A diabetic foot classification model based on radiomics features of fundus photographs / 中华内分泌代谢杂志

Objective:To construct a diabetic foot classification prediction model based on radiomics features of fundus photographs.Methods:A total of 2 035 fundus photographs of patients with type 2 diabetes diagnosed at Nanfang Hospital between December 2011 and December 2018 were retrospectively collected [282 photographs from patients with diabetic foot(DF), and 1 753 from patients with diabetes mellitus(DM)]. All fundus photographs were randomly divided into a training set(1 424 photos) and a test set(611 photos) using a computer generated random number at 7∶3. After pre-processing the fundus photographs, a total of 4 128 texture features based on the gray matrix were extracted by the Radiomic toolkit, and 11 339 other features were extracted using the ToolboxDESC toolkit. The LASSO algorithm was used to select the 30 features most relevant to DF, and then the Bootstrap + 0.632 self-sampling method was used to further select the 7 best combinations. Logistic regression analysis was used to obtain the regression coefficients and establish the final diabetic foot classification prediction model. ROC curve was drawn, and AUC, sensitivity, specificity, and accuracy of the training and test sets were calculated to verify its prediction performance. Results:We screened 7 fundus radiomics markers for diabetic foot patients, and based on this established a DF/DM classification prediction model. The AUC, sensitivity, specificity, and accuracy of the model were 0.958 6, 0.984 0, 0.920 0, and 0.928 0 in the training set, and 0.927 1, 0.988 9, 0.881 0, and 0.896 9 in the test set, respectively.Conclusion:In this study, seven DF fundus markers were screened using radiomics technology. Based on this, a highly accurate and easy-to-use DF/DM classification model was constructed. This technology has the potential to increase the efficiency of DF screening programs.

Ocular Torsion according to Fixation in Fundus Photograph

PURPOSE: We examined the torsional change of eyeballs according to fixation using fundus photography. METHODS: We took fundus photographs of both eyes of the following patients: group 1, 10 unilateral superior oblique (SO) palsy patients; group 2, 20 exotropia (XT) patients without vertical strabismus or oblique dysfunction; and group 3, 20 normal subjects, from March 2002 to February 2005, using a fundus camera with and without fixation on the internal fixation device. We examined the torsional angle (alpha) between the horizontal line through the optic disc center and the line connecting to optic disc center with fovea using the Scion Image Program, and compared the torsional change according to fixation. RESULTS: The torsional angle (alpha) was 17.92 degrees with fixation and 18.79 degrees without fixation in paretic eyes of group 1 (p=0.46), and 8.78 degrees with fixation and 9.23 degrees without fixation in sound eyes of group 1 (p=0.36). The torsional angle was 6.35 degrees with fixation and 6.86 degrees without fixation in the right eyes of group 2 (p=0.39), and 6.40 degrees with fixation and 6.95 degrees without fixation in the left eyes of group 2 (p=0.28). In group 3, torsional angle was 6.95 degrees with fixation and 7.25 degrees without fixation in the right eyes (p=0.72), and 7.42 degrees with fixation and 7.48 degrees without fixation in the left eyes (p=0.89). Torsional angle with fixation was smaller than without fixation in all groups, but the differences were not statistically significant. CONCLUSIONS: There was no torsional change according to fixation by fundus photography in unilateral SO palsy patients, XT patients without vertical strabismus or oblique dysfunction, and normal subjects.

Ocular Torsion in Normal Korean Population

PURPOSE: We determine the normal variation in cyclotorsion, the difference between the two eyes of each individual, and the torsional changes according to age and sex. METHODS: We examined the horizontal and vertical distances from the center of the disc to the fovea and the angle with fundus photographs of 100 ophthalmologically normal subjects (200 eyes). We calculated the vertical/horizontal disc ratio and the difference of the vertical disc-foveal distance and angle between the two eyes of each individual. We studied the changes of all the measured parameters according to age and sex. RESULTS: The horizontal distance from the center of the disc to the fovea was 2.62 +/- 0.25 DD in the right eye, 2.60 +/- 0.25 DD in the left eye, and the mean was 2.61 +/- 0.25 DD. The vertical distance from the center of the disc to the fovea was 0.28 +/- 0.15 DD in the right eye, 0.30 +/- 0.15 DD in the left, and the mean was 0.29 +/- 0.15 DD. The angle was 6.11 +/- 3.21degrees in the right eye, 6.67 +/- 3.18degrees in the left, and the mean was 6.39 +/- 3.20degrees. The difference of the vertical distance and the angle between the eyes was 0.13 +/- 0.09 DD and 2.98 +/- 2.07, respectively. The parameter change according to age and sex was not statistically significant (p>0.05). CONCLUSIONS: If the fovea was from 0.01 DD above center of the disc to 0.09 DD below the lower edge of the disc, we could confirm the absence of cyclotropia. There was no significant change of any of the measured parameters according to age and sex in the normal population.

The significance of fundus photographs during health mass screening

BACKGROUND: Fundus photographs performed during health check-up were reviewed to determine the usefulness in mass screening. METHODS: Subjects consistine of 3017 people who had undergone physical health check up at the Health Clinic, Dong-Eui Medical Center in 1997 and whose fundus photographs of both eyes were taken showing clearly visible posterior pole and optic disc were chosen as subjects. Medical records which included cases of reported close exam and causes, actual cases of close exam and final diagnoses. RESULTS: Of the 3017 people, reported close exam was performed in 665 (22.0%). Among them glaucoma was suspected in 460 retinal hemorrhagic lesion in 63, retinal degenerative lesion in 67 and others in 75. There were 187 actual cases(28.1%) of close exam including glaucoma suspect in 79, retinal hemorrhagic lesion 47, retinal degenerative lesion 34 and others 27. Among 187 actual cases of close exam, 140 people (1.3%) were eventually diagnosed as glaucoma suspect (60), retinal hemorrhagic lesion (38), retinal degenerative lesion (21) and others (21). CONCLUSIONS: Fundus photographs are useful in mass screening, but performing close exam to enhance the usefulness of the fundus photographs is necessary.