Table 1 Single-domain performance evaluation of HCIA compared to other SOTA models. Optimal results are highlighted in bold.

From: Hierarchical contextual information aggregation for polyp segmentation

Methods

Kvasir-SEG

CVC-ClinicDB

CVC-ColonDB

mDice

mIoU

\(S_{\alpha }\)

\(mE_{\xi }\)

mDice

mIoU

\(S_{\alpha }\)

\(mE_{\xi }\)

mDice

mIoU

\(S_{\alpha }\)

\(mE_{\xi }\)

U-Net4

0.815

0.742

0.854

0.877

0.818

0.751

0.883

0.907

0.550

0.465

0.720

0.724

UNet++5

0.817

0.740

0.858

0.882

0.792

0.727

0.873

0.890

0.498

0.411

0.685

0.718

MedSAM39

0.862

0.795

-

-

0.867

0.803

-

-

0.734

0.651

-

-

PraNet10

0.895

0.844

0.909

0.941

0.895

0.844

0.936

0.962

0.701

0.632

0.814

0.835

DCRNet40

0.889

0.825

0.911

0.935

0.893

0.840

0.933

0.965

0.723

0.650

0.817

0.862

MADGNet41

0.907

0.853

0.856

0.947

0.939

0.895

0.922

0.985

0.775

0.697

0.833

0.880

MMFIL-Net42

0.909

0.858

0.915

0.948

0.890

0.838

0.922

0.971

0.744

0.659

0.824

0.864

Polyp-LVT43

0.909

0.851

-

0.941

0.935

0.882

-

0.982

-

-

-

-

APCNet44

0.913

0.859

0.919

-

0.934

0.886

0.945

-

0.758

0.682

0.837

-

EFA-Net45

0.914

0.861

0.929

0.955

0.919

0.871

0.943

0.972

0.774

0.696

0.855

0.884

CTNet24

0.917

0.863

0.928

0.959

0.936

0.887

0.952

0.983

0.813

0.734

0.874

0.915

Polyp-PVT46

0.918

0.861

0.924

0.955

0.938

0.886

0.947

0.981

0.807

0.726

0.864

0.913

Polyp-Mamba16

0.919

0.867

0.951

0.968

0.941

0.896

0.970

0.987

0.791

0.713

0.846

0.902

Transfuse19

0.920

0.871

0.926

0.959

0.918

0.868

0.935

0.972

0.702

0.782

0.822

0.886

SRaNet47

0.921

0.870

0.932

-

0.926

0.875

0.950

-

0.814

0.734

0.877

-

BUNet48

0.923

0.873

0.926

0.965

0.935

0.890

0.948

0.989

0.813

0.731

0.861

0.911

SSFormer-S49

0.926

0.874

-

-

0.926

0.875

-

-

0.771

0.711

-

-

DSHNet50

0.929

0.881

0.936

0.965

0.942

0.896

0.954

0.987

0.815

0.733

0.870

0.915

UM-Net15

0.930

0.925

0.938

-

-

-

-

-

0.761

0.828

0.863

-

CAFE-Net23

0.933

0.889

0.939

0.967

0.943

0.899

0.957

0.986

0.820

0.740

0.874

0.914

MGCBFormer51

0.933

0.887

-

-

0.936

0.899

-

-

0.731

0.807

-

-

SSFormer-L49

0.935

0.890

-

-

0.944

0.899

-

-

0.804

0.722

-

-

FCBFormer52

0.938

0.890

0.945

0.940

0.946

0.902

0.952

0.944

0.777

0.696

0.841

0.893

CTHP22

0.939

0.891

-

-

0.947

0.902

-

-

-

-

-

-

HCIA(ours)

0.942

0.894

0.949

0.968

0.948

0.904

0.959

0.990

0.823

0.830

0.874

0.920

Back to article page