Table 1 Characteristics of patients used in this study.

From: Fully automated deep learning based auto-contouring of liver segments and spleen on contrast-enhanced CT images

Cohorts (Number of images)

Treatment type

Used for

Image

Cancer types** (Number of patients)

Median Voxel size (in x/y, mm)

Median Voxel size (in z, mm)

\({{\text{C}}}_{{\text{RTTrain}}}\)(N = 160)

Radiotherapy

Training

Contrast

HCC = 30 CC = 49 CRM = 12 mixed = 4

0.98 (0.66–1.17)

2.5 (0.63–5.0)

\({{\text{C}}}_{{\text{RTVal}}}\)(N = 40)

Radiotherapy

Validation

Contrast

HCC = 16 CC = 18 CRM = 2

0.98 (0.7–1.07)

2.5 (2.5–5.0)

\({{\text{C}}}_{{\text{LS}}}\)(N = 33)

Liver surgery evaluation

Test

Contrast

HCC = 2 CC = 5 CRM = 22

0.86 (0.51–0.98)

2.5 (2.5–5.0)

\({{\text{C}}}_{{\text{PVE}}}\)(N = 20)

Portal vein embolization

Test

Contrast

HCC = 0 CC = 3 CRM = 17

0.80 (0.70–0.98)

2.5 (1.0–2.5)

\({{\text{C}}}_{{\text{CH}}}\)(N = 25)

Liver segmentation challenge

Test

Contrast

NA

0.71 (0.60–0.96)

2 (1.0–5.0)

\({{\text{C}}}_{{\text{CNC}}}\)(N = 20)

Pre-contrast of \({{\text{C}}}_{{\text{RT}}}\)

Perturbation study

Non-contrast

HCC = 6 CC = 10 CRM = 3 mixed = 1

0.98 (0.98–1.17)

2.5 (all*)

  1. Median (min–max).
  2. *All means all the cases showed same values.
  3. **CRM colorectal or other metastasis, CC Cholangiocarcinoma, HCC Hepatocellular carcinoma, mixed more than one cancer types.