Table 1 Characteristics of cases in the KPMP, JP-AID, and UT cohorts.

From: Multiple instance learning using pathology foundation models effectively predicts kidney disease diagnosis and clinical classification

 

KPMP

JP-AID

UT

HC

AIN

DKD

HCc

AIN

DKD

HC

AIN

DKD

N

21

12

70

21

9

9

39

17

27

Age, year

 10–19

0 (0.0)

0 (0.0)

0 (0.0)

1 (4.8)

2 (22.2)

0 (0.0)

0 (0.0)

1 (5.9)

0 (0.0)

 20–29

2 (9.5)

0 (0.0)

0 (0.0)

7 (33.3)

1 (11.1)

0 (0.0)

0 (0.0)

1 (5.9)

0 (0.0)

 30–39

6 (28.6)

7 (58.3)

6 (8.6)

4 (19.0)

1 (11.1)

1 (11.1)

2 (5.1)

1 (5.9)

3 (17.6)

 40–49

7 (33.3)

1 (8.3)

12 (17.1)

3 (14.3)

0 (0.0)

0 (0.0)

8 (20.5)

1 (5.9)

4 (23.5)

 50–59

6 (28.6)

4 (33.3)

15 (21.4)

4 (19.0)

1 (11.1)

2 (22.2)

11 (28.2)

4 (23.5)

2 (11.8)

 60–69

0 (0.0)

0 (0.0)

29 (41.4)

2 (9.5)

1 (11.1)

3 (33.3)

13 (33.3)

1 (5.9)

12 (70.6)

 70–79

0 (0.0)

0 (0.0)

8 (11.4)

0 (0.0)

1 (11.1)

4 (44.4)

4 (10.3)

6 (35.3)

4 (23.5)

 80–89

0 (0.0)

0 (0.0)

0 (0.0)

0 (0.0)

2 (22.2)

0 (0.0)

1 (2.6)

1 (5.9)

2 (11.8)

Sex

 Male

10 (47.6)

10 (83.3)

41 (58.6)

13 (61.9)

6 (66.7)

6 (66.7)

15 (38.5)

5 (29.4)

15 (55.6)

 Female

11 (52.4)

2 (16.7)

29 (41.4)

8 (38.1)

3 (33.3)

3 (33.3)

24 (61.5)

12 (70.6)

12 (44.4)

 Diabetes

0 (0.0)

3 (25)

70 (100)

NA

NA

NA

1 (2.6)

3 (17.6)

27 (100)

 Hypertension

4 (19.0)

4 (33.3)

65 (92.9)

NA

NA

NA

9 (23.1)

4 (23.5)

22 (81.5)

 eGFR

NA

79.5 (74.5–89.5)a

44.5 (34.5–64.5)a

NA

NA

NA

80.9 (66.6–86.2)

22.2 (18.6–33.0)

41.4 (25.8–52.4)

Proteinuriab

 Overt

0 (0.0)

2 (16.7)

44 (62.9)

0 (0.0)

0 (0.0)

0 (0.0)

0 (0.0)

8 (47.1)

22 (81.5)

 Mild

0 (0.0)

1 (8.3)

17 (24.3)

0 (0.0)

0 (0.0)

0 (0.0)

0 (0.0)

8 (47.1)

5 (18.5)

 Absent

0 (0.0)

2 (16.7)

6 (8.6)

0 (0.0)

0 (0.0)

0 (0.0)

39 (100)

1 (5.9)

0 (0.0)

 NA

21 (100)

7 (58.3)

3 (4.3)

21 (100)

9 (100)

9 (100)

0 (0.0)

0 (0.0)

0 (0.0)

  1. KPMP the kidney precision medicine project, JP-AID Japan-pathology artificial intelligence diagnostics project, UT the University of Tokyo Hospital, HC healthy control, AIN acute interstitial nephritis, DKD diabetic kidney disease, NA not available, KDIGO kidney disease: improving global outcomes, AKI acute kidney injury.
  2. aSince eGFR data in the KPMP dataset were provided as categorical data (e.g. 40–50 ml/min/1.73 m2), the midpoint of each range was used as the estimated value. For the AIN group, the eGFR values represented baseline values.
  3. bProteinuria severity was classified as follows:
  4. Overt: proteinuria ≥ 1000 mg/gCre or albuminuria ≥ 300 mg/gCre.
  5. Mild: 150 mg/gCre ≤ proteinuria < 1000 mg/gCre or 30 mg/gCre ≤ albuminuria < 300 mg/gCre Absent: proteinuria < 150 mg/gCre or albuminuria < 30 mg/gCre.
  6. If both proteinuria and albuminuria data were available, the more overt category was assigned.
  7. cThe demographic information for the transplanted kidneys in the JP-AID HC group may originate from the recipients rather than the donors. Since the database does not provide detailed clarification, this remains uncertain and should be interpreted with caution.
  8. Data are expressed as median (interquartile range) or n (%).
Back to article page