Ipsilateral Lesion Detection Refinement for Tomosynthesis.

Computer-aided detection (CAD) frameworks for breast cancer screening have been researched for several decades. Early adoption of deep-learning models in CAD frameworks has shown greatly improved detection performance compared to traditional CAD on single-view images. Recently, studies have improved performance by merging information from multiple views within each screening exam. Clinically, the integration of lesion correspondence during screening is a complicated decision process that depends on the correct execution of several referencing steps. However, most multi-view CAD frameworks are deep-learning-based black-box techniques. Fully end-to-end designs make it very difficult to analyze model behaviors and fine-tune performance. More importantly, the black-box nature of the techniques discourages clinical adoption due to the lack of explicit reasoning for each multi-view referencing step. Therefore, there is a need for a multi-view detection framework that can not only detect cancers accurately but also provide step-by-step, multi-view reasoning. In this work, we present Ipsilateral-Matching-Refinement Networks (IMR-Net) for digital breast tomosynthesis (DBT) lesion detection across multiple views. Our proposed framework adaptively refines the single-view detection scores based on explicit ipsilateral lesion matching. IMR-Net is built on a robust, single-view detection CAD pipeline with a commercial development DBT dataset of 24675 DBT volumetric views from 8034 exams. Performance is measured using location-based, case-level receiver operating characteristic (ROC) and case-level free-response ROC (FROC) analysis.

Systemic and Ocular Comorbidities of Black, Hispanic, and White Women with Cataracts.

Background: Cataracts are one of the leading causes of blindness in the world and disproportionately affect the elderly people and women. Sex- and race-related differences in cataract formation are not well understood. Furthermore, race and socioeconomic factors can play a role in developing systemic diseases. Earlier studies have supported a link between certain systemic diseases and cataract formation. Our study examined race-related differences in ocular and systemic comorbidities and analyzed differences among races and insurance types for cataract surgery visual outcomes among female patients with cataracts. Materials and Methods: Data were collected retrospectively and patients were grouped by race and insurance classifications. Female patients at a large tertiary center with an International Classification of Disease, 9th Edition (ICD-9) or ICD-10 cataract diagnosis or cataract extraction procedure code between January 2013 and June 2018 were included. A total of 909 female patients were included in the study. Frequency of systemic and ocular comorbidities was analyzed. Demographic factors were also compared among races. Finally, characteristics of cataract surgery patients, such as age at surgery, preoperative best-corrected visual acuity (BCVA), and visual outcomes among races and insurance types were analyzed. Results: There are differences among races for frequency of smoking, hemoglobin A1c, hypertension, and diabetes mellitus in female patients with cataracts and differences among races and insurance types for preoperative BCVA for patients who underwent cataract surgery (p < 0.001 for all). Conclusions: Female minority and non-minority patients with cataracts have a high frequency of systemic and ocular comorbidities at our county hospital. Patients with no insurance and white and Hispanic patients had worse preoperative BCVA.

Cataract Surgery Considerations for Diabetic Patients.

PURPOSE OF REVIEW: Given the epidemiology and demographic trends of diabetes mellitus and cataracts, ophthalmologists are likely to encounter patients with both comorbidities at an increasing frequency. Patients with diabetes represent a higher risk population than healthy patients for cataract surgery. In this review, we discuss key risks and risk-mitigation practices when performing cataract surgery on these patients. RECENT FINDINGS: Patients with diabetes continue to represent a high-risk surgical population: Nagar et al. suggest a dose-dependent relationship may exist between number of intravitreal injections and likelihood of posterior capsular rupture. However, novel treatments are improving outcomes for patients with diabetes. Several studies have reported intracameral phenylephrine/ketorolac may reduce the incidence of post-operative cystoid macular edema while others have discussed the efficacy of pre-treatment and post-treatment with intravitreal bevacizumab on improving cataract surgery outcomes in patients with diabetic retinopathy. Pre-operatively, ophthalmologists should perform an enhanced evaluation, consider timing and lens selection decisions, and complete any appropriate pre-operative treatment. Peri-operatively, surgeons should be aware of pupillary dilation adjustments, combination surgery options, and potential complications. Post-operatively, clinicians should address pseudophakic cystoid macular edema, diabetic macular edema, diabetic retinopathy, and posterior capsular opacification.

Orbital Varix Masquerading as an Intraorbital Lymphoma.

Clinical context was paramount to the diagnosis and management of a patient with periorbital pain and a history of systemic lymphoma.

Process Mapping and Activity-Based Costing of the Intravitreal Injection Procedure.

PURPOSE/AIM OF THE STUDY: To quantify the cost of performing an intravitreal injection (IVI) utilizing activity-based costing (ABC), which allocates a cost to each resource involved in a manufacturing process. MATERIALS AND METHODS: A prospective, observational cohort study was performed at an urban, multi-specialty ophthalmology practice affiliated with an academic institution. Fourteen patients scheduled for an IVI-only visit with a retina ophthalmologist were observed from clinic entry to exit to create a process map of time and resource utilization. Indirect costs were allocated with ABC and direct costs were estimated based on process map observations, internal accounting records, employee interviews, and nationally-reported metrics. The primary outcome measure was the cost of an IVI procedure in United States dollars. Secondary outcomes included operating income (cost subtracted from revenue) of an IVI and patient-centric time utilization for an IVI. RESULTS: The total cost of performing an IVI was $128.28; average direct material, direct labor, and overhead costs were $2.14, $97.88, and $28.26, respectively. Compared to the $104.40 reimbursement set by the Centers for Medicare and Medicaid Services for Current Procedural Terminology code 67028, this results in a negative operating income of -$23.88 (-22.87%). The median clinic resource-utilizing time to complete an IVI was 32:58 minutes (range [19:24-1:28:37]); the greatest bottleneck was physician-driven electronic health record documentation. CONCLUSIONS: Our study provides an objective and accurate cost estimate of the IVI procedure and illustrates how ABC may be applied in a clinical context. Our findings suggest that IVIs may currently be undervalued by payors.

A virtual COVID-19 ophthalmology rotation.

The coronavirus (COVID-19) pandemic temporarily suspended medical student involvement in clinical rotations, resulting in the need to develop virtual clinical experiences. The cancellation of clinical ophthalmology electives and away rotations reduces opportunities for exposure to the field, to network with faculty, conduct research, and prepare for residency applications. We review the literature and discuss the impact and consequences of COVID-19 on undergraduate medical education with an emphasis on ophthalmic undergraduate medical education. We also discuss innovative learning modalities used from medical schools around the world during the COVID-19 pandemic such as virtual didactics, online cases, and telehealth. Finally, we describe a novel, virtual neuro-ophthalmology elective created to educate medical students on neuro-ophthalmology foundational principles, provide research and presentation opportunities, and build relationships with faculty members. These innovative approaches represent a step forward in further improving medical education in ophthalmology during COVID-19 pandemic and beyond.

Improving Accuracy and Efficiency with Concurrent Use of Artificial Intelligence for Digital Breast Tomosynthesis.

PURPOSE: To evaluate the use of artificial intelligence (AI) to shorten digital breast tomosynthesis (DBT) reading time while maintaining or improving accuracy. MATERIALS AND METHODS: A deep learning AI system was developed to identify suspicious soft-tissue and calcified lesions in DBT images. A reader study compared the performance of 24 radiologists (13 of whom were breast subspecialists) reading 260 DBT examinations (including 65 cancer cases) both with and without AI. Readings occurred in two sessions separated by at least 4 weeks. Area under the receiver operating characteristic curve (AUC), reading time, sensitivity, specificity, and recall rate were evaluated with statistical methods for multireader, multicase studies. RESULTS: Radiologist performance for the detection of malignant lesions, measured by mean AUC, increased 0.057 with the use of AI (95% confidence interval [CI]: 0.028, 0.087; P < .01), from 0.795 without AI to 0.852 with AI. Reading time decreased 52.7% (95% CI: 41.8%, 61.5%; P < .01), from 64.1 seconds without to 30.4 seconds with AI. Sensitivity increased from 77.0% without AI to 85.0% with AI (8.0%; 95% CI: 2.6%, 13.4%; P < .01), specificity increased from 62.7% without to 69.6% with AI (6.9%; 95% CI: 3.0%, 10.8%; noninferiority P < .01), and recall rate for noncancers decreased from 38.0% without to 30.9% with AI (7.2%; 95% CI: 3.1%, 11.2%; noninferiority P < .01). CONCLUSION: The concurrent use of an accurate DBT AI system was found to improve cancer detection efficacy in a reader study that demonstrated increases in AUC, sensitivity, and specificity and a reduction in recall rate and reading time.© RSNA, 2019See also the commentary by Hsu and Hoyt in this issue.

PAX3-FOXO1 drives miR-486-5p and represses miR-221 contributing to pathogenesis of alveolar rhabdomyosarcoma.

Rhabdomyosarcoma is the most common soft-tissue sarcoma in childhood and histologically resembles developing skeletal muscle. Alveolar rhabdomyosarcoma (ARMS) is an aggressive subtype with a higher rate of metastasis and poorer prognosis. The majority of ARMS tumors (80%) harbor a PAX3-FOXO1 or less commonly a PAX7-FOXO1 fusion gene. The presence of either the PAX3-FOXO1 or PAX7-FOXO1 fusion gene foretells a poorer prognosis resulting in clinical re-classification as either fusion-positive (FP-RMS) or fusion-negative RMS (FN-RMS). The PAX3/7-FOXO1 fusion genes result in the production of a rogue transcription factors that drive FP-RMS pathogenesis and block myogenic differentiation. Despite knowing the molecular driver of FP-RMS, targeted therapies have yet to make an impact for patients, highlighting the need for a greater understanding of the molecular consequences of PAX3-FOXO1 and its target genes including microRNAs. Here we show FP-RMS patient-derived xenografts and cell lines display a distinct microRNA expression pattern. We utilized both loss- and gain-of function approaches in human cell lines with knockdown of PAX3-FOXO1 in FP-RMS cell lines and expression of PAX3-FOXO1 in human myoblasts and identified microRNAs both positively and negatively regulated by the PAX3-FOXO1 fusion protein. We demonstrate PAX3-FOXO1 represses miR-221/222 that functions as a tumor suppressing microRNA through the negative regulation of CCND2, CDK6, and ERBB3. In contrast, miR-486-5p is transcriptionally activated by PAX3-FOXO1 and promotes FP-RMS proliferation, invasion, and clonogenic growth. Inhibition of miR-486-5p in FP-RMS xenografts decreased tumor growth, illustrating a proof of principle for future therapeutic intervention. Therefore, PAX3-FOXO1 regulates key microRNAs that may represent novel therapeutic vulnerabilities in FP-RMS.

Biallelic Dicer1 Loss Mediated by aP2-Cre Drives Angiosarcoma.

Angiosarcoma is an aggressive vascular sarcoma with an extremely poor prognosis. Because of the relative rarity of this disease, its molecular drivers and optimal treatment strategies are obscure. DICER1 is an RNase III endoribonuclease central to miRNA biogenesis, and germline DICER1 mutations result in a cancer predisposition syndrome, associated with an increased risk of many tumor types. Here, we show that biallelic Dicer1 deletion with aP2-Cre drives aggressive and metastatic angiosarcoma independent of other genetically engineered oncogenes or tumor suppressor loss. Angiosarcomas in aP2-Cre;Dicer1Flox/- mice histologically and genetically resemble human angiosarcoma. miR-23 target genes, including the oncogenes Ccnd1 as well as Adam19, Plau, and Wsb1 that promote invasiveness and metastasis, were enriched in mouse and human angiosarcoma. These studies illustrate that Dicer1 can function as a traditional loss-of-function tumor suppressor gene, and they provide a fully penetrant animal model for the study of angiosarcoma development and metastasis. Cancer Res; 77(22); 6109-18. ©2017 AACR.