Page 114 - AIH-2-2

Artificial Intelligence in Health AI in early breast cancer diagnosis: A review

Table 4. Results of validation metrics from computer‑aided detection studies
Approach to Authors Year Work Dataset Classification results
CAD Method ACC PRE REC SPE SEN F1 AUC
Deep learning 17 2022 Automated segmentation and Ultrasound images, 0.978 0.981 0.987 0.946 NR 0.984 0.99
classification of breast tumors BUSI dataset
from ultrasound images
Deep learning 18 2021 Architectural PINUM 0.95 0.9 0.89 NR 0.99 0.87 0.91
distortion-based digital CBIS-DDSM 0.97 0.94 0.98 NR 0.95 0.96 NR
mammogram classification
DDSM 0.98 0.96 0.86 NR 0.96 0.9 0.85
Deep learning 19 2023 Artificial SW 0.944 NR NR 0.454 0.945 0.254 0.945
intelligence-assisted TS NR NR NR NR 100 NR 0.956
ultrasound image analysis
DZ NR NR NR NR 80 NR 0.907
Deep learning 20 2020 Mitosis detection in breast ICPR 2012 NR 0.876 0.841 NR NR 0.858 NR
cancer histopathology ICPR 2014 NR 0.848 0.583 NR NR 0.691 NR
images
TUPAC16 NR 0.641 0.642 NR NR 0.642 NR
Deep learning 22 2022 Improved mammographic Digital Database 1 NR NR 1 1 1 1
breast mass classification for Screening
Mammography
(DDSM)
INbreast 0.9994 NR NR 0.9992 0.9996 0.9995 0.9994
Mammographic 0.9993 NR NR 1 0.9987 0.9989 0.9993
Image Analysis
Society (MIAS)
Mixed 0.9998 NR NR 0.9997 1 0.9998 0.9998
Deep learning 24 2023 Rapid diagnosis of ductal 463 BC, 241 healthy NR NR NR NR NR NR NR
carcinoma in situ and breast
cancer
Deep learning 25 2021 Tumor segmentation in A private from NR NR NR 0.9249 NR 0.8103 0.7368
breast ultrasound image Xinhua Hospital
Deep learning 26 2022 Ensemble network BCWD 0.9632 NR NR 0.9392 0.9776 0.951 0.9805
model for classifying and BCWO 0.97 NR NR 0.9715 0.9672 0.9811 0.9902
predicting breast cancer
BCC 0.9671 NR NR 0.9716 0.9588 0.9772 0.9864
Deep learning 27 2022 Intelligent breast cancer INbreast 0.98 0.969 NR 0.97 0.95 NR NR
detection
Machine learning 21 2019 Decision support for breast Wisconsin Breast 0.9777 NR NR 0.9777 0.9808 NR NR
cancer detection Cancer Database 0.965 NR NR 0.981 0.9351 NR NR
0.9748 NR NR 0.973 0.978 NR NR
Protein microarray 0.983 NR NR 0.972 0.9867 NR NR
database 0.958 NR NR 0.956 0.9587 NR NR
0.973 NR NR 0.92 0.9907 NR NR
Machine learning 23 2022 Breast cancer diagnosis Mini-MIAS 95.2 98.6 NR 98.8 91.6 NR 95.43
improvement based on Screening
image processing mammography data
Machine learning 28 2023 Filtering and classifying MIAS 96.2 NR NR NR NR NR 0.96
mammographic DDSM 99.3 NR NR NR NR NR 0.99
microcalcification images in
early cancer detection
Machine learning 29 2022 Prediction of breast cancer Coimbra dataset 80.23 82.71 78.57 78.57 NR 0.78
from risk factors
Abbreviations: ACC: Accuracy; AUC: Area under the curve; BC: Breast cancer; BCC: Breast Cancer Coimbra dataset; BCWD: Breast Cancer
Wisconsin Diagnostic; BCWO: Breast Cancer Wisconsin (Original); BUSI: Breast Ultrasound Images Dataset; CBIS-DDSM: Curated Breast Imaging
Subset DDSM Dataset; DDSM: Digital Database for Screening Mammography; DZ: Dazu People’s Hospital; ICPR: International Conference on
Pattern Recognition; MIAS: Mammographic Image Analysis Society; NR: Not reported; PRE: Precision; REC: Recall; SEN: Sensitivity; SPE: Specificity;
SW: First Affiliated Hospital of Army Medical University; TS: Tangshan People’s Hospital; TUPAC: Tumor Proliferation Assessment Challenge.

Volume 2 Issue 2 (2025) 108 doi: 10.36922/aih.4197