An Artificial Intelligence Driven Approach for Classification of Ophthalmic Images using Convolutional Neural Network: An Experimental Study.

BACKGROUND: Early disease detection is emphasized within ophthalmology now more than ever, and as a result, clinicians and innovators turn to deep learning to expedite accurate diagnosis and mitigate treatment delay. Efforts concentrate on the creation of deep learning systems that analyze clinical image data to detect disease-specific features with maximum sensitivity. Moreover, these systems hold promise of early accurate diagnosis and treatment of patients with common progressive diseases. DenseNet, ResNet, and VGG-16 are among a few of the deep learning Convolutional Neural Network (CNN) algorithms that have been introduced and are being investigated for potential application within ophthalmology. METHODS: In this study, the authors sought to create and evaluate a novel ensembled deep learning CNN model that analyzes a dataset of shuffled retinal color fundus images (RCFIs) from eyes with various ocular disease features (cataract, glaucoma, diabetic retinopathy). Our aim was to determine (1) the relative performance of our finalized model in classifying RCFIs according to disease and (2) the diagnostic potential of the finalized model to serve as a screening test for specific diseases (cataract, glaucoma, diabetic retinopathy) upon presentation of RCFIs with diverse disease manifestations. RESULTS: We found adding convolutional layers to an existing VGG-16 model, which was named as a proposed model in this article that, resulted in significantly increased performance with 98% accuracy (p<0.05), including good diagnostic potential for binary disease detection in cataract, glaucoma, diabetic retinopathy. CONCLUSION: The proposed model was found to be suitable and accurate for a decision support system in Ophthalmology Clinical Framework.

A deep learning approach for classification of COVID and pneumonia using DenseNet-201.

In the present paper, our model consists of deep learning approach: DenseNet201 for detection of COVID and Pneumonia using the Chest X-ray Images. The model is a framework consisting of the modeling software which assists in Health Insurance Portability and Accountability Act Compliance which protects and secures the Protected Health Information . The need of the proposed framework in medical facilities shall give the feedback to the radiologist for detecting COVID and pneumonia though the transfer learning methods. A Graphical User Interface tool allows the technician to upload the chest X-ray Image. The software then uploads chest X-ray radiograph (CXR) to the developed detection model for the detection. Once the radiographs are processed, the radiologist shall receive the Classification of the disease which further aids them to verify the similar CXR Images and draw the conclusion. Our model consists of the dataset from Kaggle and if we observe the results, we get an accuracy of 99.1%, sensitivity of 98.5%, and specificity of 98.95%. The proposed Bio-Medical Innovation is a user-ready framework which assists the medical providers in providing the patients with the best-suited medication regimen by looking into the previous CXR Images and confirming the results. There is a motivation to design more such applications for Medical Image Analysis in the future to serve the community and improve the patient care.

Unusual Missed Diagnosis of Foreign Body: A Case Report.

BACKGROUND: Penetrating intraorbital foreign body (IOFB) is usually associated with high-velocity trauma forces around the eye. IOFB injury to globe or optic nerve is considered a surgical emergency; an immediate diagnosis and management plan is generally indicated. METHODS: A case report (design). The patient was a 78-year-old male presented with diminution of vision of the right eye following a high-velocity injury. The patient was noted to have a closed globe injury with associated retinal detachment and vitreous hemorrhage. An initial orbital CT scan did not reveal any IOFB, despite and intact globe. However, repeat a CT head and orbit scan revealed an intracranial magnetic foreign body lodged in the right frontal lobe. CONCLUSION: A CT scan of the brain and paranasal sinuses should be obtained along with a CT orbit in case of high-velocity orbital/ocular trauma.

Central Retinal Artery Occlusion After Nasosinal Surgery - an Insight.

OBJECTIVE: To describe a case of central retinal artery occlusion (CRAO) after nasosinal surgery and subject's subsequent response to hyperbaric oxygen therapy (HBOT). DESIGN: Observational case report. RESULTS: We describe a subject with diagnosed CRAO after septoplasty, bilateral inferior turbinate reduction and balloon sinuplasty, who was given hyperbaric oxygen treatment after four days of onset of CRAO with an improvement in visual acuity and visual field. CONCLUSION: Even though CRAO has been rarely reported after ENT procedures and HBOT has been previously described for the treatment, this is the case report where hyperbaric oxygen was given after four days of onset, with a possible improvement.

A neuronal network model with simplified tonotopicity for tinnitus generation and its relief by sound therapy.

Tinnitus is the perception of sound in the ears or in the head where no external source is present. Sound therapy is one of the most effective techniques for tinnitus treatment that have been proposed. In order to investigate mechanisms of tinnitus generation and the clinical effects of sound therapy, we have proposed conceptual and computational models with plasticity using a neural oscillator or a neuronal network model. In the present paper, we propose a neuronal network model with simplified tonotopicity of the auditory system as more detailed structure. In this model an integrate-and-fire neuron model is employed and homeostatic plasticity is incorporated. The computer simulation results show that the present model can show the generation of oscillation and its cessation by external input. It suggests that the present framework is promising as a modeling for the tinnitus generation and the effects of sound therapy.

Comparison of neuronal network models for tinnitus management by sound therapy.

Tinnitus is a condition in which sounds heard in the ear or head without any external sound. There are many therapeutic approaches for tinnitus and sound therapy is one of the techniques for its treatment that have been proposed. In order to investigate mechanisms of tinnitus generation and the clinical effects of sound therapy from the viewpoint of neural engineering, we have proposed computational models with plasticity and inhibitory feedback using a neural oscillator or model neurons described by simplified Hodgkin-Huxley equations. By hypothesizing that the oscillation and the non-oscillatory state in the models correspond to generation and inhibition of tinnitus, respectively, we found out that the models could explain the fact that the habituated human auditory system temporarily halts perception of tinnitus following sound therapy. However, a simpler model without inhibitory feedback can exhibit the solutions that exist in the former models. In the present paper, outcomes of the neuronal network model, which is incorporated with inhibitory feedback, are compared with the model without inhibitory feedback. It was revealed that the former is superior since it has a larger parameter region in which the effects of sound therapy can be restored due to synaptic plasticity.

Oscillation and its inhibition in a neuronal network model for tinnitus sound therapy.

Tinnitus is the perception of phantom sounds in the ears or in the head. Sound therapy techniques for tinnitus treatment have been proposed. In order to investigate mechanisms of tinnitus generation and the clinical effects of sound therapy from the viewpoint of neural engineering, we have proposed a computational model using a neural oscillator. In the present paper, we propose another model that is composed of model neurons described by simplified Hodgkin-Huxley equations. By computer simulation it was detected that this model also has a bistable state, i.e., a stable oscillatory state and a stable equilibrium (non-oscillatory) state coexist at a certain parameter region. It was also noticed that the oscillation can be inhibited by supplying constant or pulse train stimuli, which is hypothesized as an afferent signal that is employed as an acoustical signal for tinnitus treatment. By hypothesizing that the oscillation and the equilibrium correspond to generation and inhibition of tinnitus, respectively, these phenomena could explain the fact that the habituated human auditory system temporarily halts perception of tinnitus following sound therapy.

Dynamical properties of a plastic neural network model for tinnitus therapy and inhibition of oscillation using noise stimulus.

Tinnitus is the perception of phantom sounds in the ears or in the head. Sound therapy techniques for tinnitus have been proposed. To account for mechanisms of tinnitus generation and the clinical effects of sound therapies from the viewpoint of neural engineering, we have proposed a plastic neural network model for the human auditory system. We found that this model has a bistable state, i.e., a stable oscillatory state and a stable equilibrium (non-oscillatory) state coexist at a certain parameter region. We also found that the oscillation can be inhibited by supplying sinusoidal stimulus, which is hypothesized as sound for treatment of tinnitus, to the model. By hypothesizing that the oscillation and the equilibrium correspond to generation and inhibition of tinnitus, respectively, we reported that these phenomena could explain the fact that the habituated human auditory system temporarily halts perception of tinnitus following sound therapy. This paper describes dynamical properties of the model and inhibition of the oscillation for two kinds of noise stimuli which correspond to sound for treatment of tinnitus in clinical. Through numerical simulations we found that adequate noise stimulus can inhibits the oscillation.

Developing combinatorial multi-component therapies (CMCT) of drugs that are more specific and have fewer side effects than traditional one drug therapies.

Drugs designed for a specific target are always found to have multiple effects. Rather than hope that one bullet can be designed to hit only one target, nonlinear interactions across genomic and proteomic networks could be used to design Combinatorial Multi-Component Therapies (CMCT) that are more targeted with fewer side effects. We show here how computational approaches can be used to predict which combinations of drugs would produce the best effects. Using a nonlinear model of how the output effect depends on multiple input drugs, we show that an artificial neural network can accurately predict the effect of all 215 = 32,768 combinations of drug inputs using only the limited data of the output effect of the drugs presented one-at-a-time and pairs-at-a-time.

Oscillation and its inhibition in a neural oscillator model for tinnitus.

Tinnitus is a symptom of perceiving phantom sounds. As one of its treatment techniques, tinnitus retraining therapy (TRT) has been proposed. It consists of psychotherapy by counseling and physical therapy based on masking theory by external stimuli. Our interest is to explain medical effects of the physical therapy from the viewpoint of engineering. In this paper we proposed a neural oscillator model with plasticity as a model for the tinnitus generation in the auditory central nervous system and its treatment. We investigated not only oscillatory phenomena observed in the model but also inhibition of the oscillation by external stimulus.