions (14.1%) showed invasion on final histology. Seventy-three lesions (57%) showed suspicious enhancement on MRI with a type 1 (n = 12, 16.4%), type 2 (n = 19, 26.0%) or type 3 curve, respectively (n = 42, 57.5%). At multivariate analysis, the most predictive features for excluding presence of invasive disease were absence of enhancement or a type 1 curve on MRI (negative predictive value 98.5%; A(Z) 0.80, P = 0.00006). CONCLUSION: Contrast medium uptake kinetics at MRI provide high negative predictive value to exclude presence of invasion and may be useful in primary surgical planning in patients with a preoperative diagnosis of pure DCIS. KEY POINTS: It is important to determine invasion in breast DCIS. • MRI contrast medium uptake kinetics can help exclude the presence of invasion. • However, the positive predictive value for the presence of invasion is limited. • MRI features were more accurate at predicting invasion than mammographic features alone.

Double inversion recovery MRI versus contrast-enhanced MRI for evaluation of knee synovitis in juvenile idiopathic arthritis.

BACKGROUND: Double inversion recovery (DIR) MRI has the potential to accentuate the synovium without using contrast agents, as it allows simultaneous signal suppression of fluid and fat. The purpose of this study was (1) to compare DIR MRI to conventional contrast-enhanced (CE) MRI for delineation of the synovium in the knee in children with juvenile idiopathic arthritis (JIA) and (2) to assess the agreement between DIR MRI and CE-MRI regarding maximal synovial thickness measurements. RESULTS: In this prospective study, 26 children with JIA who consecutively underwent 3.0-T knee MRI between January 2018 and January 2021 were included (presence of knee arthritis: 13 [50%]; median age: 14 years [interquartile range [IQR]: 11-17]; 14 girls). Median confidence to depict the synovium (0-100 mm visual analogue scale; scored by 2 readers [consensus based]) was 88 (IQR: 79-97) for DIR MRI versus 100 (IQR: 100-100) for CE-MRI (p value = < .001). Maximal synovial thickness per child (millimeters; scored by 4 individual readers) on DIR MRI was greater (p value = < .001) in the children with knee arthritis (2.4 mm [IQR: 2.1-3.1]) than in those without knee arthritis (1.4 mm [IQR: 1.0-1.6]). Good inter-technique agreement for maximal synovial thickness per child was observed (rs = 0.93 [p value = < .001]; inter-reader reliability: ICC DIR MRI = 0.87 [p value = < .001], ICC CE-MRI = 0.90 [p value = < .001]). CONCLUSION: DIR MRI adequately delineated the synovium in the knee of children with JIA and enabled synovial thickness measurement similar to that of CE-MRI. Our results demonstrate that DIR MRI should be considered as a child-friendly alternative to CE-MRI for evaluation of synovitis in children with (suspected) JIA.

A practical approach to manage additional lesions at preoperative breast MRI in patients eligible for breast conserving therapy: results.

The aim of this prospective study was to evaluate the efficacy of directives, established to handle additional lesions at preoperative contrast-enhanced magnetic resonance imaging (MRI). Six-hundred-and-ninety consecutive patients with pathology-proven breast cancer planned for BCT based on clinical examination and conventional imaging underwent preoperative breast MRI. The incidence of additional lesions detected at MRI and impact on management were evaluated. Additional findings were pathology-proven or considered benign by follow-up. Findings for which no pathology proof was available prior to surgery, were defined as Unidentified Breast Objects (UBOs). Patients with multicentric or contralateral UBOs underwent BCT as planned with annual follow-up. Multifocal UBOs in the vicinity of the index cancer were excised with wider local margins. Preoperative MRI detected 141 additional lesions in 121 patients (17.5%). Of these lesions, 44.0% were proven malignant. Additional findings classified as UBOs were found in 81 patients (11.7%). None of the UBOs outside the primary tumour region resulted in malignant disease at follow-up after BCT (mean follow-up time: 57.1 months). However, most multifocal UBOs (in the vicinity of the primary) were malignant (77.5%). The strategy to pursue BCT with larger wide-local excisions for multifocal UBOs and to follow-up multicentric and contralateral UBOs with conventional imaging is effective to exclude malignancy at follow-up. After second-look targeted ultrasound has been performed, MRI-guided biopsy of BIRADS-3 multicentric and contralateral additional findings may have limited complementary clinical value.

The impact of reconstruction techniques on observer performance for the detection and characterization of small pulmonary nodules in chest CT of children under 13 years.

OBJECTIVE: To compare three different reconstruction techniques of CT data for the detection of pulmonary nodules in children under 13 years. Secondly to assess the prevalence of perifissural nodular opacities. MATERIALS AND METHODS: The study consisted of chest CTs of 31 children (median age 6.9 years, range 2.1-12.7), of whom 17 had known extra-thoracic malignancies. Four observers assessed three techniques for the presence of nodules: axial 5 mm maximum intensity projections (MIPs) used in conjunction with 1 mm slices (mode A), 1 mm slices alone (mode B) and 3 mm slices (mode C). All modes were available in 3D. Per mode sensitivities were determined above a certain threshold of reader agreement. Confidence level and reader agreement for identification of an opacity as nodule served as surrogate for quality of nodule characterization. RESULTS: 103 nodules (median size 2.0 mm) were detected. Mode A yielded the highest interreader agreement (κ 0.336) and a superior sensitivity (71%, p = 0.003) compared to mode B and C (κ 0.218, sensitivity 57% and κ 0.247, sensitivity 56%, respectively). Mode B provided the highest confidence level and interreader agreement with respect to nodule identification (mean 4.3/5, κw 0.508). Double reading improved and evened interreader agreement for all modes (κ 0.450), mode A maintained the highest sensitivity (89.1%, p = 0.05-0.08). A median of 1 intrapulmonary lymph node/patient was seen in children with and without malignancy. CONCLUSION: MIP improves the detection of pulmonary nodules in chest CTs of children, but overall interreader agreement is only fair. Double reading represents a powerful tool to increase diagnostic reliability in chest CTs of children with a malignancy. Nodule characterization is best with 1 mm slices. Intrapulmonary lymph nodes occur in children with and without malignancy.

Noncirrhotic Portal Hypertension in Perinatally HIV-infected Adolescents Treated With Didanosine-containing Antiretroviral Regimens in Childhood.

BACKGROUND: Noncirrhotic portal hypertension (NCPH) has been reported in HIV-infected adults. Antiretroviral drugs, as well as genetic and thrombophilic predisposition, have been suggested as possible etiologic factors. METHODS: Clinical data were collected from 6 HIV-infected patients attending the Infectious Diseases Departments at respectively Emma Children's Hospital Academic Medical Centre in Amsterdam, The Thai Red Cross AIDS Research Centre, Bangkok, Imperial College Healthcare NHS Trust, London who were diagnosed with NCPH. All underwent extensive blood analysis, liver ultrasound, liver elastography, esophagogastroduodenoscopy and percutaneous needle liver biopsy for histological evaluation. RESULTS: We describe 6 perinatally HIV-infected adolescents, all female, who developed NCPH after prolonged exposure during childhood to a didanosine-containing antiretroviral regimen. Histology and electron microscopy showed periportal fibrosis and mitochondrial damage as key findings in their liver biopsies. One of these 6 patients required surgical intervention, the remainder have been managed conservatively to date. CONCLUSIONS: Thus, symptomatic NCPH may present in adolescence after perinatally acquired HIV-1 infection. In this case series, risk factors included female sex and prolonged exposure to antiretroviral regimens that included the nucleoside-analogue didanosine in childhood.

Additional breast lesions in patients eligible for breast-conserving therapy by MRI: impact on preoperative management and potential benefit of computerised analysis.

This study was conducted to assess the incidence and impact of additional findings from magnetic resonance imaging (MRI) on the workup of patients eligible for breast-conserving therapy (BCT) and to optimise the specificity of further workup by combining radiological reading with computerised analysis. One hundred and sixteen patients eligible for BCT underwent preoperative MRI where the gold standard was histology or follow-up (median 35 months, range 23-48). The incidence of additional findings and impact on treatment (wider excision/conversion to mastectomy) were assessed. The specificity of referral to further workup was also assessed without and with computerised analysis. Additional findings from MRI occurred in 41% of patients, requiring workup in 78%. In 22% the findings were malignant, causing change in treatment. Specificity was 33% (10/30) for radiological reading alone, and 97% (29/30) combined with computer analysis. Our findings show that additional findings preoperative MRI required workup in approximately one-third of patients and we suggest that combining radiological reading with computer analysis has the potential to accurately exclude benign lesions from further workup.

Reproducibility of the assessment of tumor extent in the breast using multiple image modalities.

The accuracy of breast-conserving therapy (BCT) is limited by uncertainties in the assessment of tumor extent. These uncertainties may result in too wide treatment volumes leading to undesirable cosmetic results, or too narrow treatment volumes leading to higher probabilities of local recurrence. The aim of this study is to quantify the reproducibility of the assessment of tumor extent in the breast at preoperative diagnostic imaging with multiple imaging modalities and at pathology, applied to (1) determining minimum surgical safety margins to reduce the probability of underestimating the tumor extent due to uncertainty in the radiological assessment, and (2) defining the minimum difference between two measurements of tumor size that indicates a significant reduction of tumor extent in response to neoadjuvant chemotherapy or hormonal therapy. Measurements of the largest tumor diameter in mammography, ultrasonography, contrast-enhanced magnetic resonance imaging, and at pathology were retrieved, retrospectively, for 105 patients eligible for BCT. An analysis of variance technique is employed to separate uncertainty at preoperative diagnostic imaging from uncertainty at pathology. The random variations are thus calculated independently of the systematic deviations, avoiding the necessity to choose a gold standard. Moreover, the technique does not require repeat measurements of tumor extent, thus allowing the use of data that is obtained in daily clinical practice, while avoiding bias due to recollection. The magnitude of the random variations is used to determine minimum surgical safety margins and to define the minimum significant difference between two measurements of tumor size. The overall random variations in the assessment of tumor extent are on the order of 3 mm (1 s.d.) with only little differences of about 0.3 mm between the four techniques. The dependence of the random variations on tumor size was found significant (p < 0.05) for mammography (2.7 mm vs 4.2 mm, 1 s.d.) and ultrasonography (2.5 mm vs 3.8 mm, 1 s.d.) for tumors up to 17 mm compared to those that are larger. A minimum surgical safety margin on the order of 5 mm for tumors up to 17 mm and 7 mm for larger tumors takes the uncertainty in radiological assessment of the tumor extent into account effectively in 95% of the performed surgical procedures. A minimum difference in largest tumor diameter of 7 mm for tumors up to 17 mm and 9 mm for those that are larger indicates a significant (p < 0.05) reduction of tumor extent in response to neoadjuvant chemotherapy or hormonal therapy. The reproducibility of the assessment of tumor extent at preoperative diagnostic imaging is of comparable magnitude to the reproducibility at pathology. The uncertainty in the preoperative assessment of tumor extent constitutes a large portion (5-7 mm) of the current safety margin in breast-conserving surgery (10 mm). In monitoring response to neoadjuvant chemotherapy or hormonal therapy using repeat imaging before and after treatment, the current clinical guidelines may produce approximately 10% false-positive responses for tumors between 20 and 30 mm.

Reproducibility of mammary gland structure during repeat setups in a supine position.

PURPOSE: In breast conserving therapy, complete excision of the tumor with an acceptable cosmetic outcome depends on accurate localization in terms of both the position of the lesion and its extent. We hypothesize that preoperative contrast-enhanced magnetic resonance (MR) imaging of the patient in a supine position may be used for accurate tumor localization and marking of its extent immediately prior to surgery. Our aims in this study are to assess the reproducibility of mammary gland structure during repeat setups in a supine position, to evaluate the effect of a breast immobilization device, and to derive reproducibility margins that take internal tissue shifts into account occurring between repeat setups. MATERIALS & METHODS: The reproducibility of mammary gland structure during repeat setups in a supine position is estimated by quantification of tissue shifts in the breasts of healthy volunteers between repeat MR setups. For each volunteer fiducials are identified and registered with their counter locations in corresponding MR volumes. The difference in position denotes the shift of breast tissue. The dependence on breast volume and the part of the breast, as well as the effect of a breast immobilization cast are studied. RESULTS: The tissue shifts are small with a mean standard deviation on the order of 1.5 mm, being slightly larger in large breasts (V> 1000 cm3), and in the posterior part (toward the pectoral muscle) of both small and large breasts. The application of a breast immobilization cast reduces the tissue shifts in large breasts. A reproducibility margin on the order of 5 mm will take the internal tissue shifts into account that occur between repeat setups. CONCLUSION: The results demonstrate a high reproducibility of mammary gland structure during repeat setups in a supine position.

Ultrasonography and fine-needle aspiration cytology can spare breast cancer patients unnecessary sentinel lymph node biopsy.

BACKGROUND: Some 30% to 40% of the breast cancer patients scheduled for sentinel node biopsy have axillary metastasis. Pilot studies suggest that ultrasonography is useful in the preoperative detection of such nodes. The aims of this study were to evaluate the sensitivity of preoperative ultrasonography and fine-needle aspiration cytology for detecting axillary metastases and to assess how often sentinel node biopsy could be avoided. METHODS: Between October 1999 and December 2003, 726 patients with clinically negative lymph nodes were eligible for sentinel node biopsy. A total of 732 axillae were examined. Preoperative ultrasonography with subsequent fine-needle aspiration cytology in case of suspicious lymph nodes was performed in all patients. The sentinel node procedure was omitted in patients with tumor-positive axillary lymph nodes in lieu of axillary lymph node dissection. RESULTS: Ultrasound and fine-needle aspiration cytology established axillary metastases in 58 (8%) of the 726 patients. These 58 were 21% of the total of 271 patients who were proven to have axillary metastasis in the end. Of the patients with ultrasonographically suspicious lymph nodes and negative cytology, 31% had tumor-positive sentinel nodes. Patients with preoperatively established metastases by ultrasonography and fine-needle aspiration cytology had more tumor-positive lymph nodes (P < .001) than patients with metastases established later on. CONCLUSIONS: The sensitivity of ultrasonography and fine-needle aspiration cytology is 21%, and unnecessary sentinel node biopsy is avoided in 8% of the patients. This approach improves the selection of patients eligible for sentinel node biopsy.

Contrast-enhanced MRI in breast cancer patients eligible for breast-conserving therapy: complementary value for subgroups of patients.

The aim of this study was to identify patients prior to breast-conserving therapy (BCT) who have complementary value of contrast-enhanced magnetic resonance imaging (MRI) over conventional imaging in the assessment of tumor extent. All patients were eligible for BCT according to conventional imaging, and underwent preoperative MRI as part of this study. One hundred and sixty-five patients (166 tumors) were included. MRI was defined to have complementary value if conventional imaging underestimated or overestimated tumor extent (by more than 10 mm compared to histology) and MRI assessed the extent accurately. Logistic regression was employed to identify characteristics that are predictive of the complementary value of preoperative MRI. MRI had complementary value in 39 cases (23%). Patients <58 years old with irregular lesion margins at mammography and discrepancy in tumor extent by more than 10 mm between mammography and ultrasonography had a 3.2x higher chance of accurate assessment at MRI (positive predictive value 50%, negative predictive value 84%, p=0.0002). Preoperative MRI in patients eligible for BCT is more accurate than conventional imaging in the assessment of tumor extent in approximately one out of four patients. Subgroups of patients in whom MRI has complementary value may be defined by the differences in clinical and imaging features.

Ultrasonography-guided fine-needle aspiration cytology before sentinel node biopsy in patients with penile carcinoma.

OBJECTIVE: To assess the accuracy of ultrasonography (US)-guided fine-needle aspiration cytology (FNAC) for detecting occult lymph node metastases in patients with squamous cell carcinoma of the penis. PATIENTS AND METHODS: Forty-three patients with 83 clinically node-negative inguinal regions were assessed with US and FNAC. The results were compared with histology from subsequent dynamic sentinel-node biopsy (DSNB) or inguinal lymph node dissection. RESULTS: Thirty-four groins in 27 patients were considered to be suspicious by US and the lymph nodes were aspirated. Nine nodes contained tumour cells and this was confirmed by subsequent lymph node dissection. The sensitivity and specificity of US-guided FNAC were 39% (nine of 23) and 100% (60 of 60), respectively. The number of groins requiring DSNB was reduced by 11% (nine of 83). CONCLUSION: US-guided FNAC can be used as the initial investigation in clinically node-negative groins. If tumour is confirmed then therapeutic inguinal lymph node dissection can be earlier and fewer DSNBs are required.

Clinically and mammographically occult breast lesions on MR images: potential effect of computerized assessment on clinical reading.

PURPOSE: To investigate if and how computerized analysis complements characterization of breast lesions with clinical reading at magnetic resonance imaging. MATERIALS AND METHODS: The institutional review board approved the use of data obtained prospectively and analyzed either prospectively with informed patient consent or retrospectively with waiver of consent. An existing computerized analysis system was retrained with 100 breast lesions (in 78 patients with mean age of 46.5 years) and tested with 136 other lesions (in 113 patients with mean age of 48.9 years; P=.15 for age difference between groups). Seventy-five lesions in the training set were previously rated by one of three radiologists in daily clinical practice. Lesion rating (as benign, probably benign, indeterminate, suspicious, or highly suggestive of malignancy) and probability of malignancy calculated with computerized analysis were included as covariates in logistic regression analysis to obtain a combined model. The performance of the model was compared with that of clinical reading alone in a set of 72 clinically and mammographically occult lesions not used to train the computerized analysis system (in 60 patients with mean age of 43.5 years; P=.09 for age difference between training and testing groups). Receiver operating characteristic (ROC) curves were plotted, and areas under the ROC curves were calculated and compared. RESULTS: Performance of reading in the clinical setting, as indicated by area under the ROC curve (Az=0.86), was similar to that of computerized analysis (Az=0.85; P=.99). Significant overall improvement in performance was obtained with the combined model (Az=0.91; P=.03). Improvement was accomplished mostly in characterization of lesions rated indeterminate or suspicious by radiologists. CONCLUSION: Computerized analysis complements clinical reading and makes computer-aided diagnosis feasible. The complementary information has the potential to increase overall performance for clinically and mammographically occult lesions.

Breast MR imaging in women at increased lifetime risk of breast cancer: clinical system for computerized assessment of breast lesions initial results.

The authors developed a clinical system for computerized delineation, rating, and classification of breast lesions depicted in contrast material-enhanced magnetic resonance images obtained in women with increased lifetime risk of breast cancer. Initial results showed negative predictive values above 98% at 50% positive predictive value with negligible interoperator differences. The system demonstrated potential to help exclude malignancy with high confidence and reproducibility with a positive predictive value that is acceptable in screening.