Table of Contents of Supplemental Material
Supplemental Table 1. Manual annotation of biopsy images
Supplemental Table 2. Coding of soft labels for deep learning training
Supplemental Table 3. Kidney failure risk, calculated by univariable and multivariable Cox Regression analysis of clinical characteristics. Clinical outcome is time to kidney failure.
Supplemental Table 4. Data table of the predictive scores and outcome of the biopsies of the
Final Test set (n = 95 biopsies). Deep Learning Predictive Score (DLPS), MEST-C score, IgAN
Prediction Tool (IIPT), Clinical Decision Support System for Estimating the Risk of End-Stage
Kidney Disease in IgA Nephropathy’(CDSS) and Outcome (kidney failure during the follow up).
Supplemental Table 1. Manual annotation of biopsy images Feature
Biopsies, n 496
Images, n 1,369
Patches, n 7199
Mean images / biopsy sections, n 3 (2-4)
Compartmentalization by manual annotation
Capsular area per image, µm
253,062 (0 – 452,594)
Cortical area per image µm
24,734,348 (3,580,022 – 6,807,961)
Medullary area per image µm
20 (0 – 3,045,781)
Data are expressed as median (25th – 75th percentile), unless otherwise indicated.
Supplemental Table 2. Coding of soft labels for deep learning training
Clinical features Label formula Label
Kidney failure during the follow up
Time to kidney failure (TtKF), years
Follow up duration (f-up), years
(range value)
False NA > 5 0 0
False NA ≤ 5 0.5 – (f-up/10) 0.5 – 1
True < 5 NA 1 1
True
≥5 and <10NA 0.5 +((10 – TtKF) / 10) 0.5 – 1
True
≥10NA 0 0
NA = not applicable, TtKF = time to kidney failure, f-up = follow up
Supplemental Table 3. Kidney failure risk, calculated by univariate and multivariate Cox Regression analysis of clinical characteristics. Clinical outcome is time to kidney failure.
Univariable Analysis
Multivariable Analysis
Parameters Hazard
Ratio
p value Hazard Ratio
p value
Creatinine, mg/dl 2.30 <0.001 2.30 <0.001 Urine Protein/Creatinine ratio 1.80 <0.001 1.59 <0.001 Mean Blood Pressure, mmHg 1.02 <0.001 1.02 <0.001
Age, yrs 1.012 0.05 1.00 0.83
Sex 0.65 0.09 0.91 0.77
Sex coded as male = 1 and female = 2
Supplemental Table 4. Data table of the predictive scores and outcome of the biopsies of the Final Test set (n = 95 biopsies). Deep Learning Predictive Score (DLPS), MEST-C score, IgAN Prediction Tool (IIPT), Clinical Decision Support System for Estimating the Risk of End-Stage Kidney Disease in IgA Nephropathy’(CDSS) and Outcome (kidney failure during the follow up).
BIOPSY ID
DLPS IIPT DCSS M E S T C OUTCOME
20 0.20 11.19 81.16 0 0 0 1 1 0
21 0.12 6.26 83.60 0 0 0 0 0 0
24 0.27 6.41 83.92 1 1 0 0 0 0
25 0.18 5.51 87.20 0 1 0 0 0 0
52 0.47 5.51 89.92 0 1 0 0 1 0
59 0.28 5.51 89.92 0 1 0 0 1 0
72 0.41 6.06 83.30 0 1 1 0 0 0
73 0.64 11.19 74.79 0 0 0 1 0 0
86 0.56 7.04 78.56 1 1 1 0 0 1
88 0.20 5.51 89.92 0 1 0 0 1 0
89 0.73 12.53 74.44 1 1 1 1 1 1
96 0.38 5.51 89.92 0 1 0 0 1 0
145 0.69 6.06 85.45 0 1 1 0 2 1
156 0.30 6.06 85.45 0 1 1 0 1 0
192 0.15 5.51 87.20 0 1 0 0 0 0
203 0.74 6.06 85.45 0 1 1 0 1 1
213 0.20 5.51 87.20 0 1 0 0 0 0
223 0.17 12.53 66.48 1 1 1 1 0 0
229 0.39 6.41 88.67 1 1 0 0 1 0
235 0.17 6.26 83.60 0 0 0 0 0 0
244 0.36 12.26 64.01 0 0 1 1 0 1
251 0.22 6.26 87.45 0 0 0 0 1 1
256 0.79 12.53 74.44 1 1 1 1 1 1
261 0.14 7.04 83.51 1 1 1 0 1 0
285 0.12 6.06 85.45 0 1 1 0 1 0
313 0.28 6.26 83.60 0 0 0 0 0 0
324 0.79 12.53 74.44 1 1 1 1 1 1
327 0.56 18.56 57.64 0 1 1 2 0 1
328 0.52 6.06 85.45 0 1 1 0 1 0
330 0.19 6.41 83.92 1 1 0 0 0 0
342 0.45 6.06 83.30 0 1 1 0 0 0
348 0.77 10.84 77.91 0 1 1 1 2 1
385 0.25 6.26 83.60 0 0 0 0 0 0
413 0.79 12.26 72.01 0 0 1 1 1 1
419 0.46 5.51 87.20 0 1 0 0 0 0
426 0.36 9.89 84.46 0 1 0 1 1 0
435 0.37 5.51 87.20 0 1 0 0 0 0
461 0.59 12.26 64.01 0 0 1 1 0 0
862 0.48 20.86 50.58 0 0 1 2 0 0
899 0.91 12.53 66.48 1 1 1 1 0 1
903 0.63 7.27 76.21 1 0 0 0 0 0
910 0.75 23.91 61.50 1 0 1 2 0 0
991 0.32 10.84 72.27 0 1 1 1 0 0
1157 0.56 23.91 54.26 1 0 1 2 1 1
1192 0.54 7.98 77.13 1 0 1 0 1 0
1239 0.34 6.26 83.60 0 0 0 0 0 0
1243 0.48 7.98 77.13 1 0 1 0 1 0
1287 0.17 7.98 67.85 1 0 1 0 0 0
1299 0.24 11.19 74.79 0 0 0 1 0 0
1323 0.56 20.86 50.58 0 0 1 2 0 1
1431 0.59 12.53 74.44 1 1 1 1 1 0
1464 0.40 7.98 77.13 1 0 1 0 1 0
1502 0.16 12.53 74.44 1 1 1 1 1 0
1845 0.32 11.19 74.79 0 0 0 1 0 1
1858 0.07 7.04 83.51 1 1 1 0 1 0
1985 0.16 6.06 83.30 0 1 1 0 0 0
2021 0.31 7.98 67.85 1 0 1 0 0 0
2122 0.12 7.04 78.56 1 1 1 0 0 0
2218 0.38 7.04 83.51 1 1 1 0 1 0
2234 0.26 12.53 74.44 1 1 1 1 1 1
2255 0.13 7.04 78.56 1 1 1 0 0 0
2308 0.77 21.96 54.18 1 0 0 2 0 1
2401 0.21 12.53 74.44 1 1 1 1 1 1
2510 0.12 12.53 74.44 1 1 1 1 1 0
2577 0.03 7.98 67.85 1 0 1 0 0 0
2649 0.19 6.88 80.92 0 0 1 0 1 0
2666 0.63 12.53 74.44 1 1 1 1 1 1
2722 0.38 12.26 64.01 0 0 1 1 0 1
2777 0.52 12.26 64.01 0 0 1 1 0 1
2801 0.09 7.98 67.85 1 0 1 0 0 0
2874 0.12 7.04 83.51 1 1 1 0 1 0
2891 0.37 12.26 64.01 0 0 1 1 0 1
2905 0.53 12.94 65.63 1 0 0 1 0 0
2921 0.16 6.88 77.40 0 0 1 0 0 1
2972 0.26 6.26 83.60 0 0 0 0 0 0
2975 0.16 10.84 72.27 0 1 1 1 0 0
3011 0.20 7.27 76.21 1 0 0 0 0 0
3018 0.19 7.04 78.56 1 1 1 0 0 0
3032 0.50 7.98 67.85 1 0 1 0 0 0
3180 0.45 NA NA 0 1 NA 0 0 0
3228 0.52 7.98 67.85 1 0 1 0 0 1
3333 0.16 NA NA 0 0 NA 0 0 0
3368 0.28 6.06 85.45 0 1 1 0 1 0
3413 0.47 6.88 77.40 0 0 1 0 0 0
3426 0.81 6.88 80.92 0 0 1 0 1 1
3435 0.41 6.26 83.60 0 0 0 0 0 0
3530 0.16 6.06 83.30 0 1 1 0 0 0
3624 0.50 7.04 78.56 1 1 1 0 0 1
3723 0.51 NA NA NA 0 1 2 1 1
3839 0.75 10.84 77.91 0 1 1 1 1 1
3885 0.67 6.88 77.40 0 0 1 0 0 1