Why Science? | Singapore Eye Research Institute

The cornea and lens focus light onto the retina, which transforms it into neural signals. Tear film, pupil, ocular shape, retinal cells and brain processing all influence measured vision.

Vision research connects optics, neuroscience, imaging, epidemiology, engineering and clinical care. A useful finding must remain valid across instruments, populations, disease stages and real patient workflows. The practical capability is vision-science literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “The eye is an optical and biological system” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A refractive measurement does not describe every cause of visual difficulty.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Next chapter

Letter charts estimate spatial resolution under specified distance, contrast and correction. Contrast sensitivity, fields, colour, depth and real-world tasks can reveal other limitations.

Eye studies often follow people over time because prevalence, onset, progression and treatment response answer different questions. Cohort design and retention affect which conclusions the data can support. The practical capability is functional measurement. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Visual acuity is one outcome, not all of vision” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A single acuity score does not capture every visual function.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Myopia occurs when optical focus and eye geometry are mismatched, commonly with excessive axial length. Studies examine genetics, environment, progression, imaging and interventions across childhood and adulthood.

Images become evidence through acquisition standards, grading rules, quality checks and clinical reference labels. Artificial intelligence can assist analysis only within the population and task for which it has been evaluated. The practical capability is myopia literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Myopia research follows eye growth and risk” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Population findings do not determine an individual treatment plan.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Following participants over time can estimate incidence and progression rather than only prevalence. Visit timing, measurement consistency, loss to follow-up and changing exposures affect conclusions.

Translation in ophthalmology is staged. Mechanism, prototype performance, safety, clinical validity, usefulness and adoption each require different evidence and oversight before a tool changes care. The practical capability is longitudinal reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Longitudinal studies separate age from change” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A cross-sectional age difference is not automatically an ageing trajectory.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

SERI’s official SEED programme combines large studies of major eye diseases and risk factors in Asian populations. Sampling, eligibility and participation determine which wider population the results can represent.

Vision research connects optics, neuroscience, imaging, epidemiology, engineering and clinical care. A useful finding must remain valid across instruments, populations, disease stages and real patient workflows. The practical capability is epidemiology literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Population cohorts make denominators visible” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A clinic sample may not estimate community prevalence.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Fundus photography, optical coherence tomography and other modalities reveal different tissues and depths. Focus, segmentation, device settings and grader rules influence measurements and comparisons.

Eye studies often follow people over time because prevalence, onset, progression and treatment response answer different questions. Cohort design and retention affect which conclusions the data can support. The practical capability is ophthalmic-imaging literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Eye imaging turns structure into quantitative data” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A high-resolution image is not a diagnosis without clinical interpretation.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Optical coherence tomography reconstructs depth-resolved retinal or anterior-segment structure from interferometric signals. Motion, media opacity and segmentation errors can create artefacts.

Images become evidence through acquisition standards, grading rules, quality checks and clinical reference labels. Artificial intelligence can assist analysis only within the population and task for which it has been evaluated. The practical capability is OCT reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “OCT uses reflected light to map tissue layers” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: An automated thickness value should be checked against image quality and anatomy.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

An eye-image model may detect, segment, classify or predict. Developers must define the target condition, reference standard, population, threshold and workflow, then test calibration and external performance.

Translation in ophthalmology is staged. Mechanism, prototype performance, safety, clinical validity, usefulness and adoption each require different evidence and oversight before a tool changes care. The practical capability is clinical-AI literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Artificial intelligence needs task-specific evaluation” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Strong performance in one dataset does not guarantee safe use elsewhere.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Sensitivity tracks detected cases while specificity tracks correctly excluded non-cases. Predictive values also depend on prevalence, and thresholds trade missed cases against false alarms.

Vision research connects optics, neuroscience, imaging, epidemiology, engineering and clinical care. A useful finding must remain valid across instruments, populations, disease stages and real patient workflows. The practical capability is diagnostic-test reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Sensitivity and specificity describe different errors” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: One accuracy percentage can hide clinically important error patterns.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Glaucoma can involve optic-nerve damage and visual-field loss, while pressure is one important risk factor rather than the disease itself. Imaging, fields, anatomy and longitudinal change contribute to assessment.

Eye studies often follow people over time because prevalence, onset, progression and treatment response answer different questions. Cohort design and retention affect which conclusions the data can support. The practical capability is glaucoma literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Glaucoma research joins structure, pressure and function” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: One pressure reading cannot confirm or exclude glaucoma.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Retinal imaging and functional tests support research on diabetic retinopathy, age-related macular degeneration and other conditions. Disease stage, systemic factors and imaging protocol shape interpretation.

Images become evidence through acquisition standards, grading rules, quality checks and clinical reference labels. Artificial intelligence can assist analysis only within the population and task for which it has been evaluated. The practical capability is retinal-science reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Retina research studies neural and vascular tissue” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A retinal image can suggest patterns but does not replace complete clinical assessment.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Transparency, curvature, nerves, epithelium, tear film and immune responses affect comfort and vision. Laboratory, imaging and clinical outcomes may move differently and need separate measures.

Translation in ophthalmology is staged. Mechanism, prototype performance, safety, clinical validity, usefulness and adoption each require different evidence and oversight before a tool changes care. The practical capability is corneal-science literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Cornea and ocular surface research spans optics and biology” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Symptom severity may not match one surface test.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Drops, injections, implants and other approaches face barriers of penetration, residence time, dosing and toxicity. Preclinical evidence, formulation and staged trials address different risks.

Vision research connects optics, neuroscience, imaging, epidemiology, engineering and clinical care. A useful finding must remain valid across instruments, populations, disease stages and real patient workflows. The practical capability is ocular-therapeutics reasoning. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Therapeutics and drug delivery must reach the target safely” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A biological mechanism does not establish a safe or effective treatment.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Imaging hardware, sensors, surgical tools and assistive systems begin with a defined user and workflow. Bench tests, human factors, reliability and clinical evaluation build evidence in stages.

Eye studies often follow people over time because prevalence, onset, progression and treatment response answer different questions. Cohort design and retention affect which conclusions the data can support. The practical capability is ophthalmic-engineering literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Bioengineering converts needs into testable devices” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: A working prototype is not an approved medical device.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Protocols define eligibility, intervention, comparator, outcomes, monitoring and analysis. Randomisation, masking and preregistration can reduce bias, while ethics review and consent protect participants.

Images become evidence through acquisition standards, grading rules, quality checks and clinical reference labels. Artificial intelligence can assist analysis only within the population and task for which it has been evaluated. The practical capability is clinical-trial literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Clinical trials protect participants and comparisons” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Trial participation and treatment decisions require qualified clinical guidance.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Eye images, genetics and health records can be sensitive. Consent, authorised access, de-identification, security, provenance and retention rules support both participant rights and reproducible research.

Translation in ophthalmology is staged. Mechanism, prototype performance, safety, clinical validity, usefulness and adoption each require different evidence and oversight before a tool changes care. The practical capability is research-ethics literacy. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Data governance protects people and validity” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Removing a name may not eliminate re-identification risk.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter · Next chapter

Archive protocol, consent and approvals, cohort, device and software version, calibration, image quality, reference grading, outcomes, exclusions, subgroup results, uncertainty and decision stage. Students can learn from public evidence while clinicians manage care.

Vision research connects optics, neuroscience, imaging, epidemiology, engineering and clinical care. A useful finding must remain valid across instruments, populations, disease stages and real patient workflows. The practical capability is vision-science career preparation. Keep the question, method, controls, observation, uncertainty, review and next decision in one evidence receipt. That record makes progress visible without enlarging the claim beyond the stage reached. SERI, SNEC and study teams control current research, recruitment and released findings; clinicians and health authorities control diagnosis, treatment, trials and patient-care decisions.

Put this into practice by turning the idea into a small, reviewable action: Build a dated evidence card for “Keep a vision-research evidence receipt” with the question, method, observation, limitation and next test. Then ask what the resulting note or artefact actually demonstrates. A useful record names context, source or instruction, action, observation, interpretation and uncertainty. It also preserves this limit: Only current clinicians, institutions and regulators can diagnose, prescribe, recruit or authorise a product.

An evidence check keeps confidence proportional to what the work can show. Separate the original observation or official statement from the learner’s explanation. Record feedback and any revision. If information is missing, write “not shown” or “not yet known” rather than guessing. That honest gap can become the next question.

Contents · Previous chapter