Systematic PD-L1 Slide Analysis Based on Multi-Objective Learning

Abstract

In treatment of cancers, especially non-small-cell lung cancers such as lung squamous cell carcinoma(LUSC), tumor proportion score(TPS) of a programmed death-ligand 1(PD-L1) slide is essential for selecting tumor therapies. Many parameters of tumor cells(TCs) are vital to cancer diagnosis. Although the indexes can be estimated via the computational analysis, there is seldom a unified system that could acquire different nucleus information simultaneously. To address the issues, multi-objective learning pipeline(MOLP) is proposed to predict TPS, cell counts, nucleus contours and categories altogether from PD-L1 slides of LUSC. The main network comprises two branches, one estimating TPS via the cell analysis and the other directly regressing TPS. It minimizes the difference between these two approximated values of TPS to gain robustness. The cell-analysis branch increases confidence of the estimated TPS by nucleus segmentation, classification and counting. It also enables the system to estimate appearance parameters of TCs for LUSC diagnosis. Experiments on a large image set show that MOLP is feasible and effective. The TPS predicted by MOLP exhibits statistically significant correlation with pathologists' scores, with a mean absolute error(MAE) of 4.97(95% confidence interval(CI):-0.56-10.49) and a Pearson correlation coefficient(PCC) of 0.97(p < 0.001).

Keywords

programmed death-ligand 1(PD-L1) slide / tumor proportion score(TPS) / multi-objective learning / classification / segmentation / counting

References

[1] DOROSHOW D B, BHALLA S, BEASLEY M B, et al. PD-L1 as a biomarker of response to immune-checkpoint inhibitors[J]. Nature Reviews Clinical Oncology, 2021,18:345-362.

[2] WANG J S, MAO X T, WANG Y, et al. Automatic generation of pathological benchmark dataset from hyperspectral images of double stained tissues[J]. Optics & Laser Technology, 2023,163:109331.

[3] LI Q L, HE X F, WANG Y T, et al. Review of spectral imaging technology in biomedical engineering:achievements and challenges[J]. Journal of Biomedical Optics, 2013, 18(10):100901.

[4] MI H B, XU K L, XIANG Y, et al. A quantitative analysis platform for PD-L1 immunohistochemistry based on point-level supervision model[C]//Proceedings of the Twenty-Eighth International Joint Conference on Artificial Intelligence. [S.l.]: International Joint Conferences on Artificial Intelligence Organization, 2019:6554-6556.

[5] PAN B J, KANG Y X, JIN Y, et al. Automated tumor proportion scoring for PD-L1 expression based on multistage ensemble strategy in non-small cell lung cancer[J]. Journal of Translational Medicine, 2021, 19(1):249.

[6] KANG Y X, LI H S, HAN X, et al. Automated tumor proportion scoring for assessment of PD-L1 expression based on multi-stage ensemble strategy[M]//Machine Learning in Medical Imaging. Cham: Springer Nature Switzerland AG, 2020:70-79.

[7] YI J R, WU P X, JIANG M L, et al. Attentive neural cell instance segmentation[J]. Medical Image Analysis, 2019,55:228-240.

[8] GRAHAM S, VU Q D, RAZA S E A, et al. Hover-Net:simultaneous segmentation and classification of nuclei in multi-tissue histology images[J]. Medical Image Analysis, 2019,58:101563.

[9] DOGAR G M, SHAHZAD M, FRAZ M M. Attention augmented distance regression and classification network for nuclei instance segmentation and type classification in histology images[J]. Biomedical Signal Processing and Control, 2023,79:104199.

[10] LIU J X, ZHENG Q, MU X, et al. Automated tumor proportion score analysis for PD-L1 (22C3) expression in lung squamous cell carcinoma[J]. Scientific Reports, 2021,11:15907.

[11] WIDMAIER M, WIESTLER T, WALKER J, et al. Comparison of continuous measures across diagnostic PD-L1 assays in non-small cell lung cancer using automated image analysis[J]. Modern Pathology, 2020, 33(3):380-390.

[12] CHEN J H, ZHU Y A, CHEN Z. Graph-embedded online learning for cell detection and tumour proportion score estimation[J]. Electronics, 2022, 11(10):1642.

[13] WU J H, LIU C L, LIU X Q, et al. Artificial intelligence-assisted system for precision diagnosis of PD-L1 expression in non-small cell lung cancer[J]. Modern Pathology, 2022, 35(3):403-411.

[14] VALANARASU J M J, PATEL V M. UNeXt:MLP-based rapid medical image segmentation network[M]//Lecture Notes in Computer Science. Cham: Springer Nature Switzerland AG, 2022:23-33.

[15] MILLETARI F, NAVAB N, AHMADI S A. V-net:fully convolutional neural networks for volumetric medical image segmentation[C]//2016 Fourth International Conference on 3D Vision (3DV). New York: IEEE, 2016:565-571.

[16] ZHANG Q, WANG Y, QIU S, et al. 3D-PulCNN:pulmonary cancer classification from hyperspectral images using convolution combination unit based CNN[J]. Journal of Biophotonics, 2021, 14(12):e202100142.

[17] KINGMA D P, BA J. Adam:a method for stochastic optimization[EB/OL].(2017-01-30)[2023-07-01]. https://arxiv.org/abs/1412.6980v9.

[18] CHEN Z, CHEN Z, LIU J X, et al. Weakly supervised histopathology image segmentation with sparse point annotations[J]. IEEE Journal of Biomedical and Health Informatics, 2021, 25(5):1673-1685.

[19] RONNEBERGER O, FISCHER P, BROX T. U-net:convolutional networks for biomedical image segmentation[M]//Lecture Notes in Computer Science. Cham: Springer International Publishing Switzerland, 2015:234-241.

[20] ZHOU Z W, SIDDIQUEE M M R, TAJBAKHSH N, et al. UNet++:redesigning skip connections to exploit multiscale features in image segmentation[J]. IEEE Transactions on Medical Imaging, 2020, 39(6):1856-1867.

[21] HE K M, ZHANG X Y, REN S Q, et al.Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). New York: IEEE, 2016:770-778.

[22] WILLMOTT C J, MATSUURA K. Advantages of the mean absolute error (MAE) over the root mean square error (RMSE) in assessing average model performance[J]. Climate Research, 2005,30:79-82.

[23] LEE RODGERS J, NICEWANDER W A. Thirteen ways to look at the correlation coefficient[J]. The American Statistician, 1988, 42(1):59-66.

[24] NASER M Z, ALAVI A H. Error metrics and performance fitness indicators for artificial intelligence and machine learning in engineering and sciences[J]. Architecture,Structures and Construction, 2023, 3(4):499-517.

[25] CARASS A, ROY S, GHERMAN A, et al. Evaluating white matter lesion segmentations with refined S&#xF8;rensen-dice analysis[J]. Scientific Reports, 2020,10:8242.

[26] HINTZE J L, NELSON R D. Violin plots:a box plot-density trace synergism[J]. The American Statistician, 1998, 52(2):181-184.

Funding

National Natural Science Foundation of China(61702094 and 62301142)

“Chenguang Program” Supported by Shanghai Education Development Foundation and Shanghai Municipal Education Commission, China(18CG38)