New grading criterion for retinal haemorrhages in term newborns based on deep convolutional neural networks.

BACKGROUND: To define a new quantitative grading criterion for retinal haemorrhages in term newborns based on the segmentation results of a deep convolutional neural network. METHODS: We constructed a dataset of 1543 retina images acquired from 847 term newborns, and developed a deep convolutional neural network to segment retinal haemorrhages, blood vessels and optic discs and locate the macular region. Based on the ratio of areas of retinal haemorrhage to optic disc, and the location of retinal haemorrhages relative to the macular region, we defined a new criterion to grade the degree of retinal haemorrhages in term newborns. RESULTS: The F1 scores of the proposed network for segmenting retinal haemorrhages, blood vessels and optic discs were 0.84, 0.73 and 0.94, respectively. Compared with two commonly used retinal haemorrhage grading criteria, this new method is more accurate, objective and quantitative, with the relative location of the retinal haemorrhages to the macula as an important factor. CONCLUSIONS: Based on a deep convolutional neural network, we can segment retinal haemorrhages, blood vessels and optic disc with high accuracy. The proposed grading criterion considers not only the area of the haemorrhages but also the locations relative to the macular region. It provides a more objective and comprehensive evaluation criterion. The developed deep convolutional neural network offers an end-to-end solution that can assist doctors to grade retinal haemorrhages in term newborns.

The natural history of retinal hemorrhage in pediatric head trauma.

BACKGROUND: Ophthalmologists are commonly asked to interpret appearance of retinal hemorrhages (RH) in children with suspected traumatic head injury. We sought to determine the natural history of RH in young children with head trauma and to identify patterns suggestive of chronicity in order to help establish timing of suspected traumatic injury. METHODS: The medical records of children <2 years of age with abusive or accidental head trauma and RH on initial fundus examination who had 1 or more follow-up examination were retrospectively reviewed. Types of RH (intraretinal, preretinal) were noted; intraretinal hemorrhage (IRH) severity was graded as mild (0-10), moderate (10-20), or severe (>20, too numerous to count [TNTC]). RESULTS: A total of 91 eyes of 52 children were studied. All eyes had IRH (62 eyes with TNTC). In all but one eye, IRH resolved to none or mild within 1-2 weeks. TNTC IRH did not persist beyond a few days. The longest an isolated IRH persisted was 32 days. Preretinal hemorrhage (PRH) was present in 68 eyes, persisting 5-111 days. On initial examination, 25% of eyes had only IRH, 75% both PRH and IRH; no eyes had only PRH. At 2 weeks, 3% had only IRH, 18% both, and 45% only PRH. In no eyes did RH worsen. CONCLUSIONS: IRH clears rapidly, whereas PRH may persist for many weeks. The presence of TNTC IRHs indicates that trauma occurred within a few days prior to examination, whereas the presence of PRH with no or few IRHs suggests days to weeks since trauma. To accurately identify these patterns, eye examinations should be completed as soon as possible after admission, preferably within 24-48 hours.

MR detection of retinal hemorrhages: correlation with graded ophthalmologic exam.

BACKGROUND: Dilated fundoscopic exam is considered the gold standard for detecting retinal hemorrhage, but expertise in obtaining this exam is not always immediately available. MRI can detect retinal hemorrhages, but correlation of the grade or severity of retinal hemorrhage on dilated fundoscopic exam with retinal hemorrhage visibility on MRI has not been described. OBJECTIVE: To determine the value of standard brain protocol MRI in detecting retinal hemorrhage and to determine whether there is any correlation with MR detection of retinal hemorrhage and the dilated fundoscopic exam grade of hemorrhage. MATERIALS AND METHODS: We conducted a retrospective chart review of 77 children <2 years old who were seen for head trauma from April 2007 to July 2013 and had both brain MRI and dilated fundoscopic exam or retinal camera images. A staff pediatric radiologist and radiology resident reviewed the MR images. Retinal hemorrhages were graded by a chief ophthalmology resident on a 12-point scale based on the retinal hemorrhage type, size, location and extent as seen on review of retinal camera images and detailed reports by ophthalmologists. Higher scores indicated increased severity of retinal hemorrhages. RESULTS: There was a statistically significant difference in the median grade of retinal hemorrhage examination between children who had retinal hemorrhage detected on MRI and children who did not have retinal hemorrhage detected on MRI (P = 0.02). When examination grade was categorized as low-grade (1-4), moderate-grade (5-8) or high-grade (>8) hemorrhage, there was a statistically significant association between exam grade and diagnosis based on MRI (P = 0.008). For example, only 14% of children with low-grade retinal hemorrhages were identified on MRI compared to 76% of children with high-grade hemorrhages. MR detection of retinal hemorrhage demonstrated a sensitivity of 61%, specificity of 100%, positive predictive value of 100% and negative predictive value of 63%. Retinal hemorrhage was best seen on the gradient recalled echo (GRE) sequences. CONCLUSION: MRI using routine brain protocol demonstrated 61% sensitivity and 100% specificity in detecting retinal hemorrhage. High-grade hemorrhage was more often detected on MRI than low-grade hemorrhage, 76% vs. 14%. GRE images were the most sensitive for detection of retinal hemorrhages. A dilated fundoscopic exam can be difficult to obtain in infancy, especially in critically ill or non-sedated children. MRI is a useful modality for added documentation of retinal hemorrhage and can be used as an alternative exam when ophthalmologic expertise or retinal camera images are unavailable. Additionally, identification of retinal hemorrhage on MRI can raise the possibility of abuse in children presenting with nonspecific findings.

Detection and Differentiation of Intraretinal Hemorrhage in Spectral Domain Optical Coherence Tomography.

PURPOSE: The purpose of this study was to classify and detect intraretinal hemorrhage (IRH) in spectral domain optical coherence tomography (SD-OCT). METHODS: Initially the presentation of IRH in BRVO-patients in SD-OCT was described by one reader comparing color-fundus (CF) and SD-OCT using dedicated software. Based on these established characteristics, the presence and the severity of IRH in SD-OCT and CF were assessed by two other masked readers and the inter-device and the inter-observer agreement were evaluated. Further the area of IRH was compared. RESULTS: About 895 single B-scans of 24 eyes were analyzed. About 61% of SD-OCT scans and 46% of the CF-images were graded for the presence of IRH (concordance: 73%, inter-device agreement: k = 0.5). However, subdivided into previously established severity levels of dense (CF: 21.3% versus SD-OCT: 34.7%, k = 0.2), flame-like (CF: 15.5% versus SD-OCT: 45.5%, k = 0.3), and dot-like (CF: 32% versus SD-OCT: 24.4%, k = 0.2) IRH, the inter-device agreement was weak. The inter-observer agreement was strong with k = 0.9 for SD-OCT and k = 0.8 for CF. The mean area of IRH detected on SD-OCT was significantly greater than on CF (SD-OCT: 11.5 ± 4.3 mm(2) versus CF: 8.1 ± 5.5 mm(2), p = 0.008). CONCLUSIONS: IRH seems to be detectable on SD-OCT; however, the previously established severity grading agreed weakly with that assessed by CF.

Grading system for retinal hemorrhages in abusive head trauma: clinical description and reliability study.

PURPOSE: There is currently no universally accepted grading system for describing retinal hemorrhages (RH) in abusive head trauma (AHT). The purpose of this study was to devise and evaluate a novel grading system and descriptive nomenclature for RH in AHT for clinical and research purposes. METHODS: A traumatic hemorrhagic retinopathy (THR) grading system was developed for assessing and quantitatively analyzing retinal findings in abusive head trauma. The criteria for the THR grade included the extent, spread, and morphology of RH. Extent was classified as region 1 (posterior pole) or region 2 (peripheral). Spread, based on number of retinal hemorrhages, was classified as mild (10 or fewer RH), moderate (more than 10 RH) and severe (more than half of involved regions covered by RH). Morphology was classified by its intraretinal or extraretinal involvement. Two independent graders calculated the THR grade from RetCam images of 38 eyes of 19 patients <3 years of age with retinal hemorrhages associated with head injury. Grading was performed on two separate occasions. Intra- and interobserver reliability was assessed with Spearman correlation coefficient (r) and intraclass correlation coefficient (ICC). RESULTS: There was a high level of intraobserver agreement across both assessments (97% agreement [Spearman r = 0.997; P < 0.0001] and 100% agreement [Spearman r = 1.0; P < 0.0000]). Intraclass correlation (ICC, 0.995; 95% CI, 0.991-0.997; P < 0.0001) confirmed a very high level of agreement overall. CONCLUSIONS: The traumatic hemorrhagic retinopathy grading system demonstrated excellent intraobserver and interobserver reliability. The nomenclature is easily understood and may be useful in medical records and medicolegal reports.

Retinal haemorrhage description tool.

BACKGROUND: Retinal haemorrhages are an important finding in children with abusive and accidental head trauma. There are no standardised and validated protocols to describe them in a consistent manner. The aim of this web-based study was to establish the reliability and validity of a new retinal haemorrhage description tool. MATERIALS AND METHOD: Our tool is a comprehensive questionnaire, which is applied using a retinal schematic that divides the retina into four independent zones. Four independent observers scored retinal haemorrhages from 80 retinal photographs. Inter- and intra-rater agreement (by repeat assessment of 10 photographs for each examiner) were calculated using Fleiss κ statistics. RESULTS: A high inter-rater agreement was noted for haemorrhages in the peripapillary zones, whereas agreement was only fair for all other zones. Intra-rater agreement was high only for the posterior pole. Photographs may be an unreliable way of documenting retinal haemorrhages particularly from the peripheral retina, thus underscoring the importance of a thorough clinical examination. CONCLUSION: This study shows that the tool achieves some validity for describing haemorrhages in the posterior retina. It performs less well in the peripheral zones.

SD-OCT to differentiate traumatic submacular hemorrhage types using automatic three-dimensional segmentation analysis.

Traumatic submacular hemorrhage may present with significant decrease in vision and may have varying outcomes. Following injury, the hemorrhage can collect either between the neurosensory retina and retinal pigment epithelium (RPE) or below the RPE. This differentiation may be important to prognosticate and to guide treatment. In two patients with post-traumatic submacular hemorrhage, Cirrius spectral domain high-definition optical coherence tomography (OCT) (Carl Zeiss Meditec, Dublin, CA) was used to differentiate traumatic submacular hemorrhage types using automation three-dimensional segmentation analysis. Based on the OCT findings, the patient with sub-RPE bleed was subjected to pneumatic displacement. En face C-scan imaging just below the RPE allowed for the diagnosis of the exact location of choroidal rupture that was masked due to hemorrhage.

An inter-observer and intra-observer study of a classification of RetCam images of retinal haemorrhages in children.

BACKGROUND: There is currently no universally accepted classification of childhood retinal haemorrhages. AIM: To measure the inter- and intra-observer agreement of clinical classifications of retinal haemorrhages in children. METHODS: Four examiners (two consultant ophthalmologists and two other clinicians) were shown 142 retinal haemorrhages on 31 RetCam photographs. The retinal haemorrhages were from children with accidental or abusive head injury, or other encephalopathies, and included retinal haemorrhages of different ages. Specified haemorrhages were initially classified by each examiner according to their clinical understanding. Altogether, 26 haemorrhages were re-presented to test intra-observer consistency. Examiners then agreed a common description for each haemorrhage type and five categories were described (vitreous, pre-retinal, nerve fibre layer, intra-retinal/sub-retinal or indeterminate) and the study repeated. RESULTS: There was 'fair agreement' initially (Fleiss' unweighted κ=0.219) and, with the agreed classification, slight improvement (0.356). Intra-observer agreement marginally improved on re-test. The two consultant ophthalmologists showed 'fair' agreement on both occasions (paired κ statistic). The other rater pair improved from 'fair' to 'substantial' agreement with the new classification. CONCLUSIONS: The classification of retinal haemorrhage in children by appearance alone shows only fair agreement between examiners. Clinicians who are not consultant ophthalmologists appear to benefit from the new succinct classification.

An interrater reliability study of a new 'zonal' classification for reporting the location of retinal haemorrhages in childhood for clinical, legal and research purposes.

BACKGROUND/AIMS: To develop and assess a zonal classification of the retina to facilitate description of the location of retinal haemorrhages in children. METHODS: A novel zonal classification of the retina was devised based on the anatomical landmarks of the optic disc and vascular arcades, by reviewing a large number of wide field digital retinal images drawn from our database of children with accidental and non-accidental head injury and other encepthalopathies. Four expert examiners then independently 'located' 142 retinal haemorrhages by zone, from 31 high quality photographs. RESULTS: Cohen's unweighted kappa scores for all possible pairs of the four raters (ie, six pairs) ranged from 0.86 to 0.92, that is 'almost perfect' agreement. Fleiss' kappa for agreement between multiple raters (four) and for multiple categories (three) was 0.8841, that is 'almost perfect' agreement. Cohen's unweighted kappa statistic for intrarater reliability gave an overall concordance that ranged from 'substantial' to 'perfect' agreement. CONCLUSION: This new retinal zone classification and the use of photographs and templates is a very reliable tool for reporting the location of retinal haemorrhages from multiple aetiologies in children, and may be useful for research and medico-legal reports.

Retinal hemorrhages: advances in understanding.

Retinal hemorrhage is a cardinal manifestation of abusive head injury characterized by repetitive acceleration-deceleration with or without blunt head impact. Detailed description of the hemorrhages and documentation are critical to diagnosis. Vitreoretinal traction appears to be the major causative factor. Outcome is largely dependent on brain and optic nerve injury.

Ocular findings in patients with Crimean-Congo hemorrhagic fever.

PURPOSE: To describe ocular findings of patients with Crimean-Congo hemorrhagic fever (CCHF). DESIGN: Prospective, interventional, consecutive case series. METHODS: This study was conducted in Sivas, a city located in the central Anatolia, between July 1 and August 31, 2007. Confirmed CCHF patients were enrolled in the study and underwent ocular examination during hospitalization. RESULTS: Nineteen confirmed CCHF patients were included in this study. All patients were classified into 2 groups in terms of disease severity (severe vs nonsevere), according to the Swanepoel predictive criteria. Fourteen patients (73.7%) were classified as "nonsevere," and the remaining 5 patients (26.3%) were classified as "severe" in this study. One patient having severe disease died. Ocular findings were present in 14 patients (73.7%) and none of the patients presented any visual complaints. Ocular examination revealed that 7 patients (36.8%) had only bilateral multiple subconjunctival hemorrhage, and 2 patients (10.5%) (1 bilateral, 1 unilateral) had retinal hemorrhage. Five patients (26.3%) had subconjunctival hemorrhage and retinal hemorrhage. Follow-up examination 1 month later showed complete resorption of the subconjunctival hemorrhage and retinal hemorrhage. There was a statistically significant difference between patients with ocular findings and patients without ocular findings for prothrombin time (P = .011). There was no evidence of uveitis, retinal edema, sheathing of retinal vessels, or intravitreal hemorrhage in our patients. CONCLUSION: From the small sample study, CCHF caused a mild form of ocular disease. CCHF must be considered when subconjunctival or superficial retinal hemorrhages are seen in association with fever in endemic areas.

Retinal hemorrhage and pediatric brain injury: etiology and review of the literature.

Retinal hemorrhages have long been linked with child abuse and, in particular, the "shaken baby/shaking-impact" syndrome. However, the presence of retinal hemorrhages is neither necessary nor sufficient for the diagnosis of child abuse. Additionally, retinal hemorrhages are also associated with an ever-expanding list of conditions, each of which carries important implications for patients and their families. To correctly interpret a patient's retinal hemorrhages, the physician requires a broad knowledge base, including of child abuse, the "shaken baby/shaking-impact" syndrome, the differential diagnosis of retinal hemorrhages and the types of retinal hemorrhage and their diagnostic implications. We review the literature regarding types of retinal hemorrhage and their associated etiologies.

A new staging system for idiopathic retinal periphlebitis.

PURPOSE: To develop and standardize a universally acceptable new staging system for idiopathic retinal periphlebitis (Eales disease). METHODS: A new staging system was established and standardized based on standard terminology and features. Idiopathic retinal periphlebitis was classified as peripheral and central types. Peripheral disease consisted of four stages. Stage 1 is periphlebitis of small (1a) and large (1b) caliber vessels with superficial retinal hemorrhages. Stage 2a denotes capillary nonperfusion and 2b neovascularization elsewhere/of the disc. Stage 3a is classified as fibrovascular proliferation and 3b vitreous hemorrhage. Stage 4a is traction/combined rhegmatogenous retinal detachment whereas 4b is rubeosis iridis, neovascular glaucoma, complicated cataract, and optic atrophy. A total of 253 cases of idiopathic retinal periphlebitis (mean age, 24.7 +/- 4.7 years, all male) presenting at this tertiary care center were classified prospectively according to the new staging system, by two independent observers (interobserver correlation = 0.7). RESULTS: The new staging system was consistent, simple, and easy to recall. Peripheral and central types of idiopathic retinal periphlebitis were found in 94.07% and 5.93% of cases, respectively. The new staging system also defined the severity of the disease. Vitreous hemorrhage was found to be the commonest presenting feature (51.68%), whereas traction/combined rhegmatogenous detachment was found in 5.88% of cases. CONCLUSIONS: The new staging system is useful in classifying and assessing the severity of disease. Management strategy can also be defined according to the stage of the disease. It is designed to promote the use of standard assessment with applications to clinical management and research.

Síndrome del bebé sacudido (SBS) / Syndrome of drinks shaken

El Síndrome del Bebé Sacudido (SBS) es una patología que plantea dificultades para su diagnóstico correcto y seguro. El subdiagnóstico y el subregistro son habituales. Presentamos dos casos de SBS en su forma clínica clásica que tuvieron los signos característicos: lactantes sanos hasta el momento de la consulta, que bruscamente presentaron un cuadro neurológico agudo, severo y polimorfo asociado a hemorragias intracraneales e intraoculares. Se recomiendan los procedimientos y estudios para arribar al diagnóstico en forma concluyente y aplicar un tratamiento adecuado

High-altitude retinopathy and altitude illness.

OBJECTIVE: To determine the relationship between high-altitude retinopathy (HAR) and other altitude-related illnesses and establish a classification system for HAR. DESIGN: Observational case series. PARTICIPANTS: All 40 climbers among 3 Himalayan expeditions who ascended to altitudes between 16,000 and 29,028 feet above sea level (summit of Mt. Everest) were examined for signs of HAR and altitude illness (AI). METHODS: All subjects had dilated fundus examinations before the ascent, intermittent fundus, and medical examinations during the climb and a dilated fundus and medical examination within 2 days after attaining their highest altitude. MAIN OUTCOME MEASURES: Careful fundus drawings or fundus photography or both were obtained for all participants. All subjects gave a subjective assessment of their symptoms of acute mountain sickness (AMS) and were assessed clinically for signs of high-altitude cerebral edema (HACE). RESULTS: Nineteen of 21 climbers who ascended above 25,000 feet developed HAR. Fourteen of 19 climbers who attained altitudes between 16,000 and 25,000 feet were found to have retinopathy. A grading system for HAR describing the severity of the retinopathy was developed. Correlation of the retinopathy with other AI showed that AMS was endemic and that a statistically significant correlation exists between HAR and HACE (P = 0.0240). CONCLUSION: Recognizing advancing grades of HAR may allow physicians to recommend initiating empiric treatment with oxygen, steroids, diuretics and immediate descent to prevent HAR progression, macular involvement, or potentially fatal HACE. High-altitude retinopathy is both a significant component of and a predictor of progressive AI.

Classification of retinal damage by a neural network based system.

The objective of this research is to provide an ophthalmologist with a helpful system, capable of classifying a degree of patients' retinal hemorrhage. The system is composed of four modules: (a) data acquisition module, (b) image Database module, (c) image processing module, (d) image classification module. The system was trained with a modular neural network on a set of 25 images, and tested on a set of 160 images. A training performance of greater than 95% was achieved. The classifying part of the system showed 79% recognition accuracy. Since the testing images were taken from independent sources, we assume that the system should also provide an accurate classification of other image types.

[Retinopathy of prematurity and immaturity . A clinical study]. / Retinopatia prematurului si dismaturului. Studiu clinic.

The retinopathy of prematurity is the most important cause of blindness at children and one of the most frequent cause of blindness in the postnatal period. 435 newborn prematures and dismatures, were examined in the first two weeks since they were born. The ophthalmological exam showed vasoconstriction of retinal vessels. The treatment consisted in cryotHerapy with good results. The most up-to-date method is the therapy with argon laser mounted on the indirect ophthalmoscope.

Retinal hemorrhage in the young child: a review of etiology, predisposed conditions, and clinical implications.

Retinal hemorrhage is a frequent finding in child abuse, but may also be associated with cardiopulmonary resuscitation, accidental trauma, and a variety of illnesses such as blood dyscrasias and infections. Although it is imperative that child abuse be considered in all children who present with retinal hemorrhages, whether retinal hemorrhages can be attributed to cardiopulmonary resuscitation in suspected cases of abuse poses a dilemma. The etiologies of retinal hemorrhage as well as the literature presently available to support or refute the various diagnoses are discussed. Guidelines for funduscopic examination in the Emergency Department as well as a clinical classification of retinal hemorrhage are provided. In addition, guidelines are suggested for the appropriate clinical investigations in children with retinal hemorrhages.