Deep Learning Models Used in the Diagnostic Workup of Keratoconus: A Systematic Review and Exploratory Meta-Analysis.

PURPOSE: The prevalence of keratoconus in the general population is reported to be up to 1 of 84. Over the past 2 decades, diagnosis and management evolved rapidly, but keratoconus screening in clinical practice is still challenging and asks for improving the accuracy of keratoconus detection. Deep learning (DL) offers considerable promise for improving the accuracy and speed of medical imaging interpretation. We establish an inventory of studies conducted with DL algorithms that have attempted to diagnose keratoconus. METHODS: This systematic review was conducted according to the recommendations of the PRISMA statement. We searched (Pre-)MEDLINE, Embase, Science Citation Index, Conference Proceedings Citation Index, arXiv document server, and Google Scholar from inception to February 18, 2022. We included studies that evaluated the performance of DL algorithms in the diagnosis of keratoconus. The main outcome was diagnostic performance measured as sensitivity and specificity, and the methodological quality of the included studies was assessed using QUADAS-2. RESULTS: Searches retrieved 4100 nonduplicate records, and we included 19 studies in the qualitative synthesis and 10 studies in the exploratory meta-analysis. The overall study quality was limited because of poor reporting of patient selection and the use of inadequate reference standards. We found a pooled sensitivity of 97.5% (95% confidence interval, 93.6%-99.0%) and a pooled specificity of 97.2% (95% confidence interval, 85.7%-99.5%) for topography images as input. CONCLUSIONS: Our systematic review found that the overall diagnostic performance of DL models to detect keratoconus was good, but the methodological quality of included studies was modest.

The effect of a video tutorial to improve patients' keratoconus knowledge - a randomized controlled trial and meta-analysis of published reports.

Objective: To investigate whether a video tutorial, highlighting important aspects of keratoconus provided prior to a scheduled follow-up consultation, has a specific effect on patients' knowledge after the consultation. Methods and Analysis: Single center, randomized controlled trial registered on ISRCTN registry (number ISCTN75317089, https://doi.org/10.1186/ISRCTN75317089). Consenting eligible keratoconus patients were randomly assigned to either receive a conventional face-to-face consultation (control group) or to an additional video tutorial (interventional group) on definition, risk factors and treatment options provided prior to the consultation. The main outcome measure was the difference of knowledge assessed by a questionnaire after the consultation. Of each participant, clinical characteristics, highest educational level and medical background were obtained. We also performed a meta-analysis of published reports assessing knowledge improvement by video-based patient education. Results: We assigned 22 patients to the interventional and 21 patients to the control group. Mean age was 29.0 years (SD 11.6), 8/43 (18.6%) were female and median disease duration was 2.5 years (interquartile range: 2-5years). Compared to the control group, knowledge was 12.0% (95%CI: 5.8%-18.2%; p<0.001) higher in the interventional group. Subjects with a university degree scored 6.8% (95%CI: 3.8%-13.3%; p=0.038) higher. There was no interaction between video information and university degree. Other parameters were not associated with patient knowledge. The meta-analysis of 566 subjects enrolled in 6 studies revealed a standardized mean difference in favor of video-based education of 0.47 (95% CI: 0.30-0.64; p<0.004). Conclusion: The results suggest that supplementary video information embedded in the clinical management of keratoconus, helps conveying relevant disease knowledge.

Sacral Neuromodulation for Neurogenic Lower Urinary Tract Dysfunction.

BACKGROUND: Neurogenic lower urinary tract dysfunction (NLUTD) is a highly prevalent and disabling condition; nevertheless, standard treatments often remain unsatisfactory. Sacral neuromodulation (SNM) is a well-established therapy for non-NLUTD, but there is a lack of randomized controlled trials to show benefit in patients with NLUTD. METHODS: For this sham-controlled, double-blind, multicenter trial, patients with refractory NLUTD (and intended SNM) were recruited at four Swiss SNM referral centers. After lead placement into the sacral foramina S3 (rarely, S4), all participants underwent SNM testing. If successful (≥50% improvement in key bladder diary variables), the neurostimulator was implanted for permanent stimulation. For 2 months, neuromodulation was optimized using subsensory stimulation with individually adjusted parameters. Thereafter, the neurostimulator remained on or was switched off (1:1 random allocation to group SNM ON or SNM OFF, respectively) for 2 months, followed by a neurourologic reevaluation. The primary outcome was success, as defined above, of SNM compared with baseline. RESULTS: Of 124 patients undergoing SNM testing, 65 (52%) had successfully improved lower urinary tract function. Of these, 60 patients (median age, 49.5 years; 43 women) were randomly assigned to the intervention. After 2 months of intervention, the SNM ON group demonstrated a success rate of 76%. In the SNM OFF group, 42% of patients showed sustained SNM effects despite their neurostimulator being switched off during the last 2 months (odds ratio, 4.35; 95% confidence interval, 1.43 to 13.21; P=0.009). During the entire study period, there were 11 adverse events (6 dropouts; no dropouts during the intervention phase). CONCLUSIONS: SNM effectively corrected refractory NLUTD in the short term in well-selected neurologic patients. (Funded by the Swiss National Science Foundation, Vontobel­Stiftung, Gottfried und Julia Bangerter­Rhyner Stiftung, Dr. Urs Mühlebach, and the Swiss Continence Foundation; ClinicalTrials.gov number, NCT02165774.)