Unsupervised feature correlation model to predict breast abnormal variation maps in longitudinal mammograms.

Breast cancer continues to be a significant cause of mortality among women globally. Timely identification and precise diagnosis of breast abnormalities are critical for enhancing patient prognosis. In this study, we focus on improving the early detection and accurate diagnosis of breast abnormalities, which is crucial for improving patient outcomes and reducing the mortality rate of breast cancer. To address the limitations of traditional screening methods, a novel unsupervised feature correlation network was developed to predict maps indicating breast abnormal variations using longitudinal 2D mammograms. The proposed model utilizes the reconstruction process of current year and prior year mammograms to extract tissue from different areas and analyze the differences between them to identify abnormal variations that may indicate the presence of cancer. The model incorporates a feature correlation module, an attention suppression gate, and a breast abnormality detection module, all working together to improve prediction accuracy. The proposed model not only provides breast abnormal variation maps but also distinguishes between normal and cancer mammograms, making it more advanced compared to the state-of-the-art baseline models. The results of the study show that the proposed model outperforms the baseline models in terms of Accuracy, Sensitivity, Specificity, Dice score, and cancer detection rate.

Feature fusion Siamese network for breast cancer detection comparing current and prior mammograms.

PURPOSE: Automatic detection of very small and nonmass abnormalities from mammogram images has remained challenging. In clinical practice for each patient, radiologists commonly not only screen the mammogram images obtained during the examination, but also compare them with previous mammogram images to make a clinical decision. To design an artificial intelligence (AI) system to mimic radiologists for better cancer detection, in this work we proposed an end-to-end enhanced Siamese convolutional neural network to detect breast cancer using previous year and current year mammogram images. METHODS: The proposed Siamese-based network uses high-resolution mammogram images and fuses features of pairs of previous year and current year mammogram images to predict cancer probabilities. The proposed approach is developed based on the concept of one-shot learning that learns the abnormal differences between current and prior images instead of abnormal objects, and as a result can perform better with small sample size data sets. We developed two variants of the proposed network. In the first model, to fuse the features of current and previous images, we designed an enhanced distance learning network that considers not only the overall distance, but also the pixel-wise distances between the features. In the other model, we concatenated the features of current and previous images to fuse them. RESULTS: We compared the performance of the proposed models with those of some baseline models that use current images only (ResNet and VGG) and also use current and prior images (long short-term memory [LSTM] and vanilla Siamese) in terms of accuracy, sensitivity, precision, F1 score, and area under the curve (AUC). Results show that the proposed models outperform the baseline models and the proposed model with the distance learning network performs the best (accuracy: 0.92, sensitivity: 0.93, precision: 0.91, specificity: 0.91, F1: 0.92 and AUC: 0.95). CONCLUSIONS: Integrating prior mammogram images improves automatic cancer classification, specially for very small and nonmass abnormalities. For classification models that integrate current and prior mammogram images, using an enhanced and effective distance learning network can advance the performance of the models.

Applying deep learning in digital breast tomosynthesis for automatic breast cancer detection: A review.

The relatively recent reintroduction of deep learning has been a revolutionary force in the interpretation of diagnostic imaging studies. However, the technology used to acquire those images is undergoing a revolution itself at the very same time. Digital breast tomosynthesis (DBT) is one such technology, which has transformed the field of breast imaging. DBT, a form of three-dimensional mammography, is rapidly replacing the traditional two-dimensional mammograms. These parallel developments in both the acquisition and interpretation of breast images present a unique case study in how modern AI systems can be designed to adapt to new imaging methods. They also present a unique opportunity for co-development of both technologies that can better improve the validity of results and patient outcomes. In this review, we explore the ways in which deep learning can be best integrated into breast cancer screening workflows using DBT. We first explain the principles behind DBT itself and why it has become the gold standard in breast screening. We then survey the foundations of deep learning methods in diagnostic imaging, and review the current state of research into AI-based DBT interpretation. Finally, we present some of the limitations of integrating AI into clinical practice and the opportunities these present in this burgeoning field.

Convolutional neural network for automated mass segmentation in mammography.

BACKGROUND: Automatic segmentation and localization of lesions in mammogram (MG) images are challenging even with employing advanced methods such as deep learning (DL) methods. We developed a new model based on the architecture of the semantic segmentation U-Net model to precisely segment mass lesions in MG images. The proposed end-to-end convolutional neural network (CNN) based model extracts contextual information by combining low-level and high-level features. We trained the proposed model using huge publicly available databases, (CBIS-DDSM, BCDR-01, and INbreast), and a private database from the University of Connecticut Health Center (UCHC). RESULTS: We compared the performance of the proposed model with those of the state-of-the-art DL models including the fully convolutional network (FCN), SegNet, Dilated-Net, original U-Net, and Faster R-CNN models and the conventional region growing (RG) method. The proposed Vanilla U-Net model outperforms the Faster R-CNN model significantly in terms of the runtime and the Intersection over Union metric (IOU). Training with digitized film-based and fully digitized MG images, the proposed Vanilla U-Net model achieves a mean test accuracy of 92.6%. The proposed model achieves a mean Dice coefficient index (DI) of 0.951 and a mean IOU of 0.909 that show how close the output segments are to the corresponding lesions in the ground truth maps. Data augmentation has been very effective in our experiments resulting in an increase in the mean DI and the mean IOU from 0.922 to 0.951 and 0.856 to 0.909, respectively. CONCLUSIONS: The proposed Vanilla U-Net based model can be used for precise segmentation of masses in MG images. This is because the segmentation process incorporates more multi-scale spatial context, and captures more local and global context to predict a precise pixel-wise segmentation map of an input full MG image. These detected maps can help radiologists in differentiating benign and malignant lesions depend on the lesion shapes. We show that using transfer learning, introducing augmentation, and modifying the architecture of the original model results in better performance in terms of the mean accuracy, the mean DI, and the mean IOU in detecting mass lesion compared to the other DL and the conventional models.

Deep convolutional neural networks for mammography: advances, challenges and applications.

BACKGROUND: The limitations of traditional computer-aided detection (CAD) systems for mammography, the extreme importance of early detection of breast cancer and the high impact of the false diagnosis of patients drive researchers to investigate deep learning (DL) methods for mammograms (MGs). Recent breakthroughs in DL, in particular, convolutional neural networks (CNNs) have achieved remarkable advances in the medical fields. Specifically, CNNs are used in mammography for lesion localization and detection, risk assessment, image retrieval, and classification tasks. CNNs also help radiologists providing more accurate diagnosis by delivering precise quantitative analysis of suspicious lesions. RESULTS: In this survey, we conducted a detailed review of the strengths, limitations, and performance of the most recent CNNs applications in analyzing MG images. It summarizes 83 research studies for applying CNNs on various tasks in mammography. It focuses on finding the best practices used in these research studies to improve the diagnosis accuracy. This survey also provides a deep insight into the architecture of CNNs used for various tasks. Furthermore, it describes the most common publicly available MG repositories and highlights their main features and strengths. CONCLUSIONS: The mammography research community can utilize this survey as a basis for their current and future studies. The given comparison among common publicly available MG repositories guides the community to select the most appropriate database for their application(s). Moreover, this survey lists the best practices that improve the performance of CNNs including the pre-processing of images and the use of multi-view images. In addition, other listed techniques like transfer learning (TL), data augmentation, batch normalization, and dropout are appealing solutions to reduce overfitting and increase the generalization of the CNN models. Finally, this survey identifies the research challenges and directions that require further investigations by the community.

Central Core Laboratory versus Site Interpretation of Coronary CT Angiography: Agreement and Association with Cardiovascular Events in the PROMISE Trial.

Purpose To assess concordance and relative prognostic utility between central core laboratory and local site interpretation for significant coronary artery disease (CAD) and cardiovascular events. Materials and Methods In the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) trial, readers at 193 North American sites interpreted coronary computed tomographic (CT) angiography as part of the clinical evaluation of stable chest pain. Readers at a central core laboratory also interpreted CT angiography blinded to clinical data, site interpretation, and outcomes. Significant CAD was defined as stenosis greater than or equal to 50%; cardiovascular events were defined as a composite of cardiovascular death or myocardial infarction. Results In 4347 patients (51.8% women; mean age ± standard deviation, 60.4 years ± 8.2), core laboratory and site interpretations were discordant in 16% (683 of 4347), most commonly because of a finding of significant CAD by site but not by core laboratory interpretation (80%, 544 of 683). Overall, core laboratory interpretation resulted in 41% fewer patients being reported as having significant CAD (14%, 595 of 4347 vs 23%, 1000 of 4347; P < .001). Over a median follow-up period of 25 months, 1.3% (57 of 4347) sustained myocardial infarction or cardiovascular death. The C statistic for future myocardial infarction or cardiovascular death was 0.61 (95% confidence interval [CI]: 0.54, 0.68) for the core laboratory and 0.63 (95% CI: 0.56, 0.70) for the sites. Conclusion Compared with interpretation by readers at 193 North American sites, standardized core laboratory interpretation classified 41% fewer patients as having significant CAD. © RSNA, 2017 Online supplemental material is available for this article. Clinical trial registration no. NCT01174550.

Effect of Adjuvant Chemotherapy on Left Ventricular Remodeling in Women with Newly Diagnosed Primary Breast Cancer: A Pilot Prospective Longitudinal Cardiac Magnetic Resonance Imaging Study.

PURPOSE: The aim of this study was to assess the left ventricular (LV) remodeling response to chemotherapy in low-cardiac-risk women with newly diagnosed nonmetastatic breast cancer. Cardiotoxic effects of chemotherapy are an increasing concern. To effectively interpret cardiac imaging studies performed for screening purposes in patients undergoing cancer therapy it is necessary to understand the normal changes in structure and function that may occur. METHODS: Twenty women without preexisting cardiovascular disease, of a mean age of 50 years, newly diagnosed with nonmetastatic breast cancer and treated with anthracycline or trastuzumab, were prospectively enrolled and evaluated at four time points (at baseline, during chemotherapy, 2 weeks after chemotherapy, and 6 months after chemotherapy) using cardiac magnetic resonance imaging, blood samples, and a clinical questionnaire. RESULTS: Over a 6-month period, the left ventricular ejection fraction (%) decreased (64.15±5.30 to 60.41±5.77, P<0.002) and the LV end-diastolic (mm) and end-systolic (mm) volumes increased (124.73±20.25 to 132.21±19.33, P<0.04 and 45.16±11.88 to 52.57±11.65, P<0.00, respectively). The LV mass (g) did not change (73.06±11.51 to 69.21±15.3, P=0.08), but the LV mass to LVEDV ratio (g/mm) decreased (0.594±0.098 to 0.530±0.124, P<0.04). CONCLUSIONS: In low-cardiac-risk women with nonmetastatic breast cancer, the increased LV volume and a mildly decreased left ventricular ejection fraction during and after chemotherapy do not seem to be associated with laboratory or clinical evidence of increased risk for heart failure.

Doxorubicin-induced cardiomyopathy 17 years after chemotherapy.

Doxorubicin, an anthracycline antibiotic commonly used as a chemotherapeutic agent for breast cancer, is well known to cause cardiotoxicity. We report the case of an active, otherwise healthy 57-year-old breast cancer survivor who, 17 years after chemotherapy, presented with symptoms of overt heart failure. She had no cardiac risk factors, and neither laboratory nor imaging findings suggested myocarditis or dilated cardiomyopathy. Echocardiographic findings and differential diagnosis led us to attribute her condition to late doxorubicin-induced cardiomyopathy. By virtue of tapered medical therapy, her left ventricular ejection fraction improved from 0.20 to 0.55 in 8 months, and she was asymptomatic after 1 year. The reversibility of left ventricular dysfunction in our patient and the very late appearance of cardiotoxicity secondary to doxorubicin therapy raise questions about the pathogenesis and prevalence of late doxorubicin-induced cardiomyopathy and how to improve outcomes in patients who present with related symptoms of heart failure.

Posterolateral corner reconstruction of the knee: evaluation of a technique with clinical outcomes and stress radiography.

BACKGROUND: Injuries to the posterolateral corner of the knee remain a challenging problem and have been cited frequently as a reason for failure of anterior and posterior cruciate ligament reconstructions. Although several reconstructive techniques currently exist, there are relatively few clinical outcomes data after reconstruction of the posterolateral corner. PURPOSE: The study was undertaken to examine the clinical outcomes and provide objective data using arthrometry and stress radiography of a posterolateral corner reconstruction technique. STUDY DESIGN: Case series; Level of evidence, 4. METHODS: A retrospective cohort study of a consecutive series of patients who underwent posterolateral corner reconstruction of the knee was evaluated. The surgery featured dual femoral tunnels, a transfibular tunnel, and a free graft to reconstruct the posterolateral corner of the knee. All patients had concomitant reconstruction of one or both cruciate ligaments. Outcomes were assessed using the Short Form-12, Lysholm, and Tegner knee scores. A clinical examination, KT-2000 arthrometry measurements, single-legged hop quotient, and varus and posterior Telos stress radiographs were obtained and compared with results for the contralateral, uninjured knees. RESULTS: Twenty-four (83%) of 29 consecutive patients were evaluated at a mean 39 months postoperatively (range, 24-81 months). The mean Lysholm and Tegner knee scores were 83 and 6, respectively. The mean difference (+/- standard deviation) in total anterior-posterior side-to-side KT arthrometry measurements was 1.4 +/- 1.3 mm. The varus stress radiographic mean side-to-side difference measured at 20 degrees of flexion was 0.2 +/- 1.9 mm. The mean radiographic posterior tibial displacement with a 15-kg stress at 90 degrees of flexion was 3.2 +/- 4.5 mm in patients undergoing posterior cruciate ligament reconstruction. CONCLUSION: This reconstruction of the posterolateral corner of the knee with concomitant cruciate ligament reconstruction restores varus and rotational stability at a minimum of 2 years postoperatively.

Neuroimaging findings and brain-behavioral correlates in a former boxer with chronic traumatic brain injury.

Chronic traumatic brain injury (CTBI) is associated with contact sports such as boxing. CTBI results from repetitive blows to the head rather than from a single impact. CTBI individuals present with motor symptoms (incoordination, spasticity, parkinsonism), cognitive impairment (executive dysfunction, memory deficits) and neuropsychiatric symptoms (irritability, affective disturbances). The structural and functional neuroimaging findings and clinical presentation of a CTBI case are described. We propose hypotheses about the pathophysiology of the observed neuroimaging findings and their relationship to the neuropsychiatric symptoms of the patients.

The 'cardiac-lung mass' artifact: an echocardiographic sign of lung atelectasis and/or pleural effusion.

INTRODUCTION: We conducted an ultrasound study to investigate echocardiographic artifacts in mechanically ventilated patients with lung pathology. METHODS: A total of 205 mechanically ventilated patients who exhibited lung atelectasis and/or pleural effusion were included in this 36-month study. The patients underwent lung echography and transthoracic echocardiography, with a linear 5 to 10 MHz and with a 1.5 to 3.6 MHz wide-angle phased-array transducer, respectively. Patients were examined by two experienced observers who were blinded to each other's interpretation. RESULTS: A total of 124 patients (60,48%) were hospitalized because of multiple trauma; 60 patients (29,26%) because of respiratory insufficiency, and 21 (10,24%) because of recent postoperative surgery. The mean duration ( +/- standard deviation) of hospitalization was 35 +/- 27 days. An intracardiac artifact was documented in 17 out of 205 patients (8.29%) by echocardiography. It was visible only in the apical views, whereas subsequent transesophageal echocardiography revealed no abnormalities. The artifact consisted of a mobile component that exhibited, on M-mode, a pattern of respiratory variation similar to the lung 'sinusoid sign'. Lung echography revealed lung atelectasis and/or pleural effusion adjacent to the heart, and a similar M-mode pattern was observed. The artifact was recorded within the left cardiac chambers in 11 cases and within the right cardiac chambers in six. CONCLUSIONS: Lung atelectasis and/or pleural effusion may create a mirror image, intracardiac artifact in mechanically ventilated patients. The latter was named the 'cardiac-lung mass' artifact to underline the important diagnostic role of both echocardiography and lung echography in these patients. TRIAL REGISTRATION: This trial is ISRCTN registered: ISRCTN 49216096.

Alcohol dementia and thermal dysregulation: a case report and review of the literature.

Wernicke's encephalopathy and Korsakoff's psychosis in alcoholics are thought to be due to thiamine deficiency. When the process goes untreated, patients may develop alcohol-induced persisting dementia. We review the literature on thermal dysregulation and the place of thiamine treatment in Wernicke's encephalopathy, Korsakoff's psychosis, and alcohol-induced persisting dementia. We describe a patient with alcohol-induced persisting dementia who showed thermal dysregulation which responded to parenteral but not oral thiamine. Subsequently, he developed aspiration pneumonia with associated fever reaction and expired. We describe the neuroimaging findings--diffuse cortical atrophy, ventricular dilatation, atrophy of the corpus callosum, hypothalamus, and medulla, and a probable arachnoid cyst in the left temporal tip. We conclude that thermal dysregulation was likely related to dysfunction of temperature regulatory brain centers, that thermal dysregulation was stabilized with parenteral but not oral thiamine, and that parenteral thiamine may have a role even in chronic cases of alcohol-induced persisting dementia.

Electroconvulsive therapy and venous angiomas: two case reports and review of the literature.

According to the American Psychiatric Association, the risk for complications related to the electroconvulsive therapy (ECT) treatment of patients with cerebrovascular malformations is small. The literature contains a number of case studies presenting the uneventful treatment of patients with cerebral aneurysms with ECT. However, there is a paucity of cases presenting ECT in the context of a cerebral venous angioma. In this article, we present 2 cases of patients treated with ECT who were found to have documented venous angiomas. This is followed with a brief review of the literature.

Germinoma-unusual presentation: a case report.

Germinoma accounts for two-thirds of germ cell tumors and about 40% of all pineal region neoplasms. This case illustrates an unusual manifestation of metatastic germinoma with spread to ventricles and meninges without a pineal mass. A 24-year-old man presented with nausea, vertigo, and left facial droop. Cerebrospinal fluid aspirate showed malignant cells most suggestive of a germinoma.