Potential Role of Convolutional Neural Network Based Algorithm in Patient Selection for DCIS Observation Trials Using a Mammogram Dataset.

RATIONALE AND OBJECTIVES: We investigated the feasibility of utilizing convolutional neural network (CNN) for predicting patients with pure Ductal Carcinoma In Situ (DCIS) versus DCIS with invasion using mammographic images. MATERIALS AND METHODS: An IRB-approved retrospective study was performed. 246 unique images from 123 patients were used for our CNN algorithm. In total, 164 images in 82 patients diagnosed with DCIS by stereotactic-guided biopsy of calcifications without any upgrade at the time of surgical excision (pure DCIS group). A total of 82 images in 41 patients with mammographic calcifications yielding occult invasive carcinoma as the final upgraded diagnosis on surgery (occult invasive group). Two standard mammographic magnification views (CC and ML/LM) of the calcifications were used for analysis. Calcifications were segmented using an open source software platform 3D Slicer and resized to fit a 128 × 128 pixel bounding box. A 15 hidden layer topology was used to implement the neural network. The network architecture contained five residual layers and dropout of 0.25 after each convolution. Five-fold cross validation was performed using training set (80%) and validation set (20%). Code was implemented in open source software Keras with TensorFlow on a Linux workstation with NVIDIA GTX 1070 Pascal GPU. RESULTS: Our CNN algorithm for predicting patients with pure DCIS achieved an overall diagnostic accuracy of 74.6% (95% CI, ±5) with area under the ROC curve of 0.71 (95% CI, ±0.04), specificity of 91.6% (95% CI, ±5%) and sensitivity of 49.4% (95% CI, ±6%). CONCLUSION: It's feasible to apply CNN to distinguish pure DCIS from DCIS with invasion with high specificity using mammographic images.

Brain metastases from prostate cancer: an 11-year analysis in the MRI era with emphasis on imaging characteristics, incidence, and prognosis.

BACKGROUND AND PURPOSE: Brain metastases from prostate cancer are uncommon and their imaging appearance has not been well defined. The main objectives of this study were to evaluate the incidence, MRI characteristics, and prognosis of parenchymal brain metastases originating in prostate cancer. METHODS: We retrospectively identified 21 patients with prostate cancer and evidence of brain metastases from 2000 to 2010. We reviewed the initial brain MRI scans and characterized the lesions according to location and appearance on MRI, while also determining patient demography, staging, and survival. RESULTS: The incidence of brain metastasis from prostate cancer was .16%. At the time of brain metastasis detection, 95% of the patients had concurrent osseous metastases, 86% lymph node metastases, and 76% liver and/or lung metastases. Brain metastases were multifocal in 71% of patients, hemorrhagic in 33%, diffusion restricted in 19%, and partially cystic/necrotic in 19%. The median overall survival after brain metastasis detection was 2.8 months. CONCLUSIONS: Brain metastasis from prostate cancer remains a rare phenomenon that most frequently occurs in the setting of widely disseminated bone and soft tissue disease. Patients with nonadenocarcinoma pathology are more likely to develop brain metastases. The MRI appearance is highly variable and prognosis is poor.

Mitogen-activated protein kinase-activated protein kinase 2 (MAPKAP-K2) as an antiinflammatory target: discovery and in vivo activity of selective pyrazolo[1,5-a]pyrimidine inhibitors using a focused library and structure-based optimization approach.

A novel class of mitogen-activated protein kinase-activated protein kinase 2 (MAPKAP-K2) inhibitors was discovered through screening a kinase-focused library. A homology model of MAPKAP-K2 was generated and used to guide the initial SAR studies and to rationalize the observed selectivity over CDK2. An X-ray crystal structure of a compound from the active series bound to crystalline MAPKAP-K2 confirmed the predicted binding mode. This has enabled the discovery of a series of pyrazolo[1,5-a]pyrimidine derivatives showing good in vitro cellular potency as anti-TNF-α agents and in vivo efficacy in a mouse model of endotoxin shock.

A retrospective analysis of joint salvage procedures for grades III and IV hallux rigidus.

UNLABELLED: The purpose of this study was to evaluate the outcome of the use of a decompression osteotomy for the treatment of end-stage hallux rigidus. We conducted a retrospective analysis of 28 feet (23 patients) with grades III and IV hallux rigidus that underwent a first metatarsal head decompression osteotomy with preservation of the articular surfaces of the first metatarsophalangeal joint. We also devised a 9-item questionnaire to explore the patients' perceptions of preoperative and postoperative pain, limitations of activity, influence on shoe wear, and the total range of motion of the first metatarsophalangeal joint. Furthermore, we used a modified version of the AOFAS forefoot scoring system to compare the patients' foot-related health status in relative to the operative repair of hallux rigidus. Comparisons of the pre- and postoperative results revealed statistically significant improvements in pain (P< .001), functional limitation (P< .001), shoe restrictions (P= .0072), total range of motion (P= .0449), and the AOFAS forefoot score (P< .001). Overall patient satisfaction with the results of the surgery was more than 85%, and the patients' chief complaint was alleviated in more than 75% of the participants. The results of this investigation demonstrated that a decompression first metatarsal osteotomy is an acceptable alternative to joint destructive procedures for the treatment of end-stage hallux rigidus. LEVEL OF CLINICAL EVIDENCE: 4.

The role of second autografts in the management of myeloma at first relapse.

We report an analysis of the value of a second high-dose melphalan autograft, performed at relapse, on a series of newly diagnosed myeloma patients entered into the high-dose program at our center. We conclude that relapse-free survival after the first autograft is a major prognostic factor in determining outcome.

Long-term outcomes of previously untreated myeloma patients: responses to induction chemotherapy and high-dose melphalan incorporated within a risk stratification model can help to direct the use of novel treatments.

Induction chemotherapy followed by high-dose melphalan (HDM) is the standard treatment for fitter patients with myeloma. The place of bortezomib and the thalidomide analogues within this treatment paradigm is yet to be established. We sought to identify patients who may benefit from the introduction of novel agents during their initial management. An intention-to-treat analysis was performed on 383 patients with newly diagnosed myeloma eligible for HDM to determine whether the extent of response to induction therapy and HDM correlated with long-term survival. Early response [complete response (CR) and partial response (PR)] to induction therapy was predictive of overall survival (OS) [median OS, 7.47 years for responders (CR and PR) versus 4.89 years for non-responders; P = 0.035]. The attainment of CR at 3 months post-HDM correlated with a prolonged progression-free survival (PFS) (median PFS, 7.4 years in CR group versus 5.3 years in non-CR group; P = 0.023). This data suggests that, at every stage of treatment, the aim should be to achieve CR. Patients with suboptimal responses could be offered alternative therapy. We propose a multiparametric risk-adapted model that includes response to induction chemotherapy and HDM, for identifying patients who may benefit from novel approaches to treatment.