A review of deep learning-based information fusion techniques for multimodal medical image classification.

Multimodal medical imaging plays a pivotal role in clinical diagnosis and research, as it combines information from various imaging modalities to provide a more comprehensive understanding of the underlying pathology. Recently, deep learning-based multimodal fusion techniques have emerged as powerful tools for improving medical image classification. This review offers a thorough analysis of the developments in deep learning-based multimodal fusion for medical classification tasks. We explore the complementary relationships among prevalent clinical modalities and outline three main fusion schemes for multimodal classification networks: input fusion, intermediate fusion (encompassing single-level fusion, hierarchical fusion, and attention-based fusion), and output fusion. By evaluating the performance of these fusion techniques, we provide insight into the suitability of different network architectures for various multimodal fusion scenarios and application domains. Furthermore, we delve into challenges related to network architecture selection, handling incomplete multimodal data management, and the potential limitations of multimodal fusion. Finally, we spotlight the promising future of Transformer-based multimodal fusion techniques and give recommendations for future research in this rapidly evolving field.

Automated tear film break-up time measurement for dry eye diagnosis using deep learning.

In the realm of ophthalmology, precise measurement of tear film break-up time (TBUT) plays a crucial role in diagnosing dry eye disease (DED). This study aims to introduce an automated approach utilizing artificial intelligence (AI) to mitigate subjectivity and enhance the reliability of TBUT measurement. We employed a dataset of 47 slit lamp videos for development, while a test dataset of 20 slit lamp videos was used for evaluating the proposed approach. The multistep approach for TBUT estimation involves the utilization of a Dual-Task Siamese Network for classifying video frames into tear film breakup or non-breakup categories. Subsequently, a postprocessing step incorporates a Gaussian filter to smooth the instant breakup/non-breakup predictions effectively. Applying a threshold to the smoothed predictions identifies the initiation of tear film breakup. Our proposed method demonstrates on the evaluation dataset a precise breakup/non-breakup classification of video frames, achieving an Area Under the Curve of 0.870. At the video level, we observed a strong Pearson correlation coefficient (r) of 0.81 between TBUT assessments conducted using our approach and the ground truth. These findings underscore the potential of AI-based approaches in quantifying TBUT, presenting a promising avenue for advancing diagnostic methodologies in ophthalmology.

DISCOVER: 2-D multiview summarization of Optical Coherence Tomography Angiography for automatic diabetic retinopathy diagnosis.

Diabetic Retinopathy (DR), an ocular complication of diabetes, is a leading cause of blindness worldwide. Traditionally, DR is monitored using Color Fundus Photography (CFP), a widespread 2-D imaging modality. However, DR classifications based on CFP have poor predictive power, resulting in suboptimal DR management. Optical Coherence Tomography Angiography (OCTA) is a recent 3-D imaging modality offering enhanced structural and functional information (blood flow) with a wider field of view. This paper investigates automatic DR severity assessment using 3-D OCTA. A straightforward solution to this task is a 3-D neural network classifier. However, 3-D architectures have numerous parameters and typically require many training samples. A lighter solution consists in using 2-D neural network classifiers processing 2-D en-face (or frontal) projections and/or 2-D cross-sectional slices. Such an approach mimics the way ophthalmologists analyze OCTA acquisitions: (1) en-face flow maps are often used to detect avascular zones and neovascularization, and (2) cross-sectional slices are commonly analyzed to detect macular edemas, for instance. However, arbitrary data reduction or selection might result in information loss. Two complementary strategies are thus proposed to optimally summarize OCTA volumes with 2-D images: (1) a parametric en-face projection optimized through deep learning and (2) a cross-sectional slice selection process controlled through gradient-based attribution. The full summarization and DR classification pipeline is trained from end to end. The automatic 2-D summary can be displayed in a viewer or printed in a report to support the decision. We show that the proposed 2-D summarization and classification pipeline outperforms direct 3-D classification with the advantage of improved interpretability.

Federated Learning for Diabetic Retinopathy Detection in a Multi-center Fundus Screening Network.

Federated learning (FL) is a machine learning framework that allows remote clients to collaboratively learn a global model while keeping their training data localized. It has emerged as an effective tool to solve the problem of data privacy protection. In particular, in the medical field, it is gaining relevance for achieving collaborative learning while protecting sensitive data. In this work, we demonstrate the feasibility of FL in the development of a deep learning model for screening diabetic retinopathy (DR) in fundus photographs. To this end, we conduct a simulated FL framework using nearly 700,000 fundus photographs collected from OPHDIAT, a French multi-center screening network for detecting DR. We develop two FL algorithms: 1) a cross-center FL algorithm using data distributed across the OPHDIAT centers and 2) a cross-grader FL algorithm using data distributed across the OPHDIAT graders. We explore and assess different FL strategies and compare them to a conventional learning algorithm, namely centralized learning (CL), where all the data is stored in a centralized repository. For the task of referable DR detection, our simulated FL algorithms achieved similar performance to CL, in terms of area under the ROC curve (AUC): AUC =0.9482 for CL, AUC = 0.9317 for cross-center FL and AUC = 0.9522 for cross-grader FL. Our work indicates that the FL algorithm is a viable and reliable framework that can be applied in a screening network.Clinical relevance- Given that data sharing is regarded as an essential component of modern medical research, achieving collaborative learning while protecting sensitive data is key.

Quantitative gait analysis and prediction using artificial intelligence for patients with gait disorders.

Quantitative Gait Analysis (QGA) is considered as an objective measure of gait performance. In this study, we aim at designing an artificial intelligence that can efficiently predict the progression of gait quality using kinematic data obtained from QGA. For this purpose, a gait database collected from 734 patients with gait disorders is used. As the patient walks, kinematic data is collected during the gait session. This data is processed to generate the Gait Profile Score (GPS) for each gait cycle. Tracking potential GPS variations enables detecting changes in gait quality. In this regard, our work is driven by predicting such future variations. Two approaches were considered: signal-based and image-based. The signal-based one uses raw gait cycles, while the image-based one employs a two-dimensional Fast Fourier Transform (2D FFT) representation of gait cycles. Several architectures were developed, and the obtained Area Under the Curve (AUC) was above 0.72 for both approaches. To the best of our knowledge, our study is the first to apply neural networks for gait prediction tasks.

Hybrid Fusion of High-Resolution and Ultra-Widefield OCTA Acquisitions for the Automatic Diagnosis of Diabetic Retinopathy.

Optical coherence tomography angiography (OCTA) can deliver enhanced diagnosis for diabetic retinopathy (DR). This study evaluated a deep learning (DL) algorithm for automatic DR severity assessment using high-resolution and ultra-widefield (UWF) OCTA. Diabetic patients were examined with 6×6 mm2 high-resolution OCTA and 15×15 mm2 UWF-OCTA using PLEX®Elite 9000. A novel DL algorithm was trained for automatic DR severity inference using both OCTA acquisitions. The algorithm employed a unique hybrid fusion framework, integrating structural and flow information from both acquisitions. It was trained on data from 875 eyes of 444 patients. Tested on 53 patients (97 eyes), the algorithm achieved a good area under the receiver operating characteristic curve (AUC) for detecting DR (0.8868), moderate non-proliferative DR (0.8276), severe non-proliferative DR (0.8376), and proliferative/treated DR (0.9070). These results significantly outperformed detection with the 6×6 mm2 (AUC = 0.8462, 0.7793, 0.7889, and 0.8104, respectively) or 15×15 mm2 (AUC = 0.8251, 0.7745, 0.7967, and 0.8786, respectively) acquisitions alone. Thus, combining high-resolution and UWF-OCTA acquisitions holds the potential for improved early and late-stage DR detection, offering a foundation for enhancing DR management and a clear path for future works involving expanded datasets and integrating additional imaging modalities.

Towards population-independent, multi-disease detection in fundus photographs.

Independent validation studies of automatic diabetic retinopathy screening systems have recently shown a drop of screening performance on external data. Beyond diabetic retinopathy, this study investigates the generalizability of deep learning (DL) algorithms for screening various ocular anomalies in fundus photographs, across heterogeneous populations and imaging protocols. The following datasets are considered: OPHDIAT (France, diabetic population), OphtaMaine (France, general population), RIADD (India, general population) and ODIR (China, general population). Two multi-disease DL algorithms were developed: a Single-Dataset (SD) network, trained on the largest dataset (OPHDIAT), and a Multiple-Dataset (MD) network, trained on multiple datasets simultaneously. To assess their generalizability, both algorithms were evaluated whenever training and test data originate from overlapping datasets or from disjoint datasets. The SD network achieved a mean per-disease area under the receiver operating characteristic curve (mAUC) of 0.9571 on OPHDIAT. However, it generalized poorly to the other three datasets (mAUC < 0.9). When all four datasets were involved in training, the MD network significantly outperformed the SD network (p = 0.0058), indicating improved generality. However, in leave-one-dataset-out experiments, performance of the MD network was significantly lower on populations unseen during training than on populations involved in training (p < 0.0001), indicating imperfect generalizability.

French national diagnostic and care protocol for Sjögren's disease.

Sjögren's disease (SD), also known as Sjögren's syndrome (SS) or Gougerot-Sjögren's syndrome in France, is a rare systemic autoimmune disease in its primary form and is characterised by tropism for the exocrine glandular epithelia, particularly the salivary and lacrimal glands. The lymphocytic infiltration of these epithelia will clinically translate into a dry syndrome which, associated with fatigue and pain, constitutes the symptom triad of the disease. In about one third of patients, SD is associated with systemic complications that can affect the joints, skin, lungs, kidneys, central or peripheral nervous system, and lymphoid organs with an increased risk of B-cell lymphoma. SD affects women more frequently than men (9/1). The peak frequency is around the age of 50. However, the disease can occur at any age, with paediatric forms occurring even though they remain rare. SD can occur alone or in association with other systemic autoimmune diseases. In its isolated or primary form, the prevalence of SD is estimated to be between 1 per 1000 and 1 per 10,000 inhabitants. The most recent classification criteria were developed in 2016 by EULAR and ACR. The course and prognosis of the disease are highly variable and depend on the presence of systemic involvement and the severity of the dryness of the eyes and mouth. The current approach is therefore to identify at an early stage those patients most at risk of systemic complications or lymphoma, who require close follow-up. On the other hand, regular monitoring of the ophthalmological damage and of the dental status should be ensured to reduce the consequences.

Cost Utility and Value of Information Analysis of Femtosecond Laser-Assisted Cataract Surgery.

Importance: The efficacy and safety of femtosecond laser-assisted cataract surgery is well documented. An important requirement for decision makers is the evaluation of the cost-effectiveness of femtosecond laser-assisted cataract surgery (FLACS) over a sufficiently long horizon. Evaluating the cost-effectiveness of this treatment was a preplanned secondary objective of the Economic Evaluation of Femtosecond Laser Assisted Cataract Surgery (FEMCAT) trial. Objective: To estimate the cost utility of FLACS compared with phacoemulsification cataract surgery (PCS) on a 12-month time horizon. Design, Setting, and Participants: This multicenter randomized clinical trial compared FLACS with PCS in parallel groups. All FLACS procedures were performed using the CATALYS precision system. Participants were recruited and treated in ambulatory surgery settings in 5 university-hospital centers in France. All consecutive patients eligible for a unilateral or bilateral cataract surgery 22 years or older with written informed consent were included. Data were collected from October 2013 to October 2018, and data were analyzed from January 2020 to June 2022. Interventions: FLACS or PCS. Main Outcomes and Measures: Utility was measured through the Health Utility Index questionnaire. Costs of cataract surgery were estimated by microcosting. All inpatient and outpatient costs were collected from the French National Health Data System. Results: Of 870 randomized patients, 543 (62.4%) were female, and the mean (SD) age at surgery was 72.3 (8.6) years. A total of 440 patients were randomized to receive FLACS and 430 to receive PCS; the rate of bilateral surgery was 63.3% (551 of 870). The mean (SD) costs of cataract surgery were €1124.0 (€162.2; US $1235) for FLACS and €565.5 (€61.4; US $621) for PCS. The total mean (SD) cost of care at 12 months was €7085 (€6700; US $7787) in participants treated with FLACS and €6502 (€7323; US $7146) in participants treated with PCS. FLACS yielded a mean (SD) of 0.788 (0.009) quality-adjusted life-years (QALYs), and PCS yielded 0.792 (0.009) QALYs. The difference in mean costs was €545.9 (95% CI, -434.1 to 1525.8; US $600), and the difference in QALYs was -0.004 (95% CI, -0.028 to 0.021). The incremental cost-effectiveness ratio (ICER) was -€136 476 (US $150 000) per QALY. The cost-effectiveness probability of FLACS compared with PCS was 15.7% for a cost-effectiveness threshold of €30 000 (US $32 973) per QALY. At this threshold, the expected value of perfect information was €246 139 079 (US $270 530 231). Conclusions and Relevance: The ICER of FLACS compared with PCS was not within the $50 000 to $100 000 per QALY range frequently cited as cost-effective. Additional research and development on FLACS is needed to improve its effectiveness and lower its price. Trial Registration: ClinicalTrials.gov Identifier: NCT01982006.

Keratoconus and the Impact of Treatment on Patients' Quality of Life: A Qualitative Study.

INTRODUCTION: Keratoconus has a significant impact on patients' quality of life (QoL), from diagnosis to the advanced stages of the disease. The aim of this research was to identify domains of QoL affected by this disease and its treatment. METHODS: Phone interviews were conducted using a semi-structured interview guide, with patients with keratoconus stratified according to their current treatment. A board of keratoconus experts helped identify the guide's main themes. RESULTS: Thirty-five patients (rigid contact lenses, n = 9; cross-linking, n = 9; corneal ring implants, n = 8; and corneal transplantation, n = 9) were interviewed by qualitative researchers. Phone interviews revealed several QoL domains affected by the disease and its treatments: "psychological", "social life", "professional life", "financial costs" and "student life". All domains were impacted, independently of the treatment history. Few differences were found between treatment regimens and keratoconus stages. Qualitative analysis enabled the development of a conceptual framework based on Wilson and Cleary's model for patient outcomes common to all patients. This conceptual model describes the relationship between patients' characteristics, their symptoms, their environment, their functional visual impairment and the impact on their QoL. CONCLUSIONS: These qualitative findings supported the generation of a questionnaire to evaluate the impact of keratoconus and its treatment on patients' QoL. Cognitive debriefings confirmed its content validity. The questionnaire is applicable for all stages of keratoconus and treatments and may help tracking change over time in regular clinical settings. Psychometric validation is yet to be performed before its use in research and clinical practices.

Sodium hyaluronate 0.30% ocular gel versus sodium hyaluronate 0.18% eye drop in the treatment of moderate to severe dry eye disease.

PURPOSE: Compare 0.30% sodium hyaluronate (0.30%HA) ocular gel with 0.18%HA eye drops in terms of improvement of ocular signs and symptoms, in patients with moderate to severe dry eye disease (DED). METHODS: This was a multicentric, randomized, investigator-masked, non-inferiority, comparative study conducted over 84 days. Three visits were scheduled, testing fluorescein corneal and conjunctival staining (Oxford and Van Bijsterveld scores), tear film break-up time (TBUT), Schirmer test, DED symptoms, 5-Item-Dry-Eye-Questionnaire (5-DEQ), patient and investigator satisfaction and frequency of instillation. RESULTS: At Day 35 (D35) and Day 84 (D84), both groups (n = 35 each) had a significant improvement in corneal staining (p < 0.001) with no inter-group difference. Van Bijsterveld score improved earlier (D35) for 0.30%HA suggesting a faster effect on conjunctival epithelium healing. There was no difference between the two concentrations in terms of TBUT or Schirmer improvements; however, the Schirmer test increase was only significant for 0.30%HA at D35 (p = 0.040). At D35 and D84, both groups showed similar improvements of DED symptoms and DEQ-5 score. Furthermore, treatment satisfaction was similar for the 2 formulations suggesting that daily use of 0.30%HA do not cause gel-related blurred vision disturbances. Frequency of instillation was similar for both groups. CONCLUSION: Our study demonstrates the non-inferiority of 0.30%HA gel compared to 0.18%HA solution in patients with moderate to severe DED. Because of its gel formulation and higher HA concentration providing prolonged comfort without causing visual disturbances, 0.30%HA gel might be adapted for bedtime use or during the day in more severe conditions.

SMILE for Hyperopia With and Without Astigmatism: Results of a Prospective Multicenter 12-Month Study.

PURPOSE: To investigate the safety and effectiveness of small incision lenticule extraction (SMILE) in patients who have hyperopia with or without astigmatism. METHODS: This was a prospective multicenter trial including 374 eyes of 199 patients treated by SMILE for hyperopia using the VisuMax femtosecond laser (Carl Zeiss Meditec AG). Inclusion criteria were sphere up to +6.00 diopters (D), cylinder up to 5.00 D, and maximum hyperopic meridian up to +7.00 D, with preoperative corrected distance visual acuity (CDVA) of 20/25 or better. The optical zone was 6.3 mm with a transition zone of 2 mm. The minimum lenticule thickness was set at 25 µm in the center and at 10 µm at the edge. Patients were examined at 1 day, 1 week, and 1, 3, 6, 9, and 12 months after surgery. Standard refractive surgery outcomes analysis was performed. RESULTS: The preoperative spherical equivalent was +3.20 ± 1.48 D (range: +0.25 to +6.50 D). At the 12-month follow-up visit, 81% of eyes treated were within ±0.50 D and 93% of eyes were within ±1.00 D of intended correction. A total of 1.2% of eyes lost two or more lines of CDVA at the 12-month follow-up visit, and 83% were at least 20/20, corresponding to a safety index of 1.005 at 12 months. Of the 219 eyes with plano target, 68.8% had an uncorrected distance visual acuity of 20/20 or better and 88% were at least 20/25 uncorrected at 12 months. There were no statistically significant changes in contrast sensitivity. CONCLUSIONS: SMILE was found to be an effective treatment method for the correction of compound hyperopic astigmatism, demonstrating a high level of efficacy, predictability, safety, and stability. [J Refract Surg. 2022;38(12):760-769.].

Comparison of Bilateral Reading Performance Among Two Presbyopia-Correcting Intraocular Lenses.

PURPOSE: To evaluate and compare the performance of a trifocal diffractive intraocular lens (IOL) and a lens combining a bifocal diffractive profile and extended depth of focus (EDOF) profile. METHODS: This non-randomized, prospective comparative study included 42 patients (84 eyes) undergoing lens surgery with implantation of either the FineVision HP trifocal IOL (PhysIOL) or TECNIS Synergy bifocal EDOF IOL (Johnson and Johnson Surgical Vision). There were 21 patients (42 eyes) in each group. The primary outcome was reading speed at high contrast and luminance. Secondary outcomes were reading speed at lower contrasts and luminances, visual acuity at all distances (distance, intermediate, and near) with and without correction, and quality of vision. RESULTS: The reading speed at high contrast (100%) and high luminance (100%) was better in the Synergy group (P = .01). This difference between the two IOLs seemed to be preserved at lower contrasts and luminances. There was no statistically significant difference between visual acuities except for monocular uncorrected intermediate visual acuity (P = .046) in favor of the FineVision HP IOL. The mean spherical equivalents in the FineVision HP and Synergy groups were 0.14 ± 0.64 and 0.10 ± 0.33 diopters without significant difference between these means (P = .78). The defocus curve was more dome-shaped for the Synergy IOL. The evaluation of visual symptoms was comparable in both groups. The glare halo (Halometry test; Aston University) was smaller in the FineVision HP group (P = .03). CONCLUSIONS: The Synergy IOL appears to provide better reading speed and is less sensitive to refractive error. Both lenses provided excellent distance, intermediate, and near vision. [J Refract Surg. 2022;38(7):428-434.].

Association of Sensitive Eyes with Sensitive Skin: A Worldwide Study of 10,743 Subjects.

INTRODUCTION: Sensitive eyes are commonly reported by patients, but there are very few epidemiological studies on this disorder. The aim of this study was the evaluation of the self-reported frequency of sensitive eyes and the association with sensitive skin. METHODS: A survey was performed on a representative sample of the population aged more than 18 years in five different countries (Brazil, China, France, Russia, and the USA). All participants answered a questionnaire on sociodemographic characteristics; skin phototype; eye color; tobacco consumption; exposure to sunlight, air pollution, or having pets; and sleep disorders. The presence of sensitive eyes, eyelids, or skin and their triggering factors were assessed with specific questions. RESULTS: A total of 10,743 individuals (5,285 men and 5,458 women) were included in the study. Among them, 48.2% reported having sensitive skin and 46.0% reported having sensitive eyes. Sensitive eyes were more frequently reported by women (46.5%) than men (39.4%) in all countries, with the exception of China. The presence of sensitive eyes was more frequent if skin was very sensitive. More than half of subjects with sensitive eyes declared that their triggering factors were exposure to sunlight, dust, touch pad screens, or computer screens or dry air. They were more exposed to pollution and tobacco. Their phototype (including eye color) was lighter. DISCUSSION/CONCLUSION: This large study shows that self-declared sensitive eyes are very frequent and commonly associated with sensitive skin. Triggering factors of sensitive eyes are more specific.

Automatic Screening for Ocular Anomalies Using Fundus Photographs.

SIGNIFICANCE: Screening for ocular anomalies using fundus photography is key to prevent vision impairment and blindness. With the growing and aging population, automated algorithms that can triage fundus photographs and provide instant referral decisions are relevant to scale-up screening and face the shortage of ophthalmic expertise. PURPOSE: This study aimed to develop a deep learning algorithm that detects any ocular anomaly in fundus photographs and to evaluate this algorithm for "normal versus anomalous" eye examination classification in the diabetic and general populations. METHODS: The deep learning algorithm was developed and evaluated in two populations: the diabetic and general populations. Our patient cohorts consist of 37,129 diabetic patients from the OPHDIAT diabetic retinopathy screening network in Paris, France, and 7356 general patients from the OphtaMaine private screening network, in Le Mans, France. Each data set was divided into a development subset and a test subset of more than 4000 examinations each. For ophthalmologist/algorithm comparison, a subset of 2014 examinations from the OphtaMaine test subset was labeled by a second ophthalmologist. First, the algorithm was trained on the OPHDIAT development subset. Then, it was fine-tuned on the OphtaMaine development subset. RESULTS: On the OPHDIAT test subset, the area under the receiver operating characteristic curve for normal versus anomalous classification was 0.9592. On the OphtaMaine test subset, the area under the receiver operating characteristic curve was 0.8347 before fine-tuning and 0.9108 after fine-tuning. On the ophthalmologist/algorithm comparison subset, the second ophthalmologist achieved a specificity of 0.8648 and a sensitivity of 0.6682. For the same specificity, the fine-tuned algorithm achieved a sensitivity of 0.8248. CONCLUSIONS: The proposed algorithm compares favorably with human performance for normal versus anomalous eye examination classification using fundus photography. Artificial intelligence, which previously targeted a few retinal pathologies, can be used to screen for ocular anomalies comprehensively.

ExplAIn: Explanatory artificial intelligence for diabetic retinopathy diagnosis.

In recent years, Artificial Intelligence (AI) has proven its relevance for medical decision support. However, the "black-box" nature of successful AI algorithms still holds back their wide-spread deployment. In this paper, we describe an eXplanatory Artificial Intelligence (XAI) that reaches the same level of performance as black-box AI, for the task of classifying Diabetic Retinopathy (DR) severity using Color Fundus Photography (CFP). This algorithm, called ExplAIn, learns to segment and categorize lesions in images; the final image-level classification directly derives from these multivariate lesion segmentations. The novelty of this explanatory framework is that it is trained from end to end, with image supervision only, just like black-box AI algorithms: the concepts of lesions and lesion categories emerge by themselves. For improved lesion localization, foreground/background separation is trained through self-supervision, in such a way that occluding foreground pixels transforms the input image into a healthy-looking image. The advantage of such an architecture is that automatic diagnoses can be explained simply by an image and/or a few sentences. ExplAIn is evaluated at the image level and at the pixel level on various CFP image datasets. We expect this new framework, which jointly offers high classification performance and explainability, to facilitate AI deployment.

Towards improved breast mass detection using dual-view mammogram matching.

Breast cancer screening benefits from the visual analysis of multiple views of routine mammograms. As for clinical practice, computer-aided diagnosis (CAD) systems could be enhanced by integrating multi-view information. In this work, we propose a new multi-tasking framework that combines craniocaudal (CC) and mediolateral-oblique (MLO) mammograms for automatic breast mass detection. Rather than addressing mass recognition only, we exploit multi-tasking properties of deep networks to jointly learn mass matching and classification, towards better detection performance. Specifically, we propose a unified Siamese network that combines patch-level mass/non-mass classification and dual-view mass matching to take full advantage of multi-view information. This model is exploited in a full image detection pipeline based on You-Only-Look-Once (YOLO) region proposals. We carry out exhaustive experiments to highlight the contribution of dual-view matching for both patch-level classification and examination-level detection scenarios. Results demonstrate that mass matching highly improves the full-pipeline detection performance by outperforming conventional single-task schemes with 94.78% as Area Under the Curve (AUC) score and a classification accuracy of 0.8791. Interestingly, mass classification also improves the performance of mass matching, which proves the complementarity of both tasks. Our method further guides clinicians by providing accurate dual-view mass correspondences, which suggests that it could act as a relevant second opinion for mammogram interpretation and breast cancer diagnosis.

Intravenous high-dose methotrexate based systemic therapy in the treatment of isolated primary vitreoretinal lymphoma: An LOC network study.

The treatment of primary vitreoretinal lymphoma (PVRL) remains controversial regarding the use of local, systemic, or combined treatments. The aim of this study was to analyze the efficacy and toxicity of intravenous high-dose methotrexate (IV HD-MTX) based systemic therapy in a uniformly treated population of PVRL patients. From a nationwide French database, we retrospectively selected 59 patients (median age: 70 years, median Karnofsky Performance Status: 90%) with isolated PVRL at diagnosis who received first-line treatment with HD-MTX between 2011 and 2018. 8/59 patients also received a local treatment. No deaths or premature discontinuations of MTX due to toxicity were reported. A complete response was obtained in 40/57 patients after chemotherapy. Before treatment, IL-10 was elevated in the aqueous humor (AH) or in the vitreous in 89% of patients. After treatment, AH IL-10 was undetectable in 87% of patients with a CR/uCR/PR and detectable in 92% of patients with PD/SD. After a median follow-up of 61 months, 42/59 (71%) patients had relapsed, including 29 isolated ocular relapses as the first relapse and a total of 22 brain relapses. The median overall survival, progression-free survival, ocular-free survival and brain-free survival were 75, 18, 29 and 73 months, respectively. IV HD-MTX based systemic therapy as a first-line treatment for isolated PVRL is feasible, with acceptable toxicity, even in an elderly population. This strategy seems efficient to prevent brain relapse with prolonged overall survival. However, the ocular relapse rate remains high. New approaches are needed to improve local control of this disease, and ocular assessment could be completed by monitoring AH IL-10.

Phase IV clinical study to evaluate the effects of an intracameral combined mydriatic and anesthetic agent and standard topical mydriatics and anesthetics on the ocular surface after cataract surgery.

PURPOSE: To compare intracameral and topical mydriatics and anesthetics in cataract surgery. SETTING: Institute of Ocular Microsurgery, Barcelona, Spain. DESIGN: Phase IV, open-label, randomized, single-center study. METHODS: Patients were randomized in a 1:1 ratio to receive intracameral mydriatic-anesthetic (Mydrane/Fydrane) and anesthetic eyedrops or control (topical eyedrops only). The other treatment was administered for the second cataract surgery. Assessments were performed at presurgery and immediately postsurgery, at 12 to 36 hours postsurgery (day 1), and 7 days postsurgery. The primary endpoint was the change from baseline in corneal/conjunctival surface staining. The secondary endpoints included assessments of epithelial alterations, point-spread function, ocular surface disease index, conjunctival hyperemia, vision breakup time, ocular symptoms/signs, adverse events (AEs), corrected distance visual acuity, intraocular pressure, patient/investigator satisfaction, and procedure time. RESULTS: A total of 50 patients undergoing sequential cataract surgery in both eyes were included. Baseline assessments were similar in each group. The difference between Fydrane and control groups for the change from baseline at day 1 in corneal and conjunctival surface staining was not statistically significant. For Fydrane, postoperative epithelial alterations were fewer at day 1 (P < .005), folliculopapillary reaction was less frequent (P < .05), some ocular symptoms were less frequent and milder (P < .05), length of procedure was shorter (P < .001), and patient and investigator satisfaction were better (P < .05). There were few AEs in both groups. CONCLUSIONS: Fydrane reduced ocular surface damage by decreasing corneal epithelial and conjunctival toxicity with faster recovery of surface integrity compared with topical eyedrops, improved patient and investigator satisfaction, and reduced procedure time.